TIBCO ActiveMatrix Adapter for Database ... - TIBCO Software

TIBCO ActiveMatrix Adapter™ for Database

User’s GuideSoftware Release 5.5.1July 2011

Important Information

SOME TIBCO SOFTWARE EMBEDS OR BUNDLES OTHER TIBCO SOFTWARE. USE OF SUCH EMBEDDED OR BUNDLED TIBCO SOFTWARE IS SOLELY TO ENABLE THE FUNCTIONALITY (OR PROVIDE LIMITED ADD-ON FUNCTIONALITY) OF THE LICENSED TIBCO SOFTWARE. THE EMBEDDED OR BUNDLED SOFTWARE IS NOT LICENSED TO BE USED OR ACCESSED BY ANY OTHER TIBCO SOFTWARE OR FOR ANY OTHER PURPOSE.USE OF TIBCO SOFTWARE AND THIS DOCUMENT IS SUBJECT TO THE TERMS AND CONDITIONS OF A LICENSE AGREEMENT FOUND IN EITHER A SEPARATELY EXECUTED SOFTWARE LICENSE AGREEMENT, OR, IF THERE IS NO SUCH SEPARATE AGREEMENT, THE CLICKWRAP END USER LICENSE AGREEMENT WHICH IS DISPLAYED DURING DOWNLOAD OR INSTALLATION OF THE SOFTWARE (AND WHICH IS DUPLICATED IN LICENSE.PDF) OR IF THERE IS NO SUCH SOFTWARE LICENSE AGREEMENT OR CLICKWRAP END USER LICENSE AGREEMENT, THE LICENSE(S) LOCATED IN THE “LICENSE” FILE(S) OF THE SOFTWARE. USE OF THIS DOCUMENT IS SUBJECT TO THOSE TERMS AND CONDITIONS, AND YOUR USE HEREOF SHALL CONSTITUTE ACCEPTANCE OF AND AN AGREEMENT TO BE BOUND BY THE SAME.This document contains confidential information that is subject to U.S. and international copyright laws and treaties. No part of this document may be reproduced in any form without the written authorization of TIBCO Software Inc.TIB, TIBCO, TIBCO Adapter, Predictive Business, Information Bus, The Power of Now, TIBCO ActiveMatrix BusinessWorks, TIBCO Administrator, TIBCO Designer, TIBCO Hawk, TIBCO IntegrationManager, TIBCO Rendezvous, and TIBCO Runtime Agent are either registered trademarks or trademarks of TIBCO Software Inc. in the United States and/or other countries.EJB, Java EE, J2EE, and all Java-based trademarks and logos are trademarks or registered trademarks of Sun Microsystems, Inc. in the U.S. and other countries.All other product and company names and marks mentioned in this document are the property of their respective owners and are mentioned for identification purposes only.THIS SOFTWARE MAY BE AVAILABLE ON MULTIPLE OPERATING SYSTEMS. HOWEVER, NOT ALL OPERATING SYSTEM PLATFORMS FOR A SPECIFIC SOFTWARE VERSION ARE RELEASED AT THE SAME TIME. SEE THE README.TXT FILE FOR THE AVAILABILITY OF THIS SOFTWARE VERSION ON A SPECIFIC OPERATING SYSTEM PLATFORM.THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT.THIS DOCUMENT COULD INCLUDE TECHNICAL INACCURACIES OR TYPOGRAPHICAL ERRORS. CHANGES ARE PERIODICALLY ADDED TO THE INFORMATION HEREIN; THESE CHANGES WILL BE INCORPORATED IN NEW EDITIONS OF THIS DOCUMENT. TIBCO SOFTWARE INC. MAY MAKE IMPROVEMENTS AND/OR CHANGES IN THE PRODUCT(S) AND/OR THE PROGRAM(S) DESCRIBED IN THIS DOCUMENT AT ANY TIME.THE CONTENTS OF THIS DOCUMENT MAY BE MODIFIED AND/OR QUALIFIED, DIRECTLY OR INDIRECTLY, BY OTHER DOCUMENTATION WHICH ACCOMPANIES THIS SOFTWARE, INCLUDING BUT NOT LIMITED TO ANY RELEASE NOTES AND "READ ME" FILES.Copyright © 1999-2011 TIBCO Software Inc. ALL RIGHTS RESERVED.TIBCO Software Inc. Confidential Information

| iii

Contents

Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Related Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviTIBCO ActiveMatrix Adapter for Database Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviOther TIBCO Product Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvi

Typographical Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

Connecting with TIBCO Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxHow to Join TIBCOmmunity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxHow to Access All TIBCO Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxHow to Contact TIBCO Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx

Chapter 1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Adapter Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Publication and Subscription Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Request-Response Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Message Transports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6TIBCO Rendezvous Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6JMS Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Adapter Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8Run-time Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8TIBCO Runtime Agent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Polling or Alerter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Referencing External Schemas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Exception Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10Loop Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Master-Master Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11Multi-file Format Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12DAT File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Fault Tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Supported SQL Operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

TIBCO ActiveMatrix Adapter for Database User’s Guide

iv | Contents

Publication Service Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Publish by Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Publish by Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15Update Feature for Parent-Child Publication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Changing a Publication Subject Without Using TIBCO Designer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16Message Grouping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Checking for Duplicate Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Message Logging Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Timely Processing of Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Chapter 2 Getting Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Required Platform and Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Before Starting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Create the Database Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Configure the Properties File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Start the Adapter Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Change the Table Values and Receive Notification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

Exit the Query Tool and Adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

Clean Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

Chapter 3 Adapter Instance Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Configuration Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Saving the Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Testing the Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Notes on Configuring an Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36Configuration Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Changing an Existing Configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37Using Global Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Adapter Instance Tabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Configuration Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Design-time Connection Tab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Using Connection Settings Templates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Saving Connection Settings for Reuse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49Using Saved Connection Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Run-time Connection Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51


Contents | v

Adapter Services Tab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53All Publication Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53All Subscription Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55All Request-Response Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

General Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Logging Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

Startup Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

Monitoring Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Chapter 4 Adapter Service Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Quality of Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68Delivery Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Wire Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Publication Service Tabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71Configuration Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71Tables Tab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74Adding Child Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76DB2/OS390 Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80DB2/AS400 Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Publisher Options Tab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Advanced Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

Subscription Service Tabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87Configuration Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87Table Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90Child Table Mappings Tab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92Child Exception Table Mappings Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Subscriber Options for DB2 on OS/390 (z/OS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Subscriber Options Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94Advanced Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Request-Response Service Tabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97Configuration Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97Call Operation Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100Advanced Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Chapter 5 Deploying and Starting the Adapter Using TIBCO Administrator . . . . . . . . . . . . . 103

Create an EAR File in TIBCO Designer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

Deploy the Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Start or Stop the Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

Monitor the Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107


vi | Contents

Chapter 6 Using an Alerter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

Setting Up and Starting the Alerter on Oracle or Microsoft SQL Server 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . 111Before Starting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113Configuring and Starting the Alerter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113Oracle Alerter Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

Setting Up and Starting the Alerter on Microsoft SQL Server 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Before Starting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Alerter Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115Using the Alerter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Setting Up and Starting the Alerter on Sybase SQL Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118Before Starting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118Alerter Setup on Solaris. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118Alerter Setup on Microsoft Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119Using the Alerter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

Chapter 7 Using the Request-Response Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

Request-Response Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126TIBCO ActiveEnterprise or XML Message Request Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127TIBCO Rendezvous Message Request Format. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127TIBCO ActiveEnterprise or XML Message Response Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130TIBCO Rendezvous Message Response Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

RPC Mode. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134Standard RPC Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135Custom RPC Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139Implementing Client RPC Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

Load Balancing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144Load Balancing in a Single Instance Using Threads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144Load Balancing Across Adapter Instances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

Oracle REF Data Type Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Chapter 8 Advanced Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153

Publishing Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154Removing Records from the Publication Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156Publication Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156Changing the Publication Trigger to Publish a Subset of Rows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

Source Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

Exception Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161Child Exception Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162Using an Exception Table as a Source Table. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163


Contents | vii

Opaque Exception Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

Incremental Parent-child Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

Database Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168Oracle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168SQL Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168Sybase and Sybase Adaptive Server Anywhere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169DB2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169Mapping to Database Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

Subject Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171Parameterized Subject Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

Preregistering a Certified Subscriber. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

Changing the Location of the Ledger File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

Sending Messages to an Adapter Instance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177adbDateTime Object. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177

Publishing by Reference Object. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179Changing Repository Objects for Parent-Child Relationships . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185Table to Record Sequence Numbers (DB2 on iSeries) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

User Callout Library . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187Using the alterMsgPub and alterMsgSub Functions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188Using the adbPreCommit Function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189Building the Callout Library. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

Unique Connection Identifier for Oracle Databases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193

Publishing in a Commitment Controlled Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

Using the Adapter with a Revision Control System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

Using a Log File for an Adapter Instance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198Fine-Tuning Tracing Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198Log File Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200Using Advanced Logging Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200

Using Database Cleanup Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203

Change the SQL Statement Terminator (DB2 for z/OS Only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

Fault Tolerance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206

SSL Data Encryption using DataDirect ODBC Drivers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

Chapter 9 Setting Encoding Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210

Setting TIBCO Messaging Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212

Configuring the Adapter to Communicate with the Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214Configuring the Adapter at Design-time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214Special Configurations for Supporting a UTF-8 Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217


viii | Contents

Run-time Environment Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218

Relevant Environment Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

Chapter 10 Monitoring the Adapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222

Starting TIBCO Hawk Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

The Auto-Discovery Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224

Invoking Microagent Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Available Microagents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228activateTraceRole() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232deactivateTraceRole() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233getAdapterServiceInformation(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234getComponents() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235getConfig() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236getConfigProperties(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237getHostInformation() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238getRvConfig(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239getStatus() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240getTraceSinks() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241getVersion() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242_onUnsolictedMsg() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243preRegisterListener() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244reviewLedger() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245setTraceSinks() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246stopApplicationInstance(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247unRegisterListener() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248getEventQueueSize(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249setDebugLevel(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250toggleVerboseFlag(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251showConfiguration(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252terminateADBagent() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253getPollingInterval() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254setPollingInterval() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255getPollingBatchSize(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256setPollingBatchSize() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257getActivityStatistics() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258getActivityStatisticsByService . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259getActivityStatisticsByOperation() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260getThreadStatistics() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261getQueueStatistics() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262getConnectionStatistics() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263resetActivityStatistics(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264resetThreadStatistics(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265


Contents | ix

resetQueueStatistics() . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266resetConnectionStatistics(). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

Appendix A Using TIBCO ActiveMatrix Adapter Service Engine for Database . . . . . . . . . . . 269

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

Installation Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272Installing in GUI Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272Installing in Console Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272Installing in Silent Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273Uninstallation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273

Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275Using Substitution Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280

Appendix B Frequently Asked Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281

General Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

Request-Response Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288

Appendix C Trace Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290Trace Message Fields. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

Status Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295

Appendix D Adapter Properties File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322Properties File Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322Run-time Adapter Properties File Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323Obfuscating or Encrypting a Password in the Properties File. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337Using Oracle Hints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341


x | Contents


Figures | xi

Figures

Figure 1 Typical Publication-Subscription Services Flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Figure 2 Typical Request-Response Service Flow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Figure 3 The TIBCO ActiveMatrix Adapter for Database Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Figure 4 Diagram of Publish-Subscribe Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Figure 5 Alerter Operation (Oracle and MSSQLServer 2005) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

Figure 6 Request-Response Load Balancing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

Figure 7 Example of Unicode Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210


xii | Figures


Tables | xiii

Tables

Table 1 General Typographical Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

Table 2 Syntax Typographical Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviii

Table 3 Values to Use for TIBCO Rendezvous and JMS Transport Types . . . . . . . . . . . . . . . . . . . . . . . . 24

Table 4 Predefined Global Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Table 5 JDBC Drivers and URLs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Table 6 REQUEST Schema Description. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Table 7 REPLY Schema Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

Table 8 Publishing Table Additional Columns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

Table 9 Mapping to Database Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

Table 10 Wire Format Parameters and Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188

Table 11 adbPreCommit Parameter Descriptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

Table 12 Compilers for Platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

Table 13 IANAAppCodePage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215

Table 14 TIBCO ActiveMatrix Adapter for Database and Oracle NLS_Lang Values . . . . . . . . . . . . . . . . . 219

Table 15 Standard Microagent Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

Table 16 Custom Microagent Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

Table 17 Custom Microagent Methods with adb.perfMon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

Table 18 Trace Message Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

Table 19 Required Run-Time Adapter Properties File Parameters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

Table 20 Additional Run-Time Adapter Properties File Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325


xiv | Tables


| xv

Preface

TIBCO ActiveMatrix Adapter for Database software is a bidirectional gateway between databases and applications configured for the TIBCO environment.

Topics

• Related Documentation, page xvi

• Typographical Conventions, page xvii

• Connecting with TIBCO Resources, page xx


xvi | Related Documentation

Related Documentation

This section lists documentation resources you may find useful.

TIBCO ActiveMatrix Adapter for Database DocumentationThe following documents form the TIBCO ActiveMatrix Adapter for Database documentation set:

• TIBCO Adapter Concepts Read this manual to gain an understanding of the product that you can apply to the various tasks you may undertake.

• TIBCO ActiveMatrix Adapter for Database Installation Read this manual for instructions on site preparation and installation.

• TIBCO ActiveMatrix Adapter for Database User’s Guide Read this manual for instructions on how to create, configure, and deploy adapter projects.

• TIBCO ActiveMatrix Adapter for Database Examples Read this manual to work through the examples provided with the adapter.

• TIBCO ActiveMatrix Adapter for Database Release Notes Read the release notes for a list of new and changed features. This document also contains lists of known issues and closed issues for this release.

Other TIBCO Product DocumentationYou may find it useful to read the documentation for the following TIBCO products:

• TIBCO Designer™ software:

• TIBCO Administrator™ software:

• TIBCO BusinessWorks™ software:

• TIBCO IntegrationManager™ software:

• TIBCO Rendezvous™ software:

• TIBCO Enterprise Message Service™ software:

• TIBCO Hawk® software:

• TIBCO Adapter SDK

• TIBCO Runtime™ Agent

• TIBCO ActiveMatrix™ Foundation


Preface | xvii

Typographical Conventions

The following typographical conventions are used in this manual.

Table 1 General Typographical Conventions

Convention Use

TIBCO_HOME

ENV_HOME

ADADB_HOME

Many TIBCO products must be installed within the same home directory. This directory is referenced in documentation as TIBCO_HOME. The default value of TIBCO_HOME depends on the operating system. For example, on Windows systems, the default value is C:\tibco.

Other TIBCO products are installed into an installation environment. Incompatible products and multiple instances of the same product are installed into different installation environments. An environment home directory is referenced in documentation as ENV_HOME. The default value of ENV_HOME depends on the operating system. For example, on Windows systems the default value is C:\tibco.

TIBCO ActiveMatrix Adapter for Database installs into a directory within TIBCO_HOME. This directory is referenced in documentation as ADADB_HOME. The default value of ADADB_HOME depends on the operating system. For example on Windows systems, the default value is C:\tibco\ADADB_HOME\5.5.1.

code font Code font identifies commands, code examples, filenames, pathnames, and output displayed in a command window. For example:

Use MyCommand to start the foo process.

bold code

font Bold code font is used in the following ways:

• In procedures, to indicate what a user types. For example: Type admin.

• In large code samples, to indicate the parts of the sample that are of particular interest.

• In command syntax, to indicate the default parameter for a command. For example, if no parameter is specified, MyCommand is enabled: MyCommand [enable | disable]


xviii | Typographical Conventions

italic font Italic font is used in the following ways:

• To indicate a document title. For example: See TIBCO ActiveMatrix BusinessWorks Concepts.

• To introduce new terms For example: A portal page may contain several portlets. Portlets are mini-applications that run in a portal.

• To indicate a variable in a command or code syntax that you must replace. For example: MyCommand PathName

Key combinations

Key name separated by a plus sign indicate keys pressed simultaneously. For example: Ctrl+C.

Key names separated by a comma and space indicate keys pressed one after the other. For example: Esc, Ctrl+Q.

The note icon indicates information that is of special interest or importance, for example, an additional action required only in certain circumstances.

The tip icon indicates an idea that could be useful, for example, a way to apply the information provided in the current section to achieve a specific result.

The warning icon indicates the potential for a damaging situation, for example, data loss or corruption if certain steps are taken or not taken.

Table 2 Syntax Typographical Conventions

Convention Use

[ ] An optional item in a command or code syntax.

For example:

MyCommand [optional_parameter] required_parameter

| A logical OR that separates multiple items of which only one may be chosen.

For example, you can select only one of the following parameters:

MyCommand para1 | param2 | param3

Table 1 General Typographical Conventions (Cont’d)

Convention Use


Preface | xix

{ } A logical group of items in a command. Other syntax notations may appear within each logical group.

For example, the following command requires two parameters, which can be either the pair param1 and param2, or the pair param3 and param4.

MyCommand {param1 param2} | {param3 param4}

In the next example, the command requires two parameters. The first parameter can be either param1 or param2 and the second can be either param3 or param4:

MyCommand {param1 | param2} {param3 | param4}

In the next example, the command can accept either two or three parameters. The first parameter must be param1. You can optionally include param2 as the second parameter. And the last parameter is either param3 or param4.

MyCommand param1 [param2] {param3 | param4}

Table 2 Syntax Typographical Conventions

Convention Use


xx | Connecting with TIBCO Resources

Connecting with TIBCO Resources

How to Join TIBCOmmunityTIBCOmmunity is an online destination for TIBCO customers, partners, and resident experts, a place to share and access the collective experience of the TIBCO community. TIBCOmmunity offers forums, blogs, and access to a variety of resources. To register, go to http://www.tibcommunity.com.

How to Access All TIBCO DocumentationAfter you join TIBCOmmunity, you can access the documentation for all supported product versions here:

http://docs.tibco.com/TibcoDoc

How to Contact TIBCO SupportFor comments or problems with this manual or the software it addresses, please contact TIBCO Support as follows.

• For an overview of TIBCO Support, and information about getting started with TIBCO Support, visit this site:

http://www.tibco.com/services/support

• If you already have a valid maintenance or support contract, visit this site:

https://support.tibco.com

Entry to this site requires a user name and password. If you do not have a user name, you can request one.


http://www.tibco.com/services/support

https://support.tibco.com

http://docs.tibco.com/TibcoDoc

http://www.tibcommunity.com

| 1

271

Chapter 1 Introduction

This chapter introduces the TIBCO ActiveMatrix Adapter for Database product by providing background information about its features, components and the applications that use it.

Topics

• Overview, page 2

• Adapter Services, page 4

• Message Transports, page 6

• Adapter Architecture, page 8

• Supported SQL Operations, page 14

• Publication Service Considerations, page 15

• Checking for Duplicate Configurations, page 18

• Message Logging Feature, page 19

• Timely Processing of Data, page 20


2 | Chapter 1 Introduction

Overview

TIBCO ActiveMatrix Adapter for Database software (the adapter) allows data changes in a database to be sent as they occur to other databases and applications. It extends publish-subscribe and request-response technology to databases, making multiple levels of delivery services available to applications that need access to these databases. ODBC and JDBC compliant databases such as Oracle, Sybase, and Microsoft SQL Server are supported. While the adapter does not run on z/OS and iSeries systems, it can remotely connect to a DB2 database running on these systems.

TIBCO ActiveMatrix Adapter for Database is written using the TIBCO Adapter SDK software, which allows the adapter to interoperate with other TIBCO products. The adapter can communicate with any application that is configured for the TIBCO environment.

CapabilitiesThe following capabilities are described in detail in this manual.

• Automatically publish data when rows in pre-specified database tables are inserted, updated or deleted:

— Publish data by creating a copy (by value). See Publish by Value on page 15.

— Publish by reference (publish data directly from the source database table without first copying the data from the source table to a publishing table). See Publish by Reference on page 15.

— Publish on parameterized subjects, which allows a subject to be created from the contents of one or more table columns. See Parameterized Subject Names on page 172.

— Preregister certified subscribers with a certified publisher. See Preregistering a Certified Subscriber on page 173.

— Update both parent and child tables within a publication. Both the parent row and all related child rows will be published if the user has set up to publish child data. See Adding Child Tables on page 76.

— Specify that a group of rows fetched from the publishing table is sent in a single message. See Message Grouping on page 16.


Overview | 3

• Automatically subscribe to data and insert, update or delete the data in pre-specified tables in a database:

— Subscribe using wildcard subject names. See Subject Names on page 171 and Preregistering a Certified Subscriber on page 173.

— Perform batch commits based on message count or time out value. See Adapter Services Tab on page 53.

• Use request-response semantics to publish SQL statements, stored procedures, or both on a specified subject:

— Execute stored procedures with IN parameters, OUT parameters, or both. (Binary OUT parameters are not supported. This is a limitation of the Oracle ODBC driver.) See Oracle REF Data Type Support on page 151.

• Support RPC (Remote Procedure Calls). The adapter can act as an operation server providing a simple means for a client to execute a single or batch of SQL statements. See Request-Response Mode on page 126 and RPC Mode on page 134.

• Change published messages for a publication or subscription service by customizing the supplied callout library. See User Callout Library on page 187.

• Customize adapter behavior to suit your needs:

— Specify relationships between tables, then publish related tables together. See Tables Tab on page 74.

— Monitor for database changes using periodic polling or notification by an alerting mechanism. See Adapter Services Tab on page 53.

• Rely on standards:

— Connect to multiple database vendors using ODBC or JDBC drivers. See Design-time Connection Tab on page 47 and Run-time Connection Tab on page 51.

— Interoperate with other TIBCO products through the use of TIBCO Adapter SDK software. See Sending Messages to an Adapter Instance on page 177.

— Monitor your adapter service using TIBCO Hawk software. See Monitoring Tab on page 65 and Chapter 10, Monitoring the Adapter, on page 221.

— Choose message transports: TIBCO Rendezvous or Java Messaging Service (JMS). See Message Transports on page 6.

• Configure the adapter easily using TIBCO Designer. See Chapter 3 on page 33.



Adapter Services

The adapter offers publication, subscription, and request-response services. Communication parameters, database connectivity parameters, and many other parameters can be configured for these adapter services using TIBCO Designer.

The following sections describe how these adapter services interact with the database environment to provide message connectivity with it.

Publication and Subscription ServicesThe following diagram illustrates publication and subscription. Database Application 1 updates a table in a database monitored by a TIBCO ActiveMatrix Adapter for Database publication service. The adapter extracts data from the changed rows of the database tables and publishes it on the specified transport (TIBCO Rendezvous or TIBCO Enterprise Message Service). This data is then available to other applications listening on these transports.

One particular type of listening application could be another instance of a TIBCO ActiveMatrix Adapter for Database adapter, running a subscription service. Upon receiving a message, it updates the relevant tables in its associated database.

Figure 1 Typical Publication-Subscription Services Flow

DatabaseApplication 1

DatabaseApplication 2

Adapter forActiveDatabase

Publisher

Adapter forActiveDatabase

Subscriber

Other TIBCOApplications

TIBCO Messaging


Adapter Services | 5

Request-Response ServiceThe request-response feature in TIBCO ActiveMatrix Adapter for Database allows an application to submit one or more SQL statements, stored procedures, or both for the adapter to execute. After execution, the adapter returns one or more result sets and a result code. The request and responses are formulated using nested self-describing messages. This data flow is illustrated in the figure below.

Figure 2 Typical Request-Response Service Flow

The adapter can also be configured to manage requests using load balancing among a set of adapters that participate in the same named queue. Additionally, the adapter has been designed to allow multiple threads to process application requests.

For more information, see Chapter 7, Using the Request-Response Service, on page 123.

TIBCO Messaging

TIBCOApplication 2

TIBCOApplication 1

Adapter

Database



Message Transports

Two message transport products are available: TIBCO Rendezvous and TIBCO Enterprise Message Service. These transports are described below.

TIBCO Rendezvous TransportThe TIBCO Rendezvous transport provides two qualities of services: Reliable and Certified. In addition, a subscription service can be configured to use distributed queues. These services are described in more detail in the TIBCO Rendezvous Concepts manual.

Supported Wire Formats

Three wire formats are supported, as described below. For detailed information, see the TIBCO Adapter SDK Concepts manual.

• ActiveEnterprise Message This is the standard TIBCO ActiveEnterprise wire format, which provides class information and packing rules for the TIBCO Adapter SDK set of data types. This wire format allows TIBCO ActiveEnterprise components to perform extra validation on messages sent or received.

• Rendezvous Message This is the basic TIBCO Rendezvous wire format.

• XML Message This conforms to specifically constructed and fully compliant XML Schema (XSD) based on the existing definition of the TIBCO ActiveEnterprise schema.

The notion of parent-child relationships is supported with the TIBCO ActiveEnterprise Message wire format. A group of related rows from multiple tables can be encoded into a single message and published. As a subscriber, the adapter can decode a complex message and insert multiple rows into a set of target tables.

An adapter does not convert messages from the TIBCO ActiveEnterprise Message wire format to the TIBCO Rendezvous Message wire format or the reverse. The wire format specified for a publishing table (in the source database) must match the wire format specified for a destination table (in the destination database). A wire format is assigned to a publishing table and to a destination table when using TIBCO Designer to configure an adapter. If message conversion between the two formats is required, use TIBCO ActiveMatrix BusinessWorks software to perform the conversion.


Message Transports | 7

JMS TransportAn adapter service that uses the Java Messaging Service (JMS) can communicate only with the TIBCO Enterprise Message Service server. Before configuring an adapter service for the this transport, you should be familiar with JMS concepts, which are explained fully in the TIBCO Enterprise Message Service User? Guide. The book is available from the machine on which the software is installed in the <install-path>\tibco\ems\doc directory.

The TIBCO Enterprise Message Service transport uses the XML message wire format only. The XML wire format conforms to specifically constructed and fully compliant XML Schema (XSD) based on the existing definition of the TIBCO ActiveEnterprise schema.



Adapter Architecture

TIBCO ActiveMatrix Adapter for Database software allows data stored in a database to be exchanged with applications configured for the TIBCO environment. For Oracle databases, the adapter includes an alerter component that can be used when the default method of polling for database changes is not appropriate. An alerter is also available for Microsoft SQL Server and Sybase and is run as a separate executable.

TIBCO Designer software is used to configure an adapter service. The adapter can be monitored using TIBCO monitoring tools.

The following diagram shows the major components in a system that uses the adapter and other parts of the TIBCO environment. Configuration data is stored in projects that can be used by one or more adapters. The components are explained in this section and the following section.

Figure 3 The TIBCO ActiveMatrix Adapter for Database Environment

Run-time AdapterA run-time adapter service acts as a bi-directional bridge between your database and the TIBCO environment. A publication service monitors your database, extracting data from relevant database tables and sending it on an appropriate subject using the configured quality of service, reliable or certified. A subscription service listens for messages sent on configured subjects and inserts, deletes or updates rows in its associated database tables with the information.

JDBC

TIBCO Adapter forActiveDatabase

ODBC

Database

Application 1 Application 2

TIBCO Messaging

TIBCODesigner


Adapter Architecture | 9

An adapter can be configured with both a publication and subscription service, sending messages and writing them to more than one table. An adapter can publish and subscribe to the same table, provided the publish destination is different from the subscribe destination.

An adapter can also be configured to use request-response semantics, which allows the adapter to act on behalf of a TIBCO application, processing its requests and returning responses. A set of adapters can be configured to participate in a queue (TIBCO Rendezvous Distributed Queue or JMS Queue) that allows load balancing, so requests are handled promptly.

Adapter services using TIBCO Rendezvous transport can use the TIBCO Rendezvous certified messaging service. Message parameters are stored in the project and can be modified for your environment. For an example, see Changing the Location of the Ledger File on page 175. Adapters using JMS can use the durable messaging service.

TIBCO Runtime AgentTIBCO Runtime Agent is required on any machine on which an adapter is installed. It provides basic connectivity between the other components and manages the components on each machine. TIBCO Runtime Agent provides connectivity to other applications configured for the TIBCO environment. TIBCO Runtime Agent provides the run-time environment for TIBCO ActiveEnterprise components. TIBCO Runtime Agent carries out two primary functions by providing:

• The run-time environment; that is, all the shared libraries including third-party libraries

• An agent that runs in the background and starts and stops processes according to deployment information.

Polling or AlerterAn adapter publication service can use periodic polling or an alerter process to monitor changes to a database table.

• The default method is polling. The adapter polls (checks) the publishing tables periodically for any new rows to be published. This method is most efficient when the publishing tables change frequently and a limited number of database operations is preferred.

Each adapter should have a unique name within a project. For more information, see Configuration Tab on page 44.



• In the case where the publishing tables change infrequently, polling may result in many unnecessary database accesses. For this situation the alerter can be used to asynchronously alert the adapter of changes in the database. This removes the need for the adapter to poll its publishing table for existence of new rows. It is recommended to use the alerter only when database changes are infrequent. When the adapter receives an alert, it publishes the new data. (Publishing on subjects is introduced in Subject Names on page 171 and explained fully in TIBCO Rendezvous Concepts.)

The alerter is not available on all databases and platforms. See Chapter 6 on page 109 for details.

Referencing External SchemasWhen configuring an adapter, you can specify the database tables to use for each publication and subscription service. The default schema for specified source and destination table names is associated with the database user account in the create_user.sql script. This script is normally executed as part of the post installation procedure.

You may need to access source and destination tables that are in another schema, and not in your default schema (the schema specified in create_user.sql). Source and destination tables in the adapter can be configured to reference tables in an external schema.

Exception TableYou can designate an exception table for a subscription service. If a subscriber adapter cannot write data to its destination table, it will write the data to the exception table.

To write to an exception table, an adapter must be started with the adb.useExceptTable option set to on in the supplied properties file and an exception table must have been specified when using TIBCO Designer to add a subscription. If an error occurs when inserting data into the destination table, the adapter attempts to insert the data into the exception table.

• If the insertion into the exception table is successful and the RVCM quality of service is specified for the subscription, the transaction commits and confirmation for the message is sent back. In the publishing table, the value of the delivery status column is C, since the message was processed. If you are using a durable JMS subscriber, a JMS confirmation will be sent back.

• If the insertion into the exception table fails, an error message displays, a rollback occurs and a TIBCO Hawk alert is published. The adapter will then terminate. The user is advised to fix the problem with the message or the table



and then to restart the adapter. The message will be delivered again after the adapter starts. (If the adapter did not terminate, after the next message was inserted successfully, its confirmation would also implicitly confirm the failed message.)

You can identify the operation that failed by checking the ADB_OPCODE column of the publishing table. The ADB_ERROR_TEXT column contains error information. For details, see Exception Table on page 161.

Loop DetectionIf a source table is used both as the source and destination table on the same subject, the loop detection feature should be enabled. The feature prevents the same changes from being published repeatedly.

If the loop detection feature is enabled, an additional column, ADB_SOURCE, is added to the source table. When an adapter receives a message, it inserts or updates the source table and enters a T in the ADB_SOURCE column to denote that this row was inserted or updated as a result of a message, rather than from user intervention.

Triggers created by the Adapter Configuration palette are defined to not copy rows with T in the ADB_SOURCE column into the publishing table, which effectively means that the row will not be published. If you must update a row that was received (that is, has a T in the ADB_SOURCE column) and want the updated row propagated, you must also change the ADB_SOURCE column to NULL. This causes the trigger to pick up the row and send it out.

Master-Master ReplicationYou can use the loop detection feature to implement a simple master-master replication scheme. Master-master replication allows multiple sites, acting as peers, to copy and maintain groups of replicated objects.

When loop detection is enabled, an adapter can be configured as both a publisher and a subscriber to the same table on the same subject. When the subscriber receives a message, it compares the adapter ID of the message to its own adapter ID. If adapter ID values match and source and destination tables are the same table in the database, the subscriber discards the message. Other subscriber adapters listening on the subject receive the message only once.

The adapter does not resolve any replication conflicts, such as when two applications each update the same row of their corresponding source table and both publish the change. For advanced replication scenarios, use the replication tools provided by your database vendor.



Multi-file Format ProjectsThis section introduces the multi-file format in which projects are stored during design-time. A detailed description of the multi-file format is in the TIBCO Designer User? Guide.

The multi-file format creates one ActiveEnterprise XML file for each logical object (such as an adapter instance, a set of related ActiveEnterprise classes, or a TIBCO ActiveMatrix BusinessWorks process flow) that occurs in the repository instance. This kind of project is referred to as a multi-file project.

Multi-file projects can be checked into a version control system, and a project can contain more than one adapter configuration. This allows a number of people to work on the same project at the same time, with different people working on each adapter configuration: a developer can check out the specific file corresponding to an object that needs to be changed, updated the file, and check it back in. TIBCO Designer accesses the local synchronized copies of the files on the developer’s hard drive.

The multi-file format provides a hierarchical structure with directories taking the place of RepoDirectory nodes. In previous releases of TIBCO ActiveMatrix Adapter for Database, repository nodes contained references to other repository nodes (such as sessions and endpoints of that instance) using a format such as:

/tibco/public/endpoint/adapter/adb/test/instance1/publisher1

In the multi-file format, these references are replaced by URIs of the following form:

ProjectRoot/multi-file_project_name/#type.name

For example, the first reference above would map to the following URI in a multi-file project:

ProjectRoot/Adapters/adb/test/instance1/publisher1

DAT File FormatFor production and for testing with run-time adapters, developers export multi-file projects to .dat format using TIBCO Designer menu commands. A project can be exported as a local project for local testing or exported to be managed by a repository server.



Fault ToleranceWithin the context of the adapter, a primary instance is the adapter instance that processes messages between the TIBCO environment and the database. The secondary instance uses the same TIBCO Designer project but runs in a stand-by mode and takes over when the primary instance goes down. The secondary instance(s) need not run on the machine or platform as the primary instance.



Supported SQL Operations

The adapter supports the following SQL operations for publishing and subscribing:

• INSERT

• UPDATE

• DELETE

• UPDATE/INSERT (update if row exists, otherwise insert)

Whenever a supported SQL statement is executed against a table monitored by a publisher adapter, the adapter instance sends a message to its subscribers, which update their destination tables.


Publication Service Considerations | 15

Publication Service Considerations

When configuring and preparing to run an adapter instance that involves a publication service, keep in mind the information in this section.

Publish by ValueWhen a user modifies a row in a source table that has been set up for publication, the data in the modified row is copied into a publishing table by a stored procedure created by the adapter during configuration. The adapter then selects the data from the publishing table, packs it into a message, and sends it.

An advantage of publish by value is its speed. However, it does not support some data types, such as Oracle LONG and LONG RAW. Publishing these data types by value from a top-level table is unsupported due to database trigger limitations.

Publish by ReferenceThe publish by reference feature allows you to publish data directly from the source table without first copying the data to a publishing table. If a source table is set up to publish by reference, the publishing table does not contain a copy of the source table data. Instead, the adapter selects the data for publishing directly from the source table itself.

For example, to publish from ORDER_TABLE with a key field ORDER_ID, the publishing table is created with the necessary adapter fields and the ORDER_ID field. When a row in ORDER_TABLE is modified, the trigger fires, populating adapter fields and copying the ORDER_ID value to the publishing table. When the adapter polls the publishing table, it detects the new row and selects from the ORDER_TABLE using the ORDER_ID value found in the publishing table. Then the message is published.

The advantage of publishing by reference is that the data to be published is stored just once. Also, this method supports data types such as Oracle LONG and LONG RAW.

The disadvantage of publish by reference is that after a row in the source table is modified, a small amount of time passes before the adapter selects the row. During this short interval, the row could be modified again. If this happens, the adapter will get only the second set of modifications when it selects the row. However, this time lapse can be minimized by using the alerter.



When publishing by reference, a key column or substitute key column is required for populating the publishing table. If the source table has no primary key, it can only be set up to publish by reference if a key column is specified when adding the publication. For more information, see Publisher Options Tab on page 81.

Update Feature for Parent-Child PublicationIf an adapter is configured to publish child data, the parent row and all related child rows are published when a parent row is updated. Upon receipt of such a message, a subscriber adapter updates the parent row and then updates all the child rows with the data that was received in the message. The adapter updates the child rows by deleting all the related child rows, then inserting child rows again based on the data in the received message.

Changing a Publication Subject Without Using TIBCO DesignerThe subject on which a publisher adapter sends messages for a publication is defined in TIBCO Designer when the publication is added (see Advanced Tab on page 86). You can change the publication subject while the adapter is running (without using TIBCO Designer) by changing the subject value in the database trigger.

By default, the subject is passed as NULL in the trigger. If the NULL value is replaced with a valid subject, it will be used when publishing. See Publication Example on page 156 for more information about triggers.

Message Grouping The default behavior when publishing rows from a database is to place each row fetched from the publishing table into its own message. This behavior is often desired because each row is a different entity and should be operated on the subscriber side individually. However, sending one message for each row can have a performance impact. Also, there are situations where the adapter should send multiple rows in a single message because those rows may represent a unit of work together.

Any changes to the child tables without a change in the parent table will not be processed. The adapter monitors only the parent table for publishing. You can define a trigger on the child table that will update the parent row whenever a change to the child table occurs.


Publication Service Considerations | 17

Message grouping allows you to specify a group of rows fetched from the publishing table into a single message. For the subscription service, when the adapter receives a message that contains multiple rows, the service considers all rows in that message to be one transaction, or treats each row individually, allowing the default batch and bulk options to determine how data is inserted.

Message grouping should not be confused with the parent-child publish and subscribe feature already present in the adapter. Parent-child publication sends one parent and all child rows in a single message. For that to work, a relationship is assumed between the parent row and each child row. In contrast, message grouping deals with multiple parent rows placed into a single message, all of which can be unrelated.

One benefit of using group messaging is to improve the adapter’s performance. For the publication service, publishing a message that contains multiple rows reduces network overhead and database operations for updating the publishing table delivery status.

Message grouping is set at the publication and subscription service level. You can specify different message grouping options for each publication service. For the subscription service, the palette automatically generates the group schemas. You can change the service class reference to the group schema when necessary, for example, when a TIBCO ActiveMatrix BusinessWorks user maps to the group schema.

A subscription service with group messaging enabled can receive a message with single row schema from another publication service. A subscription service with group messaging disabled can still receive group messages if the publication service schema is in the repository used by the subscription service.

If a pre version 5.2 subscription service receives a group message, a deserialization error is thrown and the message is not processed.



Checking for Duplicate Configurations

While multiple adapter instances can run concurrently on the same machine to split the load of subscriptions on the database, each configuration must have a unique name. During the configuration initialization cycle, a new adapter instance broadcasts a message with its name. If any running adapter has this name, it will reply to the broadcast message, causing the new adapter to abort its initialization phase and stop.

If there is running a configuration of the adapter with the same name, that configuration will respond to the broadcast message and the new configuration will not start.

To prevent same-named configurations starting on sites that have multiple TIBCO Rendezvous networks, you must allow the _ADB.DUPDETECT.adapter_instance_name subject to pass through each network by configuring the rvrd daemon for each subnet. See TIBCO Rendezvous Administration for more information about using multiple TIBCO Rendezvous networks.

You can choose to disable the detection for duplication configurations by setting the value of the adb.noDupDetection parameter in the .tra configuration file.


Message Logging Feature | 19

Message Logging Feature

You can configure an adapter instance to use the following transports and subscribe to messages in the wire formats shown below.

• TIBCO Rendezvous transport uses three wire formats:

— TIBCO ActiveEnterprise wire format

— TIBCO Rendezvous wire format

— XML wire format

• The JMS transport uses the XML wire format to communicate with the TIBCO Enterprise Message Service server.

• Opaque transport uses the TIBCO Rendezvous wire format.

Opaque Format

When the adapter receives a message in opaque format, it does not parse the message. Instead, the adapter logs the message by inserting it into a table created by TIBCO Designer. This table has two columns: a RAW column named LOG for storing the message, and a DATE column named TIMESTAMP which contains the date and time the message was inserted.

When using the TIBCO Rendezvous transport, messages can be retrieved by a database program that casts the message back into its original (uninterpreted) format. For an example of a program that selects the uninterpreted message and converts it from binary format to a TIBCO Rendezvous format, see the logtest.cpp file in the <install-path>\tibco\adapter\adadb\5.5\demo\logtest directory.



Timely Processing of Data

If there is a large volume of messages to publish, set the adb.PollingInterval parameter in the adapter properties file to help process TIBCO Rendezvous events in an efficient manner. When the parameter is used, it limits the number of messages that are picked up per poll period. The adapter returns to the event loop when it is finished sending those messages.

If the parameter is not used, the adapter picks up all available messages and then tries to send all of them before returning to the event loop. This could affect adapter performance in a high volume environment.


| 21

Chapter 2 Getting Started

This chapter helps you get familiar with the adapter by running you through a basic exercise. Work through the exercise to get a hands-on understanding on how the adapter works.

Topics


• Create the Database Tables, page 25

• Configure the Properties File, page 26

• Start the Adapter Services, page 28

• Change the Table Values and Receive Notification, page 29

• Exit the Query Tool and Adapters, page 31

• Clean Up, page 32


22 | Chapter 2 Getting Started

Overview

In this exercise, you create tables in your database and configure a publisher adapter and a subscriber adapter. Then you modify a table and observe how the publisher adapter and subscriber adapter handle the change and update the subscription table.

Required Platform and SoftwareThis exercise can be performed on any supported operating system, using adapters provided with TIBCO ActiveMatrix Adapter for Database. The exercise can be run using the TIBCO Rendezvous transport method or the JMS transport method.

The exercise can be performed with an Oracle, Microsoft SQL or Sybase database.

A sample DAT file that can be used with an Oracle database, ADBDemo2-ora.dat, is included in the <install-path>\tibco\adapter\adadb\5.5\demo\demo2 directory.

Before StartingBefore performing this exercise, make sure that TIBCO ActiveMatrix Adapter for Database is installed according to the procedures in the TIBCO ActiveMatrix Adapter for Database Installation.

You should be familiar with using TIBCO Designer to open and close projects and drag and drop resources; see the TIBCO Designer User? Guide for more information. That document also describes the multi-file format used by TIBCO Designer, and converting to and from the .dat file format used by the run-time adapter. The TIBCO Designer User? Guide is available from TIBCO Designer by clicking the Help>Designer Help menu choice.

TablesThe tables are created by running a script specific to your database vendor. The script creates the following tables, all with a common set of columns:

• Source table to which you will add new data.

• Publishing table to which the new data from the source table is copied using the trigger set on the source table. This table has additional columns (prefixed by ADB_) that are used by the adapter.


Overview | 23

• Destination table that TIBCO ActiveMatrix Adapter for Database will update with the new data.

• Exception table to which TIBCO ActiveMatrix Adapter for Database will write any errors that occur during subscription.

ActionsWhen you add data to the source table, the following occurs:

1. The insert action fires a trigger and the inserted row is copied to the publishing table.

2. The publisher adapter polls the publishing table to check if any new rows have been inserted. Newly inserted rows are fetched using ODBC, packed into a message, and published.

3. The subscriber adapter listens for messages. When a message arrives, the subscriber adapter inserts it into the destination table using ODBC.

The following diagram illustrates the activity.

Figure 4 Diagram of Publish-Subscribe Steps

TasksThis exercise consists of the following tasks:

• Create the Database Tables

PublisherAdapter

TIBCO Messaging

InsertTrigger

Publishing

Table

Destination

Table

SourceTable

Exception

Table

Subscriber

Adapter



• Configure the Properties File

• Start the Adapter Services

• Change the Table Values and Receive Notification

• Exit the Query Tool and Adapters

• Clean Up

ExerciseThis exercise uses the Oracle database. You can use the TIBCO Rendezvous or JMS transport. When performing this exercise, use the appropriate transport values for your configuration (as listed in Table 3) and follow the instructions that pertain to your particular database vendor.

You will need the following information, specific to your user environment:

• Database user ID

• Database password

• Database service

The values you will enter for the tables and the publisher and subscriber adapter instances are different, depending on which transport type you are using. (The kind of database you are using does not affect these names.)

Table 3 Values to Use for TIBCO Rendezvous and JMS Transport Types

Item TIBCO Rendezvous Transport Value

JMS Transport Value

Source Table ORDER_TABLE ITEM_TABLE

Publishing Table PUB_ORDER PUB_ITEM

Destination Table SUB_ORDER SUB_ITEM

Exception Table SUB_ORDER_EXCEP SUB_ITEM_EXCEP

Publisher Adapter rvpub jmspub

Subscriber Adapter rvsub jmssub


Create the Database Tables | 25

Create the Database Tables

The first task is to create the database tables.

1. Open a command window and change directory to the demo1 subdirectory. For example:

> cd c:\tibco\adapter\adadb\5.5\demo\demo1

2. Execute the demo1_databaseVendor.sql script in the subdirectory to create the tables for your database. Use your environment-specific user ID, password, and database service. For example:

> sqlplus userid/pswd@dbService @demo1_ora.sql

The script creates the items and displays the status. For example:

C:\TIBCO\adapter\adadb\5.5\demo\demo1>sqlplus karlh/karlh@ORCL @demo1_ora.sql

SQL*Plus: Release 8.1.7.0.0 - Production on Tue Aug 10 13:10:48 2004

(c) Copyright 2000 Oracle Corporation. All rights reserved.

Connected to:Personal Oracle8i Release 8.1.7.0.0 - ProductionWith the Partitioning optionJServer Release 8.1.7.0.0 - Production

Table created.Table created.Table created.Index created.Index created.Index created.Sequence created.Trigger created.Table created.Table created.Table created.Table created.Index created.Index created.Sequence created.Trigger created.Table created.Commit complete.SQL>



Configure the Properties File

After creating the database tables, configure the adapter’s properties file.

1. In the demo1 subdirectory, locate the correct publisher adapter properties file for the transport type you are using. Refer to Table 3 on page 24.

2. Open the publisher adapter file and set the following values for your environment.

For example:

C:\TIBCO\adapter\adadb\5.5\demo\demo1>write rvpub.tra## Sample ADB properties file# Optional properties file you can use in place of command line parameters# Usage: adbagent --propFile propFilename...# Change these settings for your login and dsn for Demo1adb.user karlhadb.password karlhadb.dsn Oracle001

Property Value

adb.user username Use the database account name used by the adapter instance. This is the same username entered during TIBCO ActiveMatrix Adapter for Database post installation procedure when creating a database account for the adapter.

adb.password pwd Use the database account password used by the adapter instance. This is the same password entered during TIBCO ActiveMatrix Adapter for Database post installation procedure when creating a database account for the adapter.

This password is not saved in the project. The global variable %%adb.password%% is saved instead.

adb.dsn data source Use the name of the ODBC system data source for the adapter. The data source is configured as a post installation procedure. See the TIBCO ActiveMatrix Adapter for Database Installation for more information.


Configure the Properties File | 27

.

.

.

3. Save and close the publisher adapter file.

4. In the same subdirectory, locate the correct subscriber adapter file for the transport type you are using.

5. Open the subscriber adapter file and make the same changes that you did in step 2.

6. Save and close the subscriber adapter file.



Start the Adapter Services

You can use TIBCO Rendezvous or TIBCO Enterprise Message Service as the message transport.

1. If you are using TIBCO Rendezvous transport, skip this step and proceed to step 2.

If you are using JMS transport, open a command window and change directory to the TIBCO Enterprise Message Service bin directory and start the server. Note that this step is not required if the server is running as a service on Microsoft Windows. For example:

cd c:\tibco\ems\bin

> tibemsd

2. Open a new command window and change directory to the adapter bin directory. For example:

> cd c:\tibco\adapter\adadb\5.5\bin

3. In separate command windows, start the publisher and subscriber adapters for your database.

If using TIBCO Rendezvous:

> adbagent --propFileC:\TIBCO\adapter\adadb\5.5\demo\demo1\rvpub.tra

> adbagent --propFileC:\TIBCO\adapter\adadb\5.5\demo\demo1\rvsub.tra

If using TIBCO Enterprise Message Service:

> adbagent --propFileC:\TIBCO\adapter\adadb\5.5\demo\demo1\jmspub.tra

> adbagent --propFileC:\TIBCO\adapter\adadb\5.5\demo\demo1\jmssub.tra


Change the Table Values and Receive Notification | 29

Change the Table Values and Receive Notification

You must change table setting in your database to trigger a notification.

1. Open a new command window and change directory to the demo1 subdirectory. For example:

> cd c:\tibco\adapter\adadb\5.5\demo\demo1

2. Invoke the query tool using the environment-specific user ID, password, and database service specified in Task .

> sqlplus userid/pswd@dbService

3. Insert three values into the source table, then commit the change.


SQL> insert into ORDER_TABLE values(111,'Oak Table',499.95);

SQL> commit;


SQL> insert into ITEM_TABLE values(111,'Oak Table',499.95);

SQL> commit;

You will see the message being sent in the publisher adapter window and, after a short delay, received in the subscriber adapter window.

4. Verify that the row in the source table has been inserted into the destination table.


SQL> select * from ORDER_TABLE;

The following example result confirms that the data has been inserted:

ORDER_ID----------ORDER_DESCRIPTION---------------------------------------------------

ORDER_PRICE-----------

111Oak Table

499.95


SQL> select * from ITEM_TABLE;

The following example result confirms that the data has been inserted:

ITEM_ID



----------ITEM_DESCRIPTION---------------------------------------------------

ITEM_PRICE-----------

111Oak Table

499.95

5. Insert additional rows of data, if you wish. If you do, be aware that the first column (containing the value 111 in the example) is a primary key and must contain a value that is unique within the table.


Exit the Query Tool and Adapters | 31

Exit the Query Tool and Adapters

After running the example, exit the query tool and adapters.

1. Exit the SQL query tool. For example:

SQL> exit

2. In a command window, use the tibrvsend application to send a message on the terminate subject to stop each adapter instance:


tibrvsend _ADB.rvpub.TERMINATE nowtibrvsend _ADB.rvsub.TERMINATE now


tibrvsend _ADB.jmspub.TERMINATE nowtibrvsend _ADB.jmssub.TERMINATE now

Note that the tibrvsend program is typically not used to terminate the adapter. Instead, a JMS program typically sends a JMS message to the JMS termination destination configured for the adapter.



Clean Up

This cleanup script removes the example tables that were created by the demo1_cleanup_databasevendor.sql script.

1. In a command window, change directory to the demo1 directory. For example:

> cd c:\tibco\adapter\adadb\5.5\demo\demo1\

2. Execute the appropriate demo_cleanup.sql script in the subdirectory.

> sqlplus userid/pswd@dbService @demo1_cleanup_databasevendor.sql

For example:

> sqlplus karlh/karlh@ORCL @demo1_cleanup_ora.sql


| 33

Chapter 3 Adapter Instance Options

This chapter explains how to create an adapter instance by configuring standard settings. All configuration tasks are performed in TIBCO Designer and the information is stored in a project that is later used by the run-time adapter.

Topics


• Configuration Tab, page 44

• Design-time Connection Tab, page 47

• Run-time Connection Tab, page 51

• Adapter Services Tab, page 53

• General Tab, page 58

• Logging Tab, page 60

• Startup Tab, page 64

• Monitoring Tab, page 65


34 | Chapter 3 Adapter Instance Options

Overview

The adapter instance tabs in TIBCO Designer allow you to create, design, and run an adapter. Adding services to an adapter is described in Chapter 4, Adapter Service Options.

Please read the following sections before starting to configure an adapter.

• Configuration Tasks, page 34

• Saving the Project, page 35

• Testing the Adapter, page 35

• Notes on Configuring an Adapter, page 36

• Configuration Recommendations, page 37

• Changing an Existing Configuration, page 37

Configuration TasksUse the following sequence to create and configure an adapter service.

1. Start TIBCO Designer and open a multi-file project. See the TIBCO Designer User’s Guide for details.

2. Drag the ActiveDatabase Adapter Configuration icon from the palettes panel to the design panel. This creates an adapter instance named, by default, ActiveDatabaseAdapterConfiguration.

3. Define the adapter instance by assigning a new name. See Configuration Tab on page 44.

4. Define and test the design-time connection options for the instance. See Design-time Connection Tab on page 47.

5. Either now or when the instance is ready to deploy, define and test the run-time connection options for the instance. See Run-time Connection Tab on page 51.


Overview | 35

6. As necessary, modify the following default values for the instance:

— Default startup state and session, metadata search URL: see Startup Tab on page 64

— Standard, class, and default MicroAgents: see Monitoring Tab on page 65

— Logging options: see Logging Tab on page 60

— Termination subject, encoding, and debugging options: see General Tab on page 58

— Threads, polling, publishing child data, bulk insert size, TIBCO Rendezvous queue size, exception table: see Adapter Services Tab on page 53

7. Add one or more services to the adapter instance by dragging a service icon from the palettes panel and dropping it in the design panel: see Chapter 4, Adapter Service Options.

8. Under the service’s Configuration tab, set the combination of options for each service. This is required.

9. As necessary, define tables, mappings, message and reply subjects, endpoints, and other service options.

10. Save the project as a server repository project.

After configuring the adapter, create the run-time adapter property file and add the project name and adapter instance name.

Also define and test the run-time connection options for the instance if you did not do it earlier. See Run-time Connection Tab on page 51.

Saving the ProjectConfiguration information for an adapter and all other parameter settings related to the adapter are saved as a project. At any time while configuring the adapter, you can save the associated project. Each time you save a project, any configuration information you have entered is saved as a project.

For detailed steps and more information about exporting or importing projects to different formats (such as .dat), see the TIBCO Designer User? Guide.

Testing the AdapterYou can use the Adapter Tester to verify that an adapter instance is configured correctly. The tester is invoked from the TIBCO Designer Tools menu and is documented in TIBCO Designer Palette Reference.



Notes on Configuring an Adapter

Required and Optional Settings

• You must set the options in the Configuration Tab for all adapters.

• The Design-time Connection Tab options are required before you can start designing the adapter and adding services.

• The remaining tabs are optional, and are customized for your adapter or services only as needed.

You should be aware of the following before starting or configuring an adapter.

• A database account must have been created for running the adapter. For details, see the TIBCO ActiveMatrix Adapter for Database Installation.

• The database server process, and the repository server process if you are using a remote repository, must be running to start an adapter instance.

• Each adapter must have configuration information defined in a project using TIBCO Designer. For instructions, see Chapter 3, Adapter Instance Options.

• Data source connection parameters for an adapter instance can be specified in the adapter properties file and in TIBCO Designer. Values given in the properties file override the same values specified in TIBCO Designer.

• An adapter instance name is defined in TIBCO Designer. Names must be alphanumeric, can contain underscores (_) and hyphens(-), and must be less then 80 characters.

• Multiple subscribers of the same instance should not be configured to listen on the same subject.

• Multiple subscribers in different instances can be configured to listen on the same subject but should not write to the same destination table.

• An adapter instance can be configured as a publisher, subscriber, or both. It can handle request-response operations if it is configured to do so.

• An adapter instance can publish or subscribe to a maximum of 1024 database columns per table. The maximum size of each record is restricted by the number of bytes per row supported by the database server, and the amount of contiguous space in memory available to build the message.

• If an adapter instance cannot access the database at startup, the adapter will not start. If the connection fails after startup, you can configure the adapter to attempt automatic reconnections (see the Run-time Connection Tab on page 51). You can also build a TIBCO Hawk rule to restart the adapter instance after the database is restarted.

• To start or stop an adapter instance see Start or Stop the Adapter on page 106.


Overview | 37

• An adapter instance logs error, warning, debug and information messages to the console window by default. Some of the output can be directed to a log file that can be located anywhere on your file system.

• Do not use ADB_ to prefix columns in the database. The prefix is reserved for this product.

Configuration RecommendationsThe following recommendations and limitations apply when using TIBCO ActiveMatrix Adapter for Database:

• The wire format for both the publishing and destination tables must be the same, otherwise an error will occur.

• Publisher and subscriber information should be defined in the same project for all cases. When using parent-child tables or the XML wire format, define the publisher and subscriber in the same repository.

• The following limitations apply to TIBCO Rendezvous transport type parameterized subject names:

— Do not insert the * or > character as an attribute value (varchar, char) in a column that is part of a parameterized subject.

— Do not insert data of Float type as an attribute value in a column that is part of a parameterized subject.

— Do not use Date type columns as part of a parameterized subject. Dates generally contain characters such as dashes and spaces that do not work well as part of a subject.

Changing an Existing ConfigurationYou can change an existing adapter instance before or after it is deployed, but doing so may have unintended consequences to the database, the schema, and to other adapter instances associated with the same schema. When making changes, be aware of the following:

• It is recommended to back up your project before making major configuration changes. To back up a project, use TIBCO Designer to export the project. The export file can be imported into TIBCO Designer, if necessary. For more information, see TIBCO Designer Palette Reference.

• Legitimate changes to the database as a result of adapter instance changes may not be immediately successful, for example, if the database was not available at the time the change was made.



• If you want adapter instance changes to be reflected in the database, make sure the Deploy On Save checkbox is selected; otherwise, make sure this checkbox is cleared. This checkbox is on the Configuration tab of the configured adapter resource.

• If you delete an adapter instance or service that is already deployed, the related database will be cleaned up as soon as you confirm the deletion, and the changes will be irreversible.

• If you change an adapter instance or service that is already deployed, the related database will be cleaned up as soon as you re-save the project.

• Deleting an adapter instance deletes all the services included in that configuration.

• When deleting an adapter service, you have the opportunity to save the service’s schema. This is important when another adapter instance is also associated with the same schema.

• If an adapter instance is renamed, the schema associated with that configuration is also renamed, and if another configuration is associated with that same schema, the association will become invalid once the schema is renamed.

If legitimate changes to the database result in error messages while you are trying to save the changes, you may need to run the cleanup script that the adapter creates during the save operation. For more information about these error messages and using the script, see Using Database Cleanup Scripts on page 203.

Using Global VariablesThe variable substitution mechanism can override global variables predefined in the project in a restricted manner. Predefined variables can be viewed and set in TIBCO Designer. Variables are specified as %%VARNAME%% and cannot contain any white space.

Variable substitution allows you to accomplish the following.

• Substitute string variables specified in the project at startup time.

• Locally define the value for a variable for a specific project. The local value takes precedence over any global value.

• Specify the value for a variable in a properties file. This overrides the project repository and values set in code, but not variables set on the command line.

• Enforce the pre-defined variables listed in Predefined Global Variables on page 40.


Overview | 39

Variables can be used anywhere in the configuration and will be replaced by the locally-defined adapter instance.

Variable Specification

The adapter can specify variables:

• In the project during configuration using TIBCO Designer.

• In a properties file.

• In TIBCO Administrator Enterprise Edition when deploying the project.

The values in the properties file or Enterprise Edition take precedence over the values set in the project through TIBCO Designer.

Specifying Variables Using TIBCO Designer

Global variables provide an easy way to set defaults for use throughout your project.

For example, you could assign the value 7474 to the predefined global variable RvDaemon. You can then use the variable in different sessions in your adapter. If you wish to change the TIBCO Rendezvous daemon for your adapter, you can globally set it to a different value or override it from the command line.

To Specify Global Variables:

1. In the project panel, select the Global Variables tab.



The project panel displays all currently defined global variables. You have these choices for modifying global variables:

— To assign or change a variable value, triple-click the variable. The variable expands so you can change either the variable name or the variable value. Press Enter when done.

— To add a new global variable group, click the group icon (on the left below the project panel). Specify the name of the group, then press Enter.

— To add a global variable to the list, click the abc icon below the project panel. A new global variable item is added to the bottom of the list. Type the variable name and, optionally, the value. Press Enter when done.

— To add a global variable to a group, select the desired group icon and click the abc icon below the project panel.

— Click the pencil icon to open the advanced editor where you can add more information to a global variable.

The global variable is now displayed in the global variables list.

2. When you want to use the global variable in the fields of a resource, enter the variable name surrounded by %% on both sides.

When the project is deployed and the configured components are run, all occurrences of the global variable name are replaced with the global variable value (unless it was overridden in a way that had higher precedence). For example, RvServiceTest would be replaced with 7800.

A number of global variables are predefined. See Predefined Global Variables in the next section for information. You may add definitions of any variables you need to the predefined variables.

Predefined Global Variables

The following table lists and explains the predefined global variables. Some global variables are automatically used within the system when an adapter instance is configured.

Table 4 Predefined Global Variables

Variable Description

Deployment Defaults to the TIBCO Designer project name. This value can be any string value. This global variable is used by the system to partially define the subject name defined for a service.


Overview | 41

DirLedger Specifies the path name of the TIBCO Rendezvous certified messaging ledger file. The default is the root installation directory.

DirTrace Specifies the path name for log file used by the adapter. The default is the root installation directory.

Domain The default value for file-based local projects is Domain. The value for server-based projects is the domain to which the project was saved.

JmsProviderUrl Tells applications where the JMS daemon is located. Setting this value mostly makes sense in early stages of a project, when only one JMS daemon is used.

JmsSslProviderUrl

Tells applications where the JMS SSL daemon is located.

RemoteRvDaemon TIBCO Rendezvous routing daemon (rvrd) to be used. See TIBCO Administrator Server Configuration Guide for details about setting up a domain using rvrd.

RvDaemon TIBCO Rendezvous daemon. Sessions use this daemon to establish communication. The default value is 7500.

RvNetwork TIBCO Rendezvous network. This variable need only be set on computers with more than one network interface. If specified, the TIBCO Rendezvous daemon uses that network for all outbound messages.

In most cases, you can leave the default.

RvService TIBCO Rendezvous service. The TIBCO Rendezvous daemon divides the network into logical partitions. Each transport communicates on a single service. A transport can communicate only on the same service with other transports.

Unless you are using a non-default TIBCO Rendezvous configuration, you should leave the default (7500).

Table 4 Predefined Global Variables (Cont’d)

Variable Description (Cont’d)



RvaHost Computer on which the TIBCO Rendezvous agent runs. This variable is only relevant if you are using the TIBCO Rendezvous Agent (rva) instead of the TIBCO Rendezvous daemon, and if you have configured a non-default setup. See TIBCO Rendezvous Administration for details about specifying the rva parameters.

RvaPort TCP port where the TIBCO Rendezvous agent (rva) listens for client connection requests. See TIBCO Rendezvous Administration for details about specifying the rva parameters. Defaults to 7501.

TIBHawkDaemon TIBCO Rendezvous daemon used in the TIBCO Hawk session. Specifies which Hawk daemon handles communication for the session. A local daemon is specified by the communications type (always tcp) and a socket number. The default configuration uses the local daemon with the TCP socket number 7474.

Specify a remote daemon by inserting its host name or IP address between the tcp entry and the port number of the daemon parameter, such as tcp:remote_computer:7800.

See the TIBCO Hawk Installation and Configuration manual for details about this parameter.

TIBHawkNetwork TIBCO Rendezvous network used by the TIBCO Hawk session. Specifies which network to use for outbound session communications when a computer is connected to more than one network, and also specifies the multicast groups to use for communication.


TIBHawkService TIBCO Rendezvous service used by the TIBCO Hawk session. The Service parameter specifies which User Datagram Protocol (UDP) service group the TIBCO Rendezvous daemon should use for session communications. The default service port is 7474.


Table 4 Predefined Global Variables (Cont’d)

Variable Description (Cont’d)


Overview | 43

Adapter Instance TabsThe following tabs are available when configuring an adapter instance.

• Configuration Tab

• Design-time Connection Tab

• Run-time Connection Tab

• Adapter Services Tab

• General Tab

• Logging Tab

• Startup Tab

• Monitoring Tab



Configuration Tab

You must define the options on this tab before other options can be configured. Click Apply to apply the changes before leaving this dialog.

Instance Name

Use the default name or replace it with a name of your choice.

• An instance name must use alphanumeric characters. An underscore (_) character can be used. The entire instance name must be less than 80 characters. The space character cannot be used in an instance name.

• An instance name cannot use global variables.

• An instance name must be unique with respect to other adapter instances for the same adapter in the project. The same instance name can be used to name an adapter instance for a different adapter in the same project. For example, a TIBCO ActiveMatrix Adapter for Database adapter instance named TEST and a TIBCO Adapter for Siebel adapter instance named TEST can coexist in the same project.

• Each instance name must be unique per adapter within a project even if each instance is defined in a different folder. That is, configuring same-named adapter instances in different folders will not make their names unique.

When you create an adapter instance, the palette automatically creates several resources for it. The names of these resources derive from the name of the instance they belong to. Changing the adapter instance name results in an automatic regeneration of the resources names. If you manually modify any resource name, that particular name will not be automatically regenerated next time your rename the adapter instance.

Description

(Optional) Provide additional information about the adapter instance.

Version

The version string indicates the TIBCO ActiveEnterprise (AE) configuration format in which the adapter instance is saved. An adapter instance can be saved in AE 4.0, AE 5.0 or AE 5.1 format.

When a new adapter instance is created in TIBCO Designer 5.x, the version string is set to AE Version 5.1. When a 4.x adapter instance is opened in Designer 5.x, the Version field is set to AE Version 4.0.


Configuration Tab | 45

• If a 4.x adapter instance is to be run against a 4.x run-time adapter, the instance must be saved with the Version field set to AE Version 4.0.

• If you are using TIBCO Designer 5.x to modify 4.x adapter instances, change only features supported by the 4.x. run-time adapter and use the validation utility to verify the instance before deploying the project. The validation utility scans the project and returns warnings if any 5.5 features are defined for 4.x adapter instances. Invoke the utility from the Project>Validate Project for Deployment menu command in Designer.

To change versions, click the Change Version button.

Message Filter

Specify a message filter, if you have configured a message filter resource for use with the adapter. The plugin allows you to manipulate incoming and outgoing data before sending it on the network or handing it to the target application. Plugins can be written using the TIBCO Adapter SDK. See the TIBCO Adapter SDK Programmer’s Guide for information about writing a message filter.

Show All Tabs

Select this box to display additional tabs for configuring advanced options.

Vendor

Select the database vendor to which the adapter connects. This populates the JDBC Driver and JDBC URL fields in the Design-time Connection Tab with the appropriate data.

DB2 AS400 Library

(Appears when DB2 AS400 is selected in the Vendor field.) When this option is selected, the adapter automatically creates an asynchronous queue named TIBADB on the AS400 machine. Your programs will be created in TIBCOSRC in the library specified in the DB2 AS400 Library field.

You must also set the trigger option under the publication service DB2/AS400 Options tab.

In the drop-down list, vendor names enclosed in parenthesis are not supported. For example Informix and INGRES.



Write to Database on Save

Select this box if you want to write configuration settings to the database when this project is saved. This is the default mode.


Design-time Connection Tab | 47

Design-time Connection Tab

You must define these JDBC information and database account options before you can design the adapter instance. I

After completing these fields, click Apply, then click Test Connection to establish a connection to the database. When the connection is successful message appears, click OK. You can now begin design-time configuration of the adapter.

Many of the following fields can use global variables. Click the Global Variables tab in the project panel to add or modify a global variable.

JDBC Driver and JDBC URL

The name and URL of the JDBC driver used during design-time configuration. The following table lists all the supported JDBC drivers and their URLs. For detailed parameter descriptions, see your JDBC driver documentation.

Instead of manually entering the following driver and URL values, you can populate these fields using a connection template or previously-stored connection parameters. For more information, see the text following this table.

Table 5 JDBC Drivers and URLs

Database Driver and URL

Oracle JDBC Driver: tibcosoftwareinc.jdbc.oracle.OracleDriver

JDBC URL: jdbc:tibcosoftwareinc:oracle://servername: 1521;SID=ORCL

Microsoft SQL Server

JDBC Driver: tibcosoftwareinc.jdbc.sqlserver.SQLServerDriver

JDBC URL: jdbc:tibcosoftwareinc:sqlserver://servername:1433;databaseName=databaseName

Note: Default port number is 1433.



User Name

Specify the database user that the design-time adapter uses to connect to the database.

Password

Specify the password for the database user.

Remember Password

If this field is checked, the password is obfuscated and saved in the project file. The next time you open this project, the Password field displays the masked password and the adapter will connect to the database without the user needing to know or enter the password.

Sybase JDBC Driver: tibcosoftwareinc.jdbc.sybase.SybaseDriver

JDBC URL: jdbc:tibcosoftwareinc:sybase://servername:5000

Note: Specify a databaseName parameter (as shown in the SQL Server description above) if connecting to a database that is not the default database.

Sybase Adaptive Server Anywhere

JDBC Driver:com.sybase.jdbc2.jdbc.SybDriver

JDBC URL: jdbc:sybase:Tds:localhost:2638/asademo

DB2 OS390 JDBC Driver: tibcosoftwareinc.jdbc.db2.DB2Driver

JDBC URL: jdbc:tibcosoftwareinc:db2://servername:port;locationName=DB2locationName;packageName=packageNamePrefix

DB2 AS400 JDBC Driver: com.ibm.as400.access.AS400JDBCDriver

JDBC URL: jdbc:as400://serverIP;libraries=lib

DB2 UDB JDBC Driver: tibcosoftwareinc.jdbc.db2.DB2Driver

JDBC URL: jdbc:tibcosoftwareinc:db2://servername:50000;databaseName=databaseName;packageName=DEF00

Table 5 JDBC Drivers and URLs (Cont’d)

Database Driver and URL (Cont’d)


Design-time Connection Tab | 49

If this field is not checked, the password must be entered here each time the project is opened. The password is saved in the project file as the global variable %%adb.password%%.

Use Design-time Connection For Run-time

This field is not active.

Using Connection Settings TemplatesTIBCO ActiveMatrix Adapter for Database provides connection settings templates with JDBC driver information for each supported database vendor. The templates populate the JDBC Driver and JDBC URL fields with the default settings as shown above. You then replace the variables with values appropriate for your configuration.

To populate the fields with default values, select ActiveDatabase>Connection templates, then select your database type from the submenu. The following screen sample shows Oracle being selected.

Saving Connection Settings for ReuseAfter you have customized the connection settings on this tab for your configuration, you can save them for use in another adapter. You can save as many sets of customized connection parameters as you need. The parameter sets can include the user name and password.

To save a set of connection parameters:

1. Fill in the connection parameters for the adapter.



2. Select ActiveDatabase>Save Connection Settings. The following dialog displays:

3. Type a name, then click OK. The following dialog displays:

4. Click either Yes or No:

— Yes saves the password in the parameter set. (If the Password field is empty, no password is saved.) When the user selects the parameter set, the Password field will be populated with the current password in clear text, even if the password is not saved in the project file or if it has been obfuscated in the project file.

— No does not save the current password in the parameter set. The user will have to enter it manually.

The connection parameters are saved.

Using Saved Connection ParametersTo use a saved set of connection parameters, select ActiveDatabase>User connections, then select a parameter set from the submenu.

The fields on the Configuration and Connection tabs are populated with the values stored in the parameter set.


Run-time Connection Tab | 51

Run-time Connection Tab

These settings must be configured before you start the run-time adapter.

After completing these fields, click Apply, then click Test Connection to establish a connection to the database. When the connection is successful message appears, click OK. The run-time adapter is now configured.

ODBC DSN

The name of an ODBC system data source for the database where the source or destination database table resides. The adapter uses this data source to send and receive information. If not specified, the data source name must be given in the adapter’s properties file or on the command line.

User Name

This is the database user that the run-time adapter uses to connect to the database. This field is automatically populated when you select a template as described in Using Connection Settings Templates on page 49. It cannot be changed here.

Password

This is the password for the database user. This field is automatically populated when you select a template as described in Using Connection Settings Templates on page 49. It cannot be changed here.

Database Disconnection Codes

Specify the database disconnection codes provided by your database. Multiple codes can be entered if each is separated by a semi-colon (;). No spaces are allowed between codes. To populate this field with the default disconnection codes for your database, click ActiveDatabase>Connection templates><database-name>.

If disconnection codes are not provided and a database operation fails, the adapter issues a test query to verify whether the database connection is valid. If the connection is not valid, the adapter attempts to reconnect and will shutdown if it cannot reconnect. This approach is database independent, but affects performance because the adapter has to send the query and wait for results.



If database disconnection codes are provided in this field, they are stored in a fatal error code dictionary. When a database operation fails, most database systems return an error code to the adapter. The adapter can look up the returned error code in the fatal error code dictionary and determine if the database connection is lost and respond accordingly. This is more efficient then sending the test query.

Maximum Number of Reconnect Attempts

Specify the total number of reconnection attempts to make after the service has been suspended. When this number is reached, the run-time adapter or adapter service will be stopped. A value of -1 means reconnection attempts will continue indefinitely.

Number of Reconnect Attempts Before Suspending Impacted Service(s)

Specify the number of reconnection attempts to make before suspending the service. This value is 1 and cannot be changed.

Interval between Reconnect Attempts (milliseconds)

Specify the time interval (in milliseconds) to elapse between each reconnection attempt.

Adapter Termination Criteria (after max number of reconnect attempts)

The adapter and all of its services is stopped if any one of its services has been unable to re-establish connection after the Maximum Number of Reconnect Attempts has been made. This option cannot be changed.

Test Connection

This option is not enabled.


Adapter Services Tab | 53

Adapter Services Tab

You are not required to change any settings in this tab. These settings affect all publishers, subscribers, and request-response services defined for the adapter, unless overridden by the individual service configurations as described in Chapter 4, Adapter Service Options.

Click Apply to apply the changes before leaving this tab.

All Publication ServicesThese settings apply to all publication services in the adapter.

Number of Publication Service Threads

This field is currently disabled. Threads are allocated on demand. For example, if the adapter instance has only a publication service configured, a publication thread and database connection will be created automatically.

Use Polling Batch Size

Check this box to activate Polling Batch Size (Maximum Rows). If Polling Batch Size is not active, the adapter picks up all new rows and then tries to send all of them before returning to the event loop. This could affect adapter performance in a high volume environment.

Polling Interval

Type a specific polling interval in milliseconds. This is how often an adapter with a publication service checks the publishing table for new rows. The default is 5000, or once every five seconds. Note that if you specify a polling interval of 0, it is assumed that you are using an alerter to manage polling.

Polling Batch Size (Maximum Rows)

(Only active when Use Polling Batch Size is checked) The maximum number of messages that are picked up per polling interval. The adapter returns to the event loop when it is finished sending those messages. Using this option helps process TIBCO Rendezvous events in an efficient manner. For example, when polling a large number of rows, the adapter works best if a fixed number of rows is specified in this field.



Batch Publish Status Updates

(Only active when Use Polling Batch Size is checked. Do not use this option when messages are published using a parameterized subject name.)

Check this box to optimize publishing performance by batching message status updates to the publishing table.

If an adapter stops before a batch update is performed, the status column is not updated. As a result, duplicate messages may be published when the adapter is restarted.

Publisher Batch Confirm Size

(Applies only to adapters with a publication service that uses certified message delivery. Do not use this option when messages are published using a parameterized subject name.)

The number of message status updates to include in a single batch.

Entering a value in this field optimizes performance. However, if an adapter stops before a batch update is performed, the status column is not updated. As a result, messages that were successfully published might still have a status of P (pending) in the publishing table when the adapter is restarted. In this case, the ledger file contains the correct status information. Smaller values in this field decrease this risk.

Publisher Batch Confirm Timeout

(Applies only to publisher adapters with publications that use certified message delivery. Do not use this option when messages are published using a parameterized subject name.)

The number of milliseconds to wait before updating the status column. After this interval, an update is performed even if the batch size value is not reached. The default value is 10000 (10 seconds). A value of 0 means that no timeout interval is used.

If an adapter instance stops before a batch update is performed, the status column is not updated. As a result, messages that were successfully published might still have a status of P (pending) in the publishing table when the adapter is restarted. In this case, the ledger file contains the correct status information. Smaller interval values decrease this risk.



Publish Child Data

This is selected by default. When selected, the parent row and all related child rows is published when a parent row is updated. Upon receipt of such a message, a subscriber adapter updates the parent row and then updates all the child rows with the data that was received in the message.

The adapter updates the child rows by deleting all the related child rows, then inserting child rows again based on the data in the received message. For information on adding related tables for a publisher adapter, see Adding Child Tables on page 76. For information on adding related tables for a subscriber adapter, see Child Table Mappings Tab on page 92.

All Subscription ServicesThese settings apply to all subscription services in the adapter.

Number of Subscription Service Threads

This field is currently disabled. Threads are allocated on demand. For example, if the adapter instance has only a subscription service configured, a subscription thread and database connection will be created automatically.

Subscriber Batch Commit Size

The number of messages to batch before invoking a commit operation. The default is 1.

Subscriber Batch Commit Timeout (milliseconds)

Specify an interval (in milliseconds) that can expire before confirmation messages regarding successful insertion into the exception table are sent back to the publisher. The default is 1.

The batch commit feature does not commit all received messages if the adapter instance terminates before the batch commit value or time-out value is met.

Any changes to the child tables without a change in the parent table will not be processed. The adapter monitors only the parent table for publishing.



Subscriber Bulk Insert Size

All incoming messages to insert are stored until this size is reached. Then, a bulk insert operation is performed on the destination table. This number must be less than or equal to the value in Subscriber Batch Commit Size. The default value is 1.

Note the following considerations:

• If any individual messages greater than 32K are published, bulk insert is automatically turned off.

• If an update statement is published while messages are being batched, the bulk insert is performed regardless of whether the size value has been reached. After records have been inserted, the update operation is performed.

Rendezvous Maximum Queue Size

Maximum number of messages to allow in the TIBCO Rendezvous event queue. The default value is 0, which means no limit is placed on event queue size. Use this option to prevent a subscriber’s memory from overflowing if the publisher is too fast.

Use Exception Table

Select this box to use an exception table. The exception table is defined when you create a subscription service.

All Request-Response ServicesThese settings apply to all request-response services in the adapter, unless the individual service is configured differently.

Number of Request-Response Service Threads

Indicate the number of database request-response threads to use. Valid values are from 1 through n. Each thread has a separate connection to the database. Specifying multiple threads allows you to load balance incoming RPC requests.

Do not use this option if LONG, LONG RAW, image, or variable-length BINARY type records are published. These records cannot be bulk inserted into a destination table.



Request Response max Rows

Specify the maximum number of rows to fetch. This can be used limit the memory usage of the adapter. The unfetched rows will be ignored by the adapter.



General Tab

This tab allows you to set a termination subject or topic and specify the encoding type, debug level, and verbose mode. Note that the adapter should communicate only with other applications that support the same code pages or Unicode.

Click Apply to apply the changes before leaving this dialog.

Termination Subject or Topic

A message sent on this subject (if TIBCO Rendezvous is the transport) or topic (if JMS is the transport) stops the adapter.

The default termination subject or topic for a 5.5 project is:

%%Domain%%.%%Deployment%%.adb.%%InstanceId%%.exit

The termination subject for a 4.0 project is shown below. Do not change it.

_ADB.%%InstanceId%%.TERMINATE

See TIBCO Rendezvous Concepts for information about specifying subject names. See the TIBCO Enterprise Message Service User’s Guide for information about publishing on a topic.

JMS Destination Prefix

Specify this prefix if you are using the JMS transport and have selected the Use Separate Service Thread option for a subscription service. The destination prefix is the subject used by the single-threaded subscription service.

By default, the adapter uses a dynamic destination that is generated using the Domain and Deployment global variables, and the adapter instance name. If you use this default dynamic destination, make sure the values for Domain and Deployment are not empty. You can override the default dynamic destination by specifying the static destination in this field. The static destination must be defined with write permissions on the TIBCO Enterprise Message Service server before it can be used by the run-time adapter.


General Tab | 59

Adapter Encoding

Select the encoding from the drop down menu. The adapter may support other encodings not shown. See Chapter 9, Setting Encoding Options, for a list of additional encodings that can be typed into this field.

Debug Level

This field is valid only when the Debug logging option is selected, as described in Logging Tab on page 60.

Select how much debugging output to provide at the console window or log file location specified in the Log File field. The options are:

• No debug information

• Only SQL commands executed against the database

• Only the ODBC data source for each SQL command

• All debug information

Generate Verbose Output

This field is valid only when the Information logging option is selected, as described in Logging Tab on page 60.

Check this box to include verbose output (all available information) at the console window or log file location specified in the Log File field.

Script File

Allows you to change the location where the SQL script file is written.

The palette does not validate encoding values that you type into the field. The run-time adapter will throw an error if the encoding value you type is not supported.



Logging Tab

Use these settings to configure a log file or log sinks, including which types of trace messages you want logged and where they are sent. Click Apply to apply the changes before leaving this dialog.

By default all error, warning, debug and information messages are printed in the console window in which the adapter was started. Alternatively, you can specify a log file and path to redirect trace output to a log file located anywhere on your file system. The default log file name is %%DirTrace%%/%%Deployment%%.%%InstanceId%%.log, and is saved in the same directory where your project (repository instance) is stored.

Logging trace messages is helpful for troubleshooting. There are four levels of trace messages that you can log: Information, Warning, Debug, and Error. Trace messages are described and listed in Appendix C, Trace Messages.

By default, the Use Advanced Logging option is not selected. In this mode, you configure a standard log file using the fields on this tab, as shown in the example below.

Most errors received by the adapter are logged. The only errors that might not be logged are any TIBCO Rendezvous or TIBCO Adapter SDK errors that appear at startup time before tracing can be initialized.

Logging affects system performance. It is recommended that you use logging only as needed.


Logging Tab | 61

When you select Use Advanced Logging, you configure log sinks using icons in the TIBCO Designer project panel. This gives you complete control on selecting the destinations and associating desired roles with each of the destinations.

Use Advanced Logging

When this box is clear, the standard log file is used. This is the default. Fill out the remaining fields on this tab. You do not need to read the rest of this field description.

When this box is selected, you can set two standard output destinations (sinks) for trace messages and set the tracing level for the roles selected. The following sink types are available:

• File

• STDIO

• Hawk

• Network

To create log sinks:

1. Check the Use Advanced Logging box.

2. Click Apply.



3. In the TIBCO Designer project panel, select the Log Sinks folder under the Advanced folder.

4. Select an existing log sink or create a new one:

— Select the File or STDIO log sink icon.

— Create a new log sink by dragging and dropping the Generic log sink icon from the palette panel into the design panel, then assign a type to it from the drop down menu in the configuration panel. Click Apply.

5. With the desired log sink icon selected in the design panel, fill in the fields in the configuration panel. You can also change the name and enter a description for each sink by right-clicking on the sink icon in the project panel.

— File sink logs the trace messages to a file. Specify the file limit, file count, and the option to append or overwrite. By default, the file limit is 30000 bytes, the file count is 3, and the mode is append.

— STDIO sink sends trace messages to stdout or stderr. By default, stdout is selected.

— Hawk sink sends each trace message to TIBCO Hawk Monitor or TIBCO Hawk Display using the Hawk session, which is created by the adapter for


Logging Tab | 63

monitoring purposes. Specify the MicroAgent Name. (For details on Hawk sessions, see Using Global Variables on page 38.)

— Network sink publishes each trace message on TIBCO Rendezvous. Specify the session and the subject on which the trace messages needs to be published.

Log to Standard I/O

When selected, trace messages are displayed in the command prompt window where the adapter is started. This is the same as creating a STDIO sink. When not selected, trace messages do not display in the window.

Log File

Specify the name of the log file to which trace messages are written. This is the same as creating a file sink. Global variables can be used to specify the location of the log file. See Using Global Variables on page 38 for more information.

Type the name and file system path, or click Browse and select an existing log file. If no file name is specified, trace information is not written to a file.

Log Info/Debug/Warning/Error Messages

Select the types of trace messages you want logged.

Debug messages should not be logged unless requested by the TIBCO Product Support Group. This option writes a lot of information to the log file and significantly reduces the speed of the adapter.



Startup Tab

Changing these settings is not allowed. This tab displays the default startup behavior.

Show Startup Banner

The startup banner displays the run-time adapter version, the infrastructure version on which the adapter is built, and copyright information in the console window when the adapter is started.

Metadata Search URL

This field is predefined and cannot be changed. It specifies the location where the adapter searches for base schemas. All schemas that have been defined and saved at this location are loaded at startup.


Monitoring Tab | 65

Monitoring Tab

These settings do not need to be configured unless TIBCO Hawk is installed.

You can use microagents to supplement the monitoring information provided by the standard logging capability. Examples of supplemental information that you can obtain with microagents include the repository URL and the command line arguments used to start the adapter.

Click Apply to apply the changes before leaving this dialog.

See Chapter 10, Monitoring the Adapter, on page 221 for a list of all supported microagents.

Many of the following fields can use global variables. Click the Global Variables tab in the project panel to add or modify a global variable.

Enable Standard Microagent

Allows you to turn on or off the standard TIBCO Hawk Microagent. Clicking the globe icon changes the checkbox to a text field, allowing you to specify a global variable. When this is a text field, turn the microagent on and off by entering true or false.

Standard Microagent Name

This is the name for the standard microagent that will be registered with the TIBCO Hawk system. In most cases the default value is used, COM.TIBCO.ADAPTER.adb.%%deployment%%.%%InstanceId%%.

The value for the %%deployment%% global variable can be set or modified by selecting the session icon and then clicking the Global Variables tab in the project panel.

The %%InstanceId%% variable does not need to be set because it is automatically set at run time by the run-time adapter.

Standard Microagent Timeout

Specifies the amount of time the Hawk Agent should wait for HMA method invocations to complete before timing them out. The default is 10000 milliseconds. Normally there is no need to change this value, however, on machines under extreme stress where method invocations are timing out, this new option allows the timeout value to be increased.



Enable Class Microagent

Allows you to turn on or off the instance-specific or class-specific standard TIBCO Hawk Microagent. You can configure how the class microagent is turned on and off in this field: clicking the globe icon changes the checkbox to a a true/false text field.

Class Microagent Name

This is the name for the class microagent that will be registered with the TIBCO Hawk system. In most cases the default value is used, COM.TIBCO.adb.%%deployment%%.%%InstanceId%%.

Class Microagent Timeout

Specifies the amount of time the Hawk Agent should wait for HMA method invocations to complete before timing them out. The default is 10000 milliseconds. Normally there is no need to change this value, however, on machines under extreme stress where method invocations are timing out, this new option allows the timeout value to be increased.

Default Microagent Session

This field is predefined and cannot be changed. The session is automatically generated by TIBCO Designer and will be used by the standard, class, and custom microagents.


| 67

Chapter 4 Adapter Service Options

After configuring an adapter instance, select one or more adapter services for the instance. All configuration tasks are performed in TIBCO Designer.

Topics


• Adding Child Tables, page 76

• Publication Service Tabs, page 71

• Subscription Service Tabs, page 87

• Request-Response Service Tabs, page 97


68 | Chapter 4 Adapter Service Options

Overview

The transport type (Rendezvous or JMS) you select for the run-time adapter determines which quality of service, delivery mode, and wire format the service can use. Only options that are compatible with a service’s transport type will be available for selection.

Quality of service, delivery mode, and wire formats are described below.

Quality of ServiceThis is the level of service that determines how messages are sent.

Possible values are:

• Reliable

(TIBCO Rendezvous transport type only) Reliable message delivery. Ensures that each multicast or broadcast message is received as long as the physical network and packet recipients are working, and that the loss of a message is detected. This choice can compensate for brief network failures because it can retransmit a message on request if the first attempt failed. This choice is appropriate when message delivery is expected but some loss can be tolerated. Messages are received without explicit confirmation.

• Certified

(TIBCO Rendezvous transport type only) Certified message delivery. Offers stronger assurances of message receipt, along with tighter control, greater flexibility and fine-grained reporting. Guarantees that every certified message reaches its intended recipient in the order sent. The message can be sent across network boundaries, and if a network fails, delivery attempts continue until delivery succeeds or until the message's time limit expires.

If certified message delivery is used, data is stored in a ledger file. The size of the ledger depends on several factors, the most important of which is the retention rate of stored data. That is, the ledger grows fastest in response to the cumulative length of undeliverable messages. You must ensure that sufficient disk space is available for the expected size of the ledger.

• Distributed Queue

(TIBCO Rendezvous transport type only) A distributed queue is a group of cooperating transport objects, each in a separate process. To obtain load balancing among servers, the adapter uses distributed queues for one-of-n delivery of messages to a group of servers. Each member of a distributed queue listens for the same subject using the TIBCO Rendezvous Distributed


Overview | 69

Queue listener objects. Even though many members listen for each inbound message (or task), only one member processes the message. For details on distributed queues, see TIBCO Rendezvous Concepts.

See Load Balancing Across Adapter Instances on page 146 and TIBCO ActiveEnterprise Concepts for more information.

• Transactional

Delivery Mode(JMS transport type only) The method of delivery for a JMS message. The semantics for these fields are somewhat more complex than the explanation given here. See the TIBCO Enterprise Message Service User’s Guide for more information.

• Persistent

In general, a message marked persistent will be available to a JMS client even if the TIBCO Enterprise Message Service server goes down. Persistent messages are held in secondary storage in the server and have guaranteed delivery when sent to a topic that has durable subscribers. (If a topic has no durable subscribers, there are no subscribers that need messages resent in the event of a server failure and therefore messages do not need to be saved.) Performance is improved because disk I/O is not required.

• Non-Persistent

(JMS transport type only) A message marked non persistent will not be available to a JMS client if the TIBCO Enterprise Message Service server goes down. These messages are never written to persistent storage.

Wire FormatsServices must use the same wire format to exchange data.

• Rendezvous Message (TIBCO Rendezvous transport type only)

A self-describing wire format used by TIBCO Rendezvous applications. Control information for validation is not sent in the message. If you use this format, the adapter is compatible with adapters not developed with TIBCO Adapter SDK.

• XML Message (TIBCO Rendezvous and JMS transport types)

The Transactional option is no longer supported. The option is included for backward compatibility with version 5.0 and 4.x adapters.



The XML Message wire format conforms to specifically constructed and fully compliant XML Schema (XSD) based on the existing definition of the TIBCO ActiveEnterprise schema.

• ActiveEnterprise Message (TIBCO Rendezvous transport type only)

An externally-described XML wire format supported by the TIBCO Adapter SDK. Control information for validation is sent in the message. If no control information is included, an exception is returned to the subscriber. TIBCO ActiveEnterprise standard wire format provides class information and packing rules for the TIBCO Adapter SDK set of data types. This format allows TIBCO ActiveEnterprise components to perform extra validation on messages sent or received.

See the TIBCO Adapter SDK Programmer’s Guide for details about the control information generated and sent with TIBCO ActiveEnterprise messages.


Publication Service Tabs | 71

Publication Service Tabs

When running as a publisher, the adapter extracts data from the changed rows from database tables and publishes them on appropriate subject names.

You can configure parameters under the following tabs:

• Configuration Tab on page 71

• Tables Tab on page 74

• DB2/OS390 Tab on page 80

• DB2/AS400 Tab on page 81

• Publisher Options Tab on page 81

• Advanced Tab on page 86

Configuration TabThe wire format for the publication and subscription services must be the same, otherwise an error will occur.

Name

You can use the default name or replace it with a name of your choice. Note the following restrictions:

• A service name must use alphanumeric characters. An underscore (_) character can be used. The entire instance name must be less than 80 characters. The space character cannot be used in an instance name.

• A service name cannot use global variables.

Transport Type

Select the transport type (JMS or TIBCO Rendezvous) to be used by the run-time adapter. This selection determines which options appear in the rest of the Configuration tab.

The transport can be configured to use a trusted store and identity resource for use in SSL (Secure Sockets Layer) configurations. TIBCO Rendezvous sessions and JMS topics have an SSL configuration field which uses a dialog to perform SSL configuration.



To enable and configure SSL, in the Project panel, expand the Advanced folder, then expand the Sessions folder. Select the TIBCO Rendezvous session or JMS topic and click Use SSL?. The SSL configuration options are explained in the online help associated with the session dialog. Click the question mark to display the online help.

Quality of Service

(Only available when TIBCO Rendezvous is selected as the transport type) Select the level of service that determines how messages are sent. See Quality of Service on page 68 for a description of these options. • Reliable

• Certified

Wire Format

The wire format in which data will be sent. Note that the wire format for both the publisher and subscriber must be the same, otherwise an error will occur. See Wire Formats on page 69 for a description of these formats.

• ActiveEnterprise Message (TIBCO Rendezvous transport type only)

• Rendezvous Message (TIBCO Rendezvous transport type only)• XML Message

Connection Factory Type

(Only available when JMS is selected as the transport type) Connection Factory objects create JMS connections for sending and receiving messages. Available choices are:

• Topic

Publish-subscribe messaging. A message published to a topic is broadcast to one or more subscribers. All messages published to the topic are received by all services that have subscribed to the topic.

• Queue

Point-to-point messaging. A message sent to a queue is consumed by one and only one receiver. Each message has only one receiver though multiple receivers may connect to the queue. The first receiver to access the queue gets the message. The other receivers do not.



Delivery Mode

(Only available when JMS is selected as the transport type) The delivery mode for each message sending operation. See Delivery Mode on page 69 for a description of each mode.• Persistent

• Non-Persistent



Tables TabYou must set publisher table options before configuring other publisher options.

Source table names can be qualified with a database user name. To access tables in other schemas, the database user specified in the Design-time Connection Tab on page 47 tab must have the proper set of permissions granted. This is described in Referencing an External Schema on page 76.

Icons

• Add Table — Click to display a dialog box that list tables available to the database user specified in the adapter Connection tab. Select the source table to publish from when data is inserted into it.

• Add Child Table — Displays a dialog box from which a secondary table can be added to the configuration.

• Add Table from Other Schema — Allows you to add a table from a different schema than the current database user schema. For instructions on granting access privileges to an external schema, see Referencing an External Schema on page 76.

• Remove Table — Deletes the selected table from the list.

• Re-find Tables from Database — Causes TIBCO Designer to refresh stored table schema information by retrieving new information from the database.

If a primary key is not defined for a table, the update and delete triggers will not be generated for the table. To define a primary key for the table, select the User Key column in a table row. The update and delete triggers will then be defined.



• Load Table Schema from Database — Allows you to load a database table schema, convert it into a TIBCO ActiveEnterprise schema, and store it in the schema folder of an instance.

Allow Key Columns Only

When this box is checked, child columns are joined only to a key column. When unchecked, child columns can be joined to any column.

Select/Deselect All Columns to Publish

When this box is checked, all of the boxes in the Use column are checked, so all columns in the tables are published. Unchecking this box unselects all of the columns. You can also individually check and uncheck the Use boxes.

Tables and Columns

The loaded tables and columns.

indicates a table or child table.

indicates a normal column.

indicates a User Key column.

Type

The primitive type.

AE Type

The primitive type mapped to an TIBCO ActiveEnterprise type.

User Key

Click to define the column as a user key.

Update Trigger?

Check the box to fire a trigger when an UPDATE statement changes a value in the column.

Use?

Click to publish this column when data is changed.



Join To

The name of a parent table column to join to for parent-child relationships.

Referencing an External Schema

To reference an external schema in TIBCO Designer, the default schema must have the proper access privileges. These are set in a command line. In the following syntax, adb_schema refers to the default schema in create_user.sql.

For Oracle, log in as system and grant create any trigger and drop any trigger permissions to the default schema. For example:

grant create any trigger to adb_schemagrant drop any trigger to adb_schema

In addition, Oracle and Sybase users must have permission to SELECT from a source table in an external schema. If table relationships are used, SELECT permission is required for both parent and child tables. SELECT, INSERT, UPDATE, and DELETE permissions are required for accessing a destination table in an external schema.

For Sybase, execute the following before creating catalog tables for the external schema:

sp_role "grant", sa_role, adb_schema

For SQL Server 7.0 and 2000, log in as sa and then execute the following before creating catalog tables for the external schema:

use masterEXEC sp_addsrvrolemember 'adb_schema', 'sysadmin'

For DB2 on AS400 you can avoid table access problems by changing the ActiveDatabase user authority to *ALLOBJ.

Adding Child TablesThis section describes how to add a related child table definition for publishing data.

Data models typically contain tables that share column data through a relationship. You can configure a publishing table to include related data from another table when it publishes. Data from the related table is not copied into the publishing table, but is fetched by reference.

When rows are inserted into the publishing table, a message that includes data from the source table and related (child) table is published. On the subscriber side, a corresponding table with the same columns as the child table associated with the publishing table must be specified.



Adding child tables requires two separate procedures, one for the publisher adapter and another for the subscriber. First, you add child tables for the source table, then you add child tables for the destination table.

The database schema must be the same for all tables, but the table names can be different. If the child table associated with the publishing table and the child table associated with the destination table have different names, you must set a mapping between the child tables.

The following restrictions apply to parent and child tables:

• The child table in the source database and child table in the destination database must have the same columns.

• Currently, child tables can be defined for publishing tables configured as follows:

— TIBCO Rendezvous transport type using the TIBCO ActiveEnterprise wire format

— JMS transport type using the XML wire format

• When parent-child relationships are defined, a subscriber adapter must use the same repository as the publisher adapter.

When you add a child table, TIBCO Designer creates a class object in the repository for the child table, and an association object for the relationship.

After adding child tables for a subscriber adapter, you create mappings between child tables on the publisher side and child tables on the subscriber side. For instructions, see Child Table Mappings Tab on page 92.

To add child tables for parent-child relationships, do the following in the Publication Service Tables Tab:

1. Click the name of the parent table to select it.

2. Click the Add Child Table icon. The Add Table dialog displays.

3. Select the related child table from the list and click OK.

4. Repeat the above steps to add more child tables.

When all of the child tables are added, you then designate a foreign key column as a key in each child table so that a relationship to the parent table can be defined.

5. Click the User Key checkbox for the foreign key column in the child table to select it. A key icon displays next to the column name.

To enable publishing child table related data, the publisher adapter properties file must have the adb.publishChildData option set to on.



You must also specify the relationship between the primary column in the parent table and the foreign key column in the child table.

6. Click in the Join To field for the child table column, and select the name of the parent table primary key column from the drop-down list.

7. Click Apply.

In the following example, a publication service that publishes newly inserted rows into the source table is created. Publisher Options Tab on page 81 explains how to configure the publisher adapter to publish data by value or by reference.



In the following example, a subscription is created such that received data will be inserted into the SUB_ORDER_DETAILS destination table. All three columns of the destination table are configured to receive data.

The following example shows a child table mapped. The left column (Subscriber Child Table Name) contains the subscriber child table and the right column (Publisher Child Table Name) contains the publisher child table.



DB2/OS390 TabThis tab appears when the Vendor field in the Adapter Instance Configuration Tab is set to DB2 OS390.

Database Name

Enter the name of the database that you want to put your publisher table in.

Table Space Name

Enter the name of the table space where the publisher table is located.

Storage Group

Optional. Enter the designator of the storage group that will hold the publisher table indexes.

Buffer Pool

Optional. Enter the name of the buffer pool to be used for indexes.

Index Suffix

Enter a suffix of your choice, up to 13 characters, that the adapter will append to each of the indexes (IDX1_, IDX2_, and IDX3_).

When the DB2 load utility loads rows to the source table, it does not activate the table’s INSERT triggers. Loaded data is not published.



Trigger Suffix

Enter a suffix of your choice, up to 5 characters, that adapter software will append to each of the triggers (T1, T2, and T3).

DB2/AS400 TabThis tab appears when the Vendor field in the Adapter Instance Configuration Tab is set to DB2 AS400.

Trigger Option

• Synchronous (deprecated) — The trigger is written in RPG. When copying from the source table to the publishing table, the prompt is not returned until all data is written.

• Asynchronous — When copying from the source table, data is inserted into a data queue and then inserted into the publishing table asynchronously. The prompt is not blocked, so you can continue working while data is inserted into the publishing table.

If you select Asynchronous, you must go to the adapter’s Design-time Connection tab and add ;transaction isolation level=none to the end of the JDBC URL field value. For example:

• SQL — The trigger is written in SQL. When copying from the source table to the publishing table, the prompt is not returned until all data is written.

Publisher Options TabWhen you add a new publishing table, a new entry is inserted into the repository for the publisher adapter. The entry includes the name of a publishing table along with a sequence, stored procedure, and trigger. A class object for the publication is created in the repository.



Storage Mode

• Publish by Value copies all specified columns in the source table to the publishing table.

• Publish by Reference copies only key column values to the publishing table. If no key column is defined in the database, a substitute non-key column must be defined to publish by reference.

For each publisher service, you must specify a Storage Mode: Publish by Value or Publish by Reference. These options are described next.

Publish by Value

With Publish by Value, all specified columns in the source table are copied to the publishing table. Publishing by value is fast, but does not support some data types, for example Oracle LONG and LONG RAW.

Note the following restrictions on publishing tables when you publish by value:

• Publishing tables cannot contain columns with LONG data types. If you have a source table that contains a column with a LONG data type, that column cannot be specified for inclusion in the publishing table. This is because the trigger generated by the palette cannot refer to the LONG column via the :new construct.

This is an Oracle restriction documented in the Oracle SQL Reference manual. The problem is not detected by Oracle during trigger creation. However, when the trigger fires and it attempts to copy the LONG column value to the publishing table, the database connection will hang for some time and then eventually terminate.

• If you define parent-child relationships between tables, the publishing table that is created for a parent table should not contain a column with a LONG data type. However, a child table can contain a column with a LONG data type. This is because data on child table rows is not copied using the :new construct.

• LONG RAW data is not allowed in the publishing table.

These restrictions do not apply to publishing tables when you publish by reference and LONG or LONG RAW are non-key types.

In the following example, the publishing table P_CUSTOMER is created for the source table, CUSTOMER. When CUSTOMER is updated, the new data will be copied to P_CUSTOMER. The adapter will poll P_CUSTOMER and publish the new data.

In this example, loop detection is enabled. If a subscription exists that uses the same subject and CUSTOMER as the destination table, any changes to CUSTOMER will not be published repeatedly.



Publish by Reference

With Publish by Reference, only key column values are copied to the publishing table. The publish by reference feature allows you to publish data directly from the source table without first copying the data from the source table to a publishing table. A trigger, stored procedure, and publishing table are created, but the publishing table contains the necessary adapter fields and only the key fields of the source table. For more information, see Start or Stop the Adapter on page 106.

The advantage of publishing by reference is that the data to be published is stored just once. Also, data types such as Oracle LONG and LONG RAW are supported for publishing by reference.

A key column or substitute key column is required when publishing by reference, since the publishing table contains only key values. If no column is specified, the publication is not added.

To use a view or other database object as the source table, you can configure the adapter to publish by reference object, where key columns are stored in the publishing table and data to publish is selected from the reference object. For details, see Publishing by Reference Object on page 179.



In the following example, the publisher adapter is configured to publish from the P_CUSTOMER table with a key field CUST_ID. The publishing table is created with the necessary fields and the CUST_ID field. When a row in the P_CUSTOMER table is modified, the trigger fires, populating fields and copying the CUST_ID value to the publishing table. When the adapter polls the publishing table, it detects the new row and selects from the P_CUSTOMER table using the CUST_ID value found in the publishing table. Then the message is published.

Publishing Table

Name of the database table used to store a copy of data to be published. The table name can be qualified using the <schema>.<tableName> format. The publishing table cannot contain any user-created columns where the column name starts with ADB_. These characters are reserved for use by the adapter.

A common practice is to use the publishing table name prefixed by P_. For example, if your source table is MY_ORDER, its publishing table should be named P_MY_ORDER.

A publishing table name must be less than 64 characters.

Update Mode

Select the method by which tables are updated.

• Update updates a row in the destination table only if the row exists.

• Upsert updates a row in the destination table if the row exists. If no such row exists, it performs an insert.



Enable Loop Detection

Select to enable loop detection and prevent an infinite loop from occurring when the publishing and destination table are the same table. Loop detection is disabled for DB2 on z/OS because of DB2 on z/OS feature limitations.

Do Not Generate Triggers

Select to prevent the generation of triggers. Although this option is not recommended, it allows you to manually manage the insertion of data into the publishing table.

Use Alerter

Select to use the alerter. This option appears only if you are using an Oracle database. See Chapter 6, Using an Alerter, on page 109 for details.

Enable Group Messaging

Select to use group messaging.

Group Size

Appears only when Enable Group Messaging is selected. Specify the number of rows to publish in a single message.



Advanced Tab

Destination

(Only available when JMS is selected as the transport type in the Adapter Instance Configuration Tab)

The publisher destination. A service uses a default destination generated using the Domain and Deployment global variables, the adapter acronym, the adapter instance name and the service name. If you use this default destination, make sure the values for Domain and Deployment are not empty.

Alternatively, you can manually enter a destination in this field. The destination does not have to be predefined in the TIBCO Enterprise Message Service server. The destination can be static or dynamic.

See the TIBCO Enterprise Message Service User’s Guide for information about destinations.

Message Subject

(Only available when TIBCO Rendezvous is selected as the transport type in the Adapter Instance Configuration Tab)

Publisher subject. By default, a service uses a message subject that is generated using the Domain and Deployment global variables, the adapter acronym, the adapter instance name and the service name. If you use this default subject, make sure the values for Domain and Deployment are not empty. You can type a TIBCO Rendezvous subject name different from the default in this field.

See TIBCO Rendezvous Concepts for information about specifying subject names.

Endpoint Reference

Displays the endpoint reference. Click the Browse icon to change the endpoint reference, the Go To icon to reconfigure the existing reference, or the Delete icon to delete the reference. Endpoint reference objects are explained in TIBCO Designer Palette Reference.

Class Reference

Click the Go To icon to reconfigure the existing reference. "Clear reference" can be used to remove the reference under Endpoint Reference. Class reference objects are explained in TIBCO Designer Palette Reference.


Subscription Service Tabs | 87

Subscription Service Tabs

When running as a subscriber, the adapter listens on a subject, receives messages and updates the relevant tables in its associated database. The data is then available to other applications that have access to the database.



• Table Tab on page 90

• Child Table Mappings Tab on page 92

• Child Exception Table Mappings Tab on page 93

• Subscriber Options Tab on page 94


Configuration Tab

Name

Type a name unique among all subscribers defined for this project. You can use the default name or replace it with a name of your choice.

• A service name must use alphanumeric characters, including underscore (_). You cannot use a blank space. The entire instance name must be less than 80 characters.


Transport Type

Select the transport type (JMS or TIBCO Rendezvous) to be used by the run-time adapter. This selection determines which options appear in the rest of the Configuration tab.





Quality of Service

(Only available when TIBCO Rendezvous is selected as the transport type) Select the level of service that determines how messages are sent. See Quality of Service on page 68 for a description of these options.

• Certified

• Reliable


Wire Format


• Rendezvous Message (TIBCO Rendezvous only)

• XML Message (TIBCO Rendezvous or JMS)

• ActiveEnterprise Message (TIBCO Rendezvous only)


Connection Factory objects create JMS connections for sending and receiving messages.

• Topic (JMS only)

A message published to a topic is broadcast to one or more subscribers. All messages published to the topic are received by all services that have subscribed to the topic. This messaging model is known as publish-subscribe.

• Queue (JMS only)

A message sent to a queue is consumed by one and only one receiver. Each message has only one receiver though multiple receivers may connect to the queue. The first receiver to access the queue gets the message. The other receivers do not. This messaging model is known as point-to-point.



Delivery Mode

(Only available when JMS is the transport type and Topic is the connection factory type) The delivery mode for each message sending operation. See Delivery Mode on page 69 for a description of each mode.• Persistent

• Non-Persistent



Table TabYou must set subscriber table options before configuring other subscriber options.

An incoming message need not contain data for all the columns defined in the subscriber table. An adapter can be configured to expect only a subset of the columns. The adapter checks the repository for attributes defined in the subscriber table's class object definition. When a message arrives, the adapter iterates through the attributes in the subscriber table’s class object definition and looks for those same attributes in the incoming message.

Subscriber table names can be qualified with a schema name, such as SCOTT.EMP. To access tables in other schemas, the database user defined in the adapter Connection tab must have the proper set of permissions granted. For details, see page 10.

When configuring the destination table, subscribe only to columns that should be updated. If you subscribe to a column that should not be updated and a message arrives with no data for that column, a NULL will be written to that column. For example:

• If a source table is configured to send data for columns, c1, c2 and c3 and the destination table is configured to receive data for columns c1, c2, c3, c4, and c5:

— For the TIBCO Rendezvous message wire format, columns c1, c2, and c3 will get the data and columns c4 and c5 will get a NULL value.

— For the TIBCO ActiveEnterprise message and XML message wire format, columns c4 and c5 will be ignored and take on whatever default values they are supposed to have.

• If a source table is configured to send data for columns, c1, c2, and c3 and the destination table is configured to receive data for columns c1, c2, and c3 but not configured to receive data for columns c4 and c5, columns c4 and c5 will retain the defaults applicable to both tables.

When the publisher table is configured to use parent-child relationships, the subscriber adapter must use the same repository as the publisher adapter.



Icons

Add Table — Click to display a dialog box that list tables available to the database user specified in the adapter Connection tab. Select the table to be used as the subscription table where data is inserted when received.

Add Child Table — Displays a dialog box from which a secondary table can be added to the configuration.

Add Table from Other Schema — Allows you to enter a schema. For instructions on granting access privileges to an external schema, see Referencing an External Schema on page 76.

Remove Table — Deletes the selected table from the configuration.

Re-find Table from Database — Causes TIBCO Designer to refresh stored table schema information by retrieving new information from the database.

Allow Key Columns Only — If selected, child columns are joined only to a key colulmn. If unselected, child columns can be joined to any column.

Columns

Tables and Columns — Lists the table and columns.

Type — Lists the primitive type.

AE Type — Lists the primitive type mapped to an TIBCO ActiveEnterprise type.

User Key — Click to define as a user key.

Use? — Click to enable the column to be updated when a message arrives. Subscribe only to columns that should be updated. Columns not configured to



receive data will retain their default value. For columns that have been configured to receive data but do not get updated:

— In the TIBCO Rendezvous wire format, the columns get a NULL value.

— In the TIBCO ActiveEnterprise and XML wire formats, the columns are ignored in the statement.

Join To — Joins two tables.

Child Table Mappings TabChild tables between a publisher and subscriber must be mapped unless the tables have the same name.

The following example shows a child table mapped. The left column (Subscriber Child Table Name) contains the subscriber child table and the right column (Publisher Child Table Name) contains the publisher child table.

Subscriber Child Table Name

Displays the child tables that can be mapped. Entries are automatically created when you add child tables for a subscription service.

Publisher Child Table Name

Click in this column and type the name of a publisher child table. Prefix the table name with a database user if your database requires it.

These options are only active when child tables have been specified on the Table Tab.



Child Exception Table Mappings TabEach child table can be configured to use an exception table.

Subscriber Child Table Name

Displays the child tables that can be mapped. Entries are automatically created when you add child tables for a subscription service.

Child Exception Table Name

Click in this column and type the name of the exception table. Prefix the exception table name with a database user if your database requires it.

Subscriber Options for DB2 on OS/390 (z/OS)This tab appears when the Vendor field in the Adapter Instance Configuration Tab is set to DB2 OS390.

Database Name

Enter the name of the database that you want to put your exception table in.

Table Space Name

Enter the name of the table space where the exception table is located.

These options are only active when child tables have been specified on the Table Tab.



Subscriber Options Tab

Exceptions Table

Name of exception table where data is written if the adapter cannot write to the subscriber table. This table will hold messages that caused an exception. If the table does not exist, the subscription service creates it. For details, see Exception Table on page 161.

The exception table cannot contain any user-created columns where the column name starts with ADB_. These characters are reserved for use by the adapter.

Use Opaque Exceptions Table

Select the box to use an opaque exceptions table. The table records each message (entirely) into a column, along with the error message. A message is logged in the exceptions table if the subscription service fails to generate records in the destination table or the adapter fails to insert a message into an exception table.

For DB2/OS390 databases, you must create a Large Objects (LOB) tablespace before using the opaque exceptions table, which will use the LOB tablespace.

Opaque Exceptions Table

Type the name for the opaque exceptions table.

Pre Commit Stored Procedure

The value entered here represents the name of a stored procedure the subscriber will call after the database insert, update, or delete and prior to commit. You can use this stored procedure as a hook to accomplish further processing inside the database and return the results to the adapter. No value means the adapter will not call a stored procedure.

Reply Sender Quality of Service

If the subscriber must send a reply to the sender, this value identifies the quality of service or delivery mode to use when sending the reply. See Quality of Service on page 68 and Delivery Mode on page 69 for a description of these fields.

• Reliable (TIBCO Rendezvous transport type only)

• Certified (TIBCO Rendezvous transport type only)

• Persistent (JMS transport type only)

• Non-Persistent (JMS transport type only)



Use Separate Service Thread

When selected, a new session is created and the service endpoint is moved to this session. You can select other subscription service sessions by changing the session reference field (under the Advanced tab). This allow multiple services to share the same session and dispatcher.

If you are using the JMS transport, you must set the JMS Destination Prefix. See JMS Destination Prefix on page 58 for details.

Note that:

• All subscription service threads reconnect separately if the database connection is lost.

• TIBCO Hawk statistics display all subscription service thread information and operation count.

• Debug messages provide the subscription service thread name in order to distinguish which message is from which subscription service thread.

• If a subscription service thread encounter fatal errors such as failed to rollback transaction or failed to update the exception table, the adapter terminates.

• At startup the subscription service dispatchers wait for all component to start before dispatching subscriber messages.

Advanced Tab

Message Subject

(TIBCO Rendezvous transport type only)

By default a service uses a message subject that is generated using the Domain and Deployment global variables, the adapter acronym, the adapter instance name and the service name. If you use this default subject, make sure the values for Domain and Deployment are not empty. You can type a TIBCO Rendezvous subject name different from the default in this field. See TIBCO Rendezvous Concepts for information about specifying subject names.

Destination

(JMS transport type only)

The subscriber destination. A service uses a default destination generated using the Domain and Deployment global variables, the adapter acronym, the adapter instance name and the service name. If you use this default destination, make sure the values for Domain and Deployment are not empty.





Endpoint Reference

Displays the endpoint reference. Click the Browse icon to change the endpoint reference or the Go To icon to reconfigure the existing reference. You can also click the Delete icon to remove the reference. Endpoint reference objects are explained in TIBCO Designer Palette Reference.

Class Reference

Click the Go To icon to reconfigure the existing reference. Class reference objects are explained in TIBCO Designer Palette Reference.


Request-Response Service Tabs | 97

Request-Response Service Tabs

This service is often called a Request Reply Server or RPC (Remote Procedural Call) Server. When running as a Request-Response Service, the adapter receives requests from other applications, parses them, calls the appropriate component interface API to set the input fields, then calls another set of component interface APIs to get the output fields. The output fields are wrapped in a schema and sent back to the caller.

A Request-Response Service is renamed ADBServer when you drag its icon into the design panel. Configuration involves specifying its name, quality of service and wire format. A server operation allows the adapter to process requests from client applications and return results in a response to the client.

When using the RPC server, you can configure custom operations. These are configured under the Call Operation Tab.



• Call Operation Tab on page 100


Configuration Tab

Name

Type a name unique among all request-response services defined for this project. You can use the default name or replace it with a name of your choice.

• A service name must use alphanumeric characters, including underscore (_). You cannot use a blank space. The entire instance name must be less than 80 characters.


Transport Type

Select the transport type (JMS or TIBCO Rendezvous) to be used by the run-time adapter. This selection determines which options appear in the rest of the Configuration tab. Only the options that are available with the selected Transport Type, Connection Factory Type, Mode or Wire Format are displayed.





Quality of Service

Select the level of service that determines how messages are sent. See Quality of Service on page 68 for a description of these options.• Certified

• Reliable



Connection Factory objects create JMS connections for sending and receiving messages.

• Topic

Publish and subscribe messaging. A message published to a topic is broadcast to one or more subscribers. All messages published to the topic are received by all services that have subscribed to the topic.

• Queue

Point-to-point messaging. A message sent to a queue is consumed by one and only one receiver. Each message has only one receiver though multiple receivers may connect to the queue. The first receiver to access the queue gets the message. The other receivers do not.

Delivery Mode

(Only available when JMS is selected as the transport type) The delivery mode for each message sending operation. See Delivery Mode on page 69 for a description of Persistent and Non-Persistent.

• Durable (see Persistent)

• Non-Durable (see Non-Persistent)



Mode

The server mode.

• Request Reply is the basic request-reply mode. The server subscribes to the request and publishes the reply.

• RPC allows you to configure an adapter to act as an RPC (remote procedure call) server on behalf of a client.

Wire Format


• Rendezvous Message (TIBCO Rendezvous only)

• XML Message (TIBCO Rendezvous or JMS)

• ActiveEnterprise Message (TIBCO Rendezvous only)

Use Custom Operations

This checkbox appears when the Mode field is set to RPC. Selecting this checkbox and clicking Apply creates the Call Operation Tab.



Call Operation TabThis tab appears when the mode is set to RPC and the Use Custom Operations box is checked in the Request-Response Service Configuration tab.

Click New to enter data into the fields. After making changes, click Apply.

TIBCO Designer retrieves the signature of each stored procedure from the database.

Name

Enter a unique name for the call operation.

Catalog/Package

(Optional, only applicable to databases that have more than one catalog or package.) The catalog or package in which the procedure resides. This name is used to resolve naming conflicts if more than one catalog or package in the database has the selected procedure with the same name. See your database documentation for more information about catalogs and packages.

Schema

(Optional) The schema in which the procedure resides. This name is used to resolve naming conflicts if more than one schema in the database has the selected procedure with the same name.

Procedure Name

Name of the database procedure to call.

If you change the stored procedure or database connection while editing your project, you must return to this dialog and click Refresh to retrieve the changes from the database.



Select Procedure

Queries the database for available procedures for the Procedure Name field. Clicking this button displays a dialog showing the available procedures.

To select a procedure:

a. Select a call operation from the list on the left. Its input and output parameters, if any, are displayed in the fields on the right side of the dialog.

b. Click Select Procedure.

Maximum Rows

(Optional) The maximum number of rows to retrieve.

One way

(Optional) Check this box if you want to execute this procedure using a one-way operation. If left unchecked, the procedure will execute using a two-way operation.



Advanced Tab

Message Subject

(TIBCO Rendezvous transport type only)

By default a service uses a message subject that is generated using the Domain and Deployment global variables, the adapter acronym, the adapter instance name and the service name. If you use this default subject, make sure the values for Domain and Deployment are not empty. You can type a TIBCO Rendezvous subject name different from the default in this field. See TIBCO Rendezvous Concepts for information about specifying subject names.

Destination

(JMS transport type only)

The subscriber destination. A service uses a default destination generated using the Domain and Deployment global variables, the adapter acronym, the adapter instance name and the service name. If you use this default destination, make sure the values for Domain and Deployment are not empty.



Reply Subject

Type a TIBCO Rendezvous subject name that the adapter uses to respond, if no response subject is specified in the request message.

Endpoint Reference

Displays the endpoint reference. Click the Browse icon to change the endpoint reference or the Go To icon to reconfigure the existing reference. You can also click the Delete icon to remove the reference. Endpoint reference objects are explained in TIBCO Designer Palette Reference.

Class Reference

Click the Go To icon to reconfigure the existing reference. Class reference objects are explained in TIBCO Designer Palette Reference.


| 103

Chapter 5 Deploying and Starting the Adapter Using TIBCO Administrator

Topics

• Create an EAR File in TIBCO Designer, page 104

• Deploy the Project, page 105

• Start or Stop the Adapter, page 106

• Monitor the Adapter, page 107

This chapter provides an overview about deploying, starting, stopping, and monitoring adapter services using the TIBCO Administrator web interface.


104 | Chapter 5 Deploying and Starting the Adapter Using TIBCO Administrator

Create an EAR File in TIBCO Designer

Generate an Enterprise Archive file (EAR) that contains information about the adapter services to deploy.

The EAR file contains information on what you wish to deploy. This could be one or more adapter services, one or more TIBCO ActiveMatrix BusinessWorks process engines, or both.

In TIBCO Designer, follow these steps to create an EAR:

1. Configure the adapter services.

2. Drag and drop the Enterprise Archive resource from the palette panel to the design panel. If there are any configured adapter services in your project, an Adapter Archive resource becomes available in the palette panel.

3. Drag the Adapter Archive into the design panel and specify information in the Configuration tab, then click Apply.

4. Go to the Enterprise Archive and click Build Archive to create the archive file.

See Also

See the TIBCO Designer User’s Guide for more information about this procedure. The guide is available from the Designer Help menu.

Building an archive creates the EAR file, which you can then deploy from TIBCO Administrator. If you make changes to the business processes or adapter services included in the archive, you need to rebuild the archive. Saving the project does not affect the archive.


Deploy the Project | 105

Deploy the Project

Before deploying a project, the machine on which the adapter is installed must be part of a TIBCO administration domain. After you have installed the TIBCO Administration Server, any machine on which you install TIBCO Runtime Agent (required by an adapter) can be added to the administration domain. The TIBCO software installed on the machine is then visible and accessible via the TIBCO Administrator GUI.

When you deploy a project, startup scripts and other information about the different components are sent to the machines to which the components were assigned. The project data store and TIBCO Administration Server are updated with the deployed components.

To deploy a project:

1. Import the EAR file into TIBCO Administrator Enterprise Edition.

2. Assign adapter archives in the EAR file to adapters installed in the administration domain and likewise assign process archives to process engines.

3. Specify startup options for each adapter service.

Password Handling

At design-time, the adapter uses a password to connect to the backend application and fetch metadata. At run-time, the adapter uses a password to connect to the back-end application and interoperate with it. If you create a 4.x configuration using TIBCO Designer 5.1, and use the configuration against a 4.x adapter version, some special considerations are required for security.

When deploying the adapter check the Service property of the global variable in the global variables section, then go to the Advanced tab of the adapter archive and set the password value under the Run-Time Variables section.

See Also

See the TIBCO Administrator User’s Guide for an introduction to the TIBCO administration domain and detailed information about the above steps.

Do not set the password to type Password in the global variables section for adapter configurations that are set to AE Version 4.0 or AE Version 5.0 (in the Configuration tab Version field) or any intermediate version.



Start or Stop the Adapter

The TIBCO Administrator Application Management module allows you to start, and stop deployed applications.

To start an adapter service from the module:

1. In the Administrator GUI left pane, expand Application Management > Application-Name > Service Instances.

2. In the Service Instance panel, select the check box next to the adapter service.

3. Click the Start Selected button.

The status changes from Stopped to Starting up to Started.

4. To stop the adapter service, click the Stop Selected button.

See Also

See the TIBCO Administrator User’s Guide for more information.


Monitor the Adapter | 107

Monitor the Adapter

TIBCO Administrator offers a number of monitoring options.

• Specify alerts and TIBCO Hawk rulebases for each machine in the domain.

• Specify alerts and Hawk rulebases for each adapter service.

• View the log for each adapter service.

See Also

See the TIBCO Administrator User’s Guide for information about configuring the above monitoring options.




| 109

Chapter 6 Using an Alerter

This chapter explains how to configure and start an alerter for each supported database.

Topics

• Introduction, page 110

• Setting Up and Starting the Alerter on Oracle or Microsoft SQL Server 2005, page 111

• Setting Up and Starting the Alerter on Microsoft SQL Server 2000, page 115

• Setting Up and Starting the Alerter on Sybase SQL Server, page 118


110 | Chapter 6 Using an Alerter

Introduction

The alerter process is used to asynchronously notify running publication service instances of database changes. The adapter does not need to poll its publishing table for existence of new rows. Use the alerter only when database changes are infrequent.

The procedures can be executed on the SQL command line or through any supported Application Programming Interfaces (the procedures cannot be invoked successfully from within a trigger). The procedures commit the inserts into the database table and notify the adapter of the commit.

Because the alerter process runs against a database instance, not user schema, an adapter instance cannot poll the publishing table name for just one user schema. For example, if the same publishing table exists for two user schemas and two instances are monitoring the same publication table, each instance with a different user, both instances will check the same publication table whenever the alerter checks for changes in the publication table.

The alerter is available on the following databases. The alerter is not available on DB2 or SybaseAdapter Server Anywhere.

Oracle and Microsoft SQL Server 2005

An alerter for Oracle is part of the adapter and available on all operating systems supported by the adapter. The alerter can be used in the TIBCO ActiveMatrix Adapter for Database environment in cases where changes are infrequent and polling by the publication service would result in unnecessary and expensive operations. The alerter is supported on both the Rendezvous and JMS transports.

Microsoft SQL Server 2000

An alerter function for Microsoft SQL Server is available on all operating systems supported by the adapter. The procedures are executed on the SQL command line itself, and there is no separate program. The alerter functionality is embedded in a dynamic linked library. The alerter is only supported on the Rendezvous transport.

Sybase SQL Server

An alerter function for Sybase SQL Server is available on Microsoft Windows along with Solaris. The procedures are executed on the SQL command line itself, and there is no separate program. The alerter functionality is embedded in a dynamic linked library. The alerter is only supported on the Rendezvous transport.


Setting Up and Starting the Alerter on Oracle or Microsoft SQL Server 2005 | 111

Setting Up and Starting the Alerter on Oracle or Microsoft SQL

Server 2005

The alerter on Oracle uses the Oracle AQ package and on MS SQLServer 2005 uses the Service Broker component to send and receive information between sessions, asynchronously allowing two or more sessions in the same database instance to communicate.

As shown in the following figure, when a source table is updated with data and the commit_and_notify procedure executed, a trigger copies the data to the publishing table and notifies the Oracle AQ or MS SQLServer 2005 that a publishing table has changed. The alerter gets the notification from Oracle AQ or MS SQLServer 2005 and sends a notification message to the adapter that a publishing table has changed. The adapter then queries all its configured publishing tables for the new data and sends it on a subject to the TIBCO transport. One or more instances can be notified.

Figure 5 Alerter Operation (Oracle and MSSQLServer 2005)

If there are multiple publishing tables under the same database account, you can use the commit_and_notify_table procedure to specify that only a particular table be checked by the adapter. This prevents the adapter from needlessly checking all its publishing tables for updates when only one table has been updated with new data. This notification can be sent to one or more instances.

commit_and_notify

Source Table

Publishing Table

Oracle AQ/ MS Microsoft SQL Server 2005

Service Broker

Alerter

TIBCO Adapter for ActiveDatabase

1. Send Notification

2. Receive Notification

3. Poll



The following object types and stored procedures are available for committing messages when the alerter is used:

Name Type Description

abd_alerter_qtbl

(Oracle only)AQ queue table

Queue table for all alerters.

adb_alerter_q

(Oracle only)AQ queue Queue for all alerters.

adb_alerter_typ Service Broker Message Type

Object type for all messages.

adb_alerter_contract

(Microsoft SQL Server 2005 only)

Service Broker Contract

The contract type used. The contract specifies the message types that can be used.

<instanceId>_q


Service Broker Queue

Created by the config_alerter stored procedure.

<instanceId>_send_service


Service Broker Service


<instanceId>_rcv_service


Service Broker Service


adb_alerter_typ Object Object type for the payload message.

commit_and_notify Procedure Sends a change notification to all adapters.

commit_and_notify_agent <adapter-name>

Procedure Sends a change notification to the named adapter.

commit_and_notify_table <tablename>

Procedure Sends a change notification for a named table.

stop_alerter Procedure Stops all running alerters.

stop_alerter_agent <adapter-name>

Procedure Stops the alerter for the named adapter.

config_alerter Procedure Alerter call to register itself as an AQ subscriber.


Setting Up and Starting the Alerter on Oracle or Microsoft SQL Server 2005 | 113

Before StartingWhen using the alerter functionality with Oracle database, make sure you have executed the create_user.sql and alerter.sql files against the database account used by the adapter, as described in Post-Installation Tasks in the TIBCO ActiveMatrix Adapter for Database Installation.

Configuring and Starting the AlerterTo use the alerter, you must install the Oracle AQ or Microsoft SQL Server 2005 package. See your Oracle documentation for information about installing the Oracle AQ or Microsoft SQLServer 2005package. When configuring the publication service, you must enable the alerter by selecting the Use Alerter option under the Publisher Options tab. When the adapter starts, the alerter will also start.

Oracle Alerter ExampleThis section describes how to execute the commit_and_notify_table stored procedure. It assumes the rvpub.tra configuration has its data source name, user account name (demo) and password defined in the repository. The rvpub adapter publication service has been configured with the Use Alerter option selected.

1. Start the adapter publication service instance by typing:

adbagent --propFile rvpub.tra rvpub.tra

2. After inserting a message, execute the commit_and_notify_table stored procedure against the publishing table monitored by the adapter. For example, if the adapter is monitoring the PUB_ORDER table:

sqlplus demo/demo insert into ORDER_TABLE values(111,'Oak Table',499.95);SQL> execute commit_and_notify_table (’PUB_ORDER’);

cleanup_alerter Procedure Alerter call to unregister itself as an AQ subscriber.

listen_pipe Procedure Alerter call to block on alerts.

Name Type Description

If an adapter that uses the alerter is not shutdown cleanly, you must call the cleanup_alerter stored procedure before restarting the adapter. The procedure is normally called by the adapter when it shuts down cleanly.



The procedure puts a message on an Oracle AQ. The publisher adapter then reads its publishing table and sends a message containing the changed data on its configured subject.

This example shows how to tell the adapter instance to poll a single table, PUB_ORDER, for changes. Use the commit_and_notify stored procedure to poll all publishing tables for changes.


Setting Up and Starting the Alerter on Microsoft SQL Server 2000 | 115

Setting Up and Starting the Alerter on Microsoft SQL Server 2000

On Microsoft SQL Server the alerter mechanism is an extended stored procedure in a dynamic link library (provided with installation media) that is dynamically loaded into SQL Server.

Two stored procedures are available for sending messages when using the alerter:

• notify. Sends a TIBCO Rendezvous message to the given instance or instances that one or more publishing tables have been updated.

• notifytable table_name. Sends a TIBCO Rendezvous message to the given instance or instances that the given publishing table has been updated.

Before StartingA database account must have been created for running the adapter. This is typically done after installation. See the TIBCO ActiveMatrix Adapter for Database Installation for details.

Alerter SetupBefore using the alerter you must copy the alerter dynamic link library adbalerter_mssql.dll to the Binn directory where SQL Server is installed, add the notify and notifytable external stored procedures to the master database and set execute permission on the procedures for the database account used by the adapter.

1. Open a command window and change directory to the <install-path>/tibco/adapter/adadb/5.5/lib directory.

2. Copy the alerter library to the Binn directory where SQL Server is installed, for example:

copy adbalerter_mssql.dll \Mssql7\Binn

3. Ensure the procedures are not already defined in the master database:

isql -Usa -Ppassword1> sp_dropextendedproc notify2> go1> sp_dropextendedproc notifytable2> go1> DBCC adbalerter_mssql(FREE)2> goDBCC execution completed. If DBCC printed error messages,

contact your system administrator.



4. Add the procedures to the master database:

1> sp_addextendedproc 'notify','adbalerter_mssql.dll'2> go1> sp_addextendedproc 'notifytable','adbalerter_mssql.dll'2> go

5. Set access permissions for the database account used by the adapter:

1> grant execute on notify to database_account2> go1> grant execute on notifytable to database_account2> go1> exit

Using the AlerterWhen using the alerter, the adapter instance must be started with polling disabled. In this example, after adding a row in a table monitored by the adapter, the alerter procedure is invoked and the adapter checks for updates in its publishing table. The database server name is itaska and database name is activedb.

1. In a command window, start an adapter instance with polling disabled (set the adb.PollingInterval value to 0 in the properties file):

adbagent --propFile propFilename

2. In another command window, insert a row into the database table being monitored by the adapter, commit the transaction, then invoke the notify procedure:

isql -Uadb -Padb -dactivedb1> insert into ORDER_TABLE values(406,'walnut table',9899.89)2> go 1> execute itaska.master.dbo.notify @agentname='rvpub'2> goNotifying: rvpub

You can also execute the notify procedure for more then one adapter instance. Each configuration name must be separated using a space character. For example:

1> execute itaska.master.dbo.notify @agentname='rvpub rvpub2 rvpub3'

2> go

The notifytable procedure can be invoked for one or more adapter instances and one publishing table. If the database used by the adapter and master database are on the same database server, the server prefix need not be given.

1> execute itaska.master.dbo.notifytable @agentname='rvpub rvpub2',@tablename='P_ORDER_TABLE’


Setting Up and Starting the Alerter on Microsoft SQL Server 2000 | 117

2> go

Modifying RV Configuration

The RV configuration parameters used by the alerter to connect to a remote daemon can be modified by including the configuration parameters when invoking the store procedure.

For example,

execute master.dbo.notify @agentname='rvpub', @service='7500', @network=’lan0’, @daemon='10.105.176.21:7500'

execute master.dbo.notifytable @agentname='rvpub rvpub2', @pubtable='ADB_ORDER', @service='7500', @network=’lan0’, @daemon='10.105.176.21:7500'



Setting Up and Starting the Alerter on Sybase SQL Server

On Sybase SQL Server the alerter mechanism is an extended stored procedure in a dynamic link library (provided with the installation media) that is dynamically loaded into Sybase SQL Server. Two stored procedures are available for sending messages when the alerter is used:

• notify. Sends a TIBCO Rendezvous message to the given configuration or configurations that one or more publishing tables have been updated.

• notifytable table_name. Sends a TIBCO Rendezvous message to the given configuration or configurations that the given publishing table has been updated.

Before StartingA database account must be created for running the adapter. This is typically done after installation. See the TIBCO ActiveMatrix Adapter for Database Installation for details.

Alerter Setup on SolarisBefore using the alerter you must copy the alerter shared library to the Sybase/lib directory, add the notify and notifytable external stored procedures to the master database and set execute permission on the procedures for the database account used by the adapter. The name of the shared library file is adbalerter_sybase.so.

1. Open a command window and change directory to the <install-path>/tibco/adapter/adadb/5.5/lib directory. Copy the alerter library as follows:

cp adbalerter_sybase.so sybase/lib

2. Ensure that the procedures are not already defined in the master database:

isql -Usa -Ppassword1> sp_dropextendedproc notify2> go1> sp_dropextendedproc notifytable2> go1> sp_freedll "adbalerter_sybase"2> go

You may need to shut down and restart your XP SERVER after copying the shared library to the sybase/lib directory.


Setting Up and Starting the Alerter on Sybase SQL Server | 119


1> sp_addextendedproc 'notify','adbalerter_sybase.so'2> go1> sp_addextendedproc 'notifytable','adbalerter_sybase.so'2> go

This example shows how to add procedures on Solaris.



5. Add the <install-path>/tibco/tibrv/lib/libtibrv.so to the Sybase LD_LIBRARY_PATH variable.

6. Add <install-path>/tibco/tibrv/bin/rvd to the Sybase PATH variable.

Alerter Setup on Microsoft WindowsBefore using the alerter you must copy the alerter dynamic link library, add the external stored procedures to the master database and set execute permission on the procedures for the database account used by the adapter.

1. Open a command window and change directory to the <install-path>\tibco\adapter\adadb\5.5\lib directory. Copy the alerter library as follows:

copy adbalerter_sybase.dll \Sybase\dll

2. Ensure that the procedures are not already defined in the master database:

isql -Usa -Ppassword1> sp_dropextendedproc notify2> go1> sp_dropextendedproc notifytable2> go1> sp_freedll "adbalerter_sybase"2> go


1> sp_addextendedproc 'notify','adbalerter_mssql.dll'2> go1> sp_addextendedproc 'notifytable','adbalerter_mssql.dll'2> go

You may need to shutdown and restart your XP SERVER after copying adbalerter_sybase.dll to the \Sybase\dll directory.





5. Add the <install-path>\tibco\tibrv\lib\libtibrv.lib to the Sybase LD_LIBRARY_PATH variable.

6. Add <install-path>\tibco\tibrv\bin\rvd to the Sybase PATH variable.

Using the AlerterWhen using the alerter, an adapter instance must be started with polling disabled. In this example, after adding a row in a table monitored by the adapter, the alerter procedure is invoked and the adapter checks for updates in its publishing table. The database server name is itaska and database name is activedb.

1. In a command window, start the adapter instance with polling disabled (set the adb.PollingInterval value to 0 in the properties file):adbagent --propFile propFilename

2. In another command window, insert a row into the database table being monitored by the adapter, commit the transaction, then invoke the notify procedure:

isql -Uadb -Padb -dactivedb1> insert into ORDER_TABLE values(406,'walnut table',9899.89)2> go1> execute itaska.master.dbo.notify @agentname='rvpub'2> goNotifying: rvpub

You can also execute the notify procedure for more than one adapter instance. Each configuration name must be separated using a space character. For example:

1> execute itaska.master.dbo.notify @agentname='rvpub rvpub2 rvpub3'

2> go

The notifytable procedure can be invoked for one or more adapter instances and one publishing table. For example:

1> execute itaska.master.dbo.notifytable@agentname='rvpub rvpub2',@tablename='P_ORDER_TABLE'

2> go

If the database used by the adapter and master database are on the same database server, the server prefix need not be given.


Setting Up and Starting the Alerter on Sybase SQL Server | 121

Modifying RV Configuration

The RV configuration parameters used by the alerter to connect to a remote daemon can be modified by including the configuration parameters when invoking the store procedure.

For example,

execute master.dbo.notify @agentname='rvpub', @service='7500', @network=’lan0’, @daemon='10.105.176.21:7500'

execute master.dbo.notifytable @agentname='rvpub rvpub2', @pubtable='ADB_ORDER', @service='7500', @network=’lan0’, @daemon='10.105.176.21:7500'




| 123

Chapter 7 Using the Request-Response Service

This chapter explains the request-response service, load balancing, and Oracle REF data type support features.

See the TIBCO ActiveMatrix Adapter for Database Examples Guide for request-response exercises.

Topics

• Introduction, page 124

• Request-Response Mode, page 126

• RPC Mode, page 134

• Load Balancing, page 144

• Oracle REF Data Type Support, page 151


124 | Chapter 7 Using the Request-Response Service

Introduction

Request-response operations are set up using the Request-Response Service dialog. This dialog is described in Configuration Tab on page 97.

Using request-response semantics, client applications can send SQL statements, stored procedures, or both on a specified subject to an adapter instance. The adapter processes the request and returns the results in a response to the client.

1. A request may be a query to the database or any DDL or DML command to be performed on the database. Multiple statements, procedures or both can be sent in one message.

2. The adapter executes the statement, procedure or both. Requests within a message are processed as a single transaction.

3. The adapter sends back the response to the Inbox, subject or response subject that was set up by the client application. The response consists of a result code or one or more result sets, based on the request. A response can also be an error code or error description, if the request was not successful.


Introduction | 125

Multiple adapter instances can be configured so that any one of the configurations receives and processes the request:

• For a single configuration, the number of threads can be configured that will be responsible for processing application requests. See Load Balancing in a Single Instance Using Threads on page 144 for details.

• For multiple configurations, TIBCO Rendezvous Distributed Queuing can be used to balance the load across a number of configurations in a queue. The task will be assigned to the least loaded member of the queue. See Load Balancing Across Adapter Instances on page 146 for details.

Use of Quotes in Microsoft SQL Server

When constructing a request in your application that will be processed by the adapter against a Microsoft SQL Server database, you must take care when using quotes.

For example, the following procedure is part of a request from an application. Double quotation marks are used, which is incorrect. An error will be returned.

select @qry = "Update " + @tablename + " set ORDER_DESCRIPTION = 'UPDATE TEST'" + ", ORDER_PRICE = 10109.25"

The following procedure is the same as above, but uses single quotes. It will be correctly processed.

select @qry = 'Update ' + @tablename + ' set ORDER_DESCRIPTION = ''UPDATE TEST'', ORDER_PRICE = 10109.25'

See Delimited Identifiers in your Microsoft SQL Server documentation for details.



Request-Response Mode

TIBCO ActiveMatrix Adapter for Database request-response mode supports the Rendezvous, ActiveEnterprise and XML message formats.

Requests

A request can contain one or more SQL statements, stored procedures, or both to be executed as a transaction. The text of the SQL statement follows the conventions for ODBC SQL syntax. All SQL statements supported by the DBMS are allowed and placeholders (represented by a question mark, ‘?’) are permitted in the SQL statement, conforming to the ODBC rules. For performance reasons, it is recommended to use a SQL statement. The ‘?’convention should only be used to bind binary data or call stored procedures.

An adapter instance is configured for request-response activity as described in Request-Response Service Tabs on page 97. In so doing, the adapter listens on the subject on which the application sends requests. The application uses TIBCO Rendezvous or the TIBCO Adapter SDK to send a self-describing message containing a request on the agreed subject to the adapter instance.

The following are supported:

• DDL and DML SQL statements

• The column size of the return resultset should not exceed 128

• Multiple statements or procedures in a single transaction that send a nested TIBCO Rendezvous message

• rv_Send(), rv_SendWithReply() and rv_Rpc() calls

• TIBCO Rendezvous C, C++, and Java APIs

Requests are described in the following sections:

• TIBCO ActiveEnterprise or XML Message Request Format, page 127

• TIBCO Rendezvous Message Request Format, page 127

If rv_SendWithReply() or rv_Rpc() calls are used, the adapter sends the response to the Inbox or subject that was set by the application for the response. Otherwise, the adapter sends the response on the response subject that was set at configuration time.


Request-Response Mode | 127

Responses

A response from the adapter to a client application has a result code and one or more result sets. Each result set contains nested self-describing messages, each of which encodes a result row, such as that returned from a query. A response can also return an error code and error description if the request was not successful.

Responses are described in the following sections:

• TIBCO ActiveEnterprise or XML Message Response Format, page 130

• TIBCO Rendezvous Message Response Format, page 130

TIBCO ActiveEnterprise or XML Message Request FormatIn these formats, the input class is SQL_REQUEST. The SQL_REQUEST class is described below. Also see SQL_STATEMENT Class on page 138.

<object name="SQL_REQUEST" lastModified="1036435805361" id="503"> <assoc name="attribute"> <string name="name" value="STATEMENTS"/> <ref name="attributeType"value="/tibco/public/sequence/ae/class/ae/ADB/adbmetadata/sequence[SQL_STATEMENT]"/> <string name="isKey" value="false"/> <string name="isReadable" value="true"/> <string name="isWriteable" value="true"/> </assoc> <assoc name="attribute"> <string name="name" value="CLOSURE"/> <ref name="attributeType" value="/tibco/public/scalar/ae/any"/> <string name="isKey" value="false"/> <string name="isReadable" value="true"/> <string name="isWriteable" value="true"/> </assoc> <string name="family" value="ae"/> <string name="objectType" value="class"/></object>

TIBCO Rendezvous Message Request FormatThis is an example structure of the nested self-describing request message sent by an application to the adapter.

Request{rv_Name = “closure”, rvmsg_Type = RVMSG_OPAQUE, rvmsg_Data = optional closure datarv_Name = “stmt”, rvmsg_Type = RVMSG_RVMSG, rvmsg_Data = Statementrv_Name = “stmt”, rvmsg_Type = RVMSG_RVMSG, rvmsg_Data = Statementrv_Name = “stmt”, rvmsg_Type = RVMSG_RVMSG, rvmsg_Data = Statement. . .}



The closure field is an optional field. If included in the request, the response returns the same closure argument untouched, so the client application can use this information as a means of matching requests with responses.

The value Statement is a TIBCO Rendezvous Message of the following structure:Statement{rv_Name = “sql”, rvmsg_Type = RVMSG_STRING, rvmsg_Data = SQL_statement_with_possible_bind_variablesrv_Name = “maxrows”, rvmsg_Type = RVMSG_INT, rvmsg_Data = Optional_max_number_of_rows_to_fetchrv_Name = “bind”, rvmsg_Type = RVMSG_RVMSG, rvmsg_Data = Bind_datarv_Name = “bind”, rvmsg_Type = RVMSG_RVMSG, rvmsg_Data = Bind_datarv_Name = “bind”, rvmsg_Type = RVMSG_RVMSG, rvmsg_Data = Bind_data. . .}

The value Bind_data is a TIBCO Rendezvous Message with the following structure:Bind_data{rv_Name = "position", rvmsg_Type = RVMSG_INT, rvmsg_Data = position_of_placeholder_starting_with_1_from_left_to_rightrv_Name = "column", rvmsg_Type = RVMSG_STRING, rvmsg_Data = in_format_table-name.column-name_whose_column_type_matches_this_bound_variablerv_Name = "data", rvmsg_Type = type_of_bound_data, rvmsg_Data = value_of_bound_data}

If the position field is not specified, the order received is used.

Example TIBCO Rendezvous Message Requests

The section contains example requests that are based on SQL statements made by a client application and sent to an adapter instance for processing.

Simple SQL Request Statement

Given that a client application submits the following SQL statement:

INSERT INTO order_table (order_id, order_description,order_price) values (1, ‘Order 1’, 1.10)

The corresponding request could be formatted as a TIBCO Rendezvous Message similar to the following:

{name = “closure” value = optional_closure_argumentname = “stmt” value = { name = “sql” value = “INSERT INTO order_table (order_id, order_description, order_price) values (1, ‘Order 1’, 1.10)” }}



Bind Request Statement

Assume that a client application submits the following SQL statement, where ? is a placeholder for an adbDateTime value:

INSERT INTO order_table (order_id, order_date) values (2, ?)

The corresponding request could be formatted as a message in TIBCO Rendezvous Message format similar to the following:

{name = “closure” value = optional_closure_argumentname = “stmt” value = { name = “sql” value = “INSERT INTO order_table (order_id, order_date) values (2, ?)” name = “bind” value = { name = “order_table.order_date” type = RVMSG_STRING value = “1999-08-20 10:20:01” } }}

To create a bind parameter for binary data, the actual data would be sent as an RVMSG_OPAQUE.

Using Bind Entries

Bind entries are specific only to the stmt {} that contains them. For example, if you want to insert three rows into REPLYTEST, you need to have three stmt {} blocks (and thus, three sets of bind statements):

request: { stmt={ sql="INSERT INTO REPLYTEST (id, timestamp, bincol) VALUES (5, ?, ?)" bind={ column="REPLYTEST.TIMESTAMP" data="1991-10-10 12:10:10"} bind={ column="REPLYTEST.BINCOL" data=[3 opaque bytes]}} stmt={ sql="INSERT INTO REPLYTEST (id, timestamp, bincol) VALUES (5, ?, ?)" bind={ column="REPLYTEST.TIMESTAMP" data="1991-11-10 12:10:10"} bind={column="REPLYTEST.BINCOL" data=[4 opaque bytes]}} stmt={ sql="INSERT INTO REPLYTEST (id, timestamp, bincol) VALUES (5, ?, ?)" bind={ column="REPLYTEST.TIMESTAMP" data="1991-12-10 12:10:10"} bind={ column="REPLYTEST.BINCOL" data=[5 opaque bytes]}}}

Not every data value sent to an SQL statement must be bound. The client can embed all these values into the actual text of the statement instead. Embedding the values directly into the SQL text results in better performance. However, binary values must be bound. Types such as adbDateTime can either be bound or included in the SQL text by using the native DBMS date conversion function.

For example, in ORACLE, you could use the TO-DATE function to enter a date:

INSERT INTO order_table (order_id, order_date) values (2,TO_DATE(’1999-08-20 10:20:01’, ’YYYY-MM-DD HH24:MI:SS’));



TIBCO ActiveEnterprise or XML Message Response FormatIn the TIBCO ActiveEnterprise Message or XML format, the input class is SQL_BATCHRETURN. This class is described in SQL_ BATCHRETURN Class on page 138.

TIBCO Rendezvous Message Response FormatIf rv_SendWithReply() or rv_Rpc() is used to send the request, the response is sent back on the subject or Inbox that was associated with the request.

If rv_Send() is used to send the request, because no response subject was specified with the request, the response subject set when the adapter was configured is used. To receive a response from the adapter, the client application must listen on the response subject.

The structure of a response message in TIBCO Rendezvous Message format is shown below.

Reply{rv_Name = “status” rvmsg_Type = RVMSG_INT rvmsg_Data = 0rv_Name = “results” rvmsg_Type = RVMSG_RVMSG rvmsg_Data = Resultrv_Name = “closure”, rvmsg_Type = RVMSG_OPAQUE, rvmsg_Data = optional_closure_data}

where Result is a message in TIBCO Rendezvous Message format of the following structure:

Result{name = “row” type = RVMSG_RVMSG value = List_of_columnsname = “row” type = RVMSG_RVMSG value = List_of_columnsname = “row” type = RVMSG_RVMSG value = List_of_columns. . .}

where List_of_columns is a message in TIBCO Rendezvous Message format of the following structure:

List_of_columns{rv_Name = column_name, rvmsg_Type = type_of_bound_data, rvmsg_Data = value_of_bound_data



rv_Name = column_name, rvmsg_Type = type_of_bound_data, rvmsg_Data = value_of_bound_datarv_Name = column_name, rvmsg_Type = type_of_bound_data, rvmsg_Data = value_of_bound_data. . .}

If the request processing was not successful, the response could also return an error code and error description as shown next:

Reply{rv_Name = “status” rvmsg_Type = RVMSG_INT rvmsg_Data = nonzero_numberrv_Name = “sql” rvmsg_Type = RVMSG_STRING rvmsg_Data = SQL_statement_that_caused_errorrv_Name = “error” rvmsg_Type = RVMSG_STRING rvmsg_Data = error_textrv_Name = “closure”, rvmsg_Type = RVMSG_OPAQUE, rvmsg_Data = optional_closure_data}

The status value is an integer specifying success or error. Possible values are:

0: ok // No error1: noMem // Out of Memory2: notInitialized // Object never initialized3: typeConversion // Type conversion error4: dbNotFound // Database not registered5: serverError // Error reported by server6: serverMessage // Message from server7: vendorLib // Error in vendor's library8: notConnected // Lost connection9: endOfFetch // End of fetch10: invalidUsage // invalid usage of object11: columnNotFound // Column does not exist12: invalidPosition // invalid positioning within object,i.e.bounds err13: notSupported // Unsupported feature14: nullReference // Null reference parameter15: notFound // Database Object not found16: missing // Required piece of information is missing17: noMultiReaders // This object cannot support multiple readers18: noDeleter // This object cannot support deletions19: noInserter // This object cannot support insertions20: noUpdater // This object cannot support updates21: noReader // This object cannot support readers22: noIndex // This object cannot support indices23: noDrop // This object cannot be dropped24: wrongConnection // Incorrect connection was supplied25: noPrivilege // This object cannot support privileges26: noCursor // This object cannot support cursors27: cantOpen // Unable to open28: applicationError // For errors produced at the application

level29: notReady // For future use



Example TIBCO Rendezvous Message Responses

This section has two examples. The first shows a successful query response and the second shows an unsuccessful query response.

Query Example

As an example, assume the SQL statement, SELECT * FROM ORDER_TABLE produces the following result:

SQL> SELECT * FROM ORDER_TABLE;

ORDER_ID ORDER_DESCRIPTION ORDER_PRICE 1 Order 1 1.00 2 Order 2 2.00

In this case, the self-describing message in TIBCO Rendezvous Message format produced by the adapter would have the following structure:

{name = “closure” value = optional_closure_argumentname = “status” value = 0name = “results” value = { name = “row” value = { name = “order_id” type = RVMSG_INT value = 1 name = “order_description” type = RVMSG_STRING value = “Order 1” name = “order_price” type = RVMSG_REAL value = 1.00 } name = “row” value = { name = “order_id” type = RVMSG_INT value = 2 name = “order_description” type = RVMSG_STRING value = “Order 2” name = “order_price” type = RVMSG_REAL value = 2.00 } }}

Error Example

If the request that was processed resulted in an error, the adapter returns a message containing an error code and description as shown below. In this example, assume the SQL statement SELECT * FROM ORDER_TABLE produced an error because there was no table named ORDER_TABLE defined in the database:

SQL> SELECT * FROM ORDER_TABLE;SELECT * FROM ORDER_TABLE *ERROR at line 1:ORA-00942: table or view does not exist



The self-describing message produced by the adapter would have the following structure:

{name = “closure” value = optional_closure_argumentname = “status” value = 5name = “sql” value = “SELECT * FROM ORDER_TABLE”name = "error" value = "[INTERSOLV][ODBC SequeLinkdriver][DB2/400]ORDER_TABLE in *LIBL type *FILE not found.}



RPC Mode

TIBCO ActiveMatrix Adapter for Database uses the TIBCO Adapter SDK Operations API for RPCServer.

In TIBCO Designer, you can configure an adapter to act as an RPC (remote procedure call) server on behalf of a client. Selecting RPC mode creates an object in the repository describing the RPC server that the adapter instance will start.

This section presents the structure of the TIBCO Repository class objects that provide MOperation support. Based on these class descriptions, you can create an RPC client to send requests to the adapter in the expected structure.

For convenience, class structure is shown in XML format.

Schema Types

Two schema types can be used for the server object created in the repository. For Standard RPC Operation, the standard request and reply object schema are pre-defined by TIBCO ActiveMatrix Adapter for Database. This schema can be used to describe the input and output of any database request, allowing a request/response service to process database operations on any tables or execute any stored procedures.

In some cases, using a standard schema can make the request/reply service difficult to integrate with third-party applications, as standard request and reply object schema do not describe the actual input and output values of a database operation or a stored procedure. The third-party application has to do its own parsing and mapping of the request and reply objects. The Custom RPC Operation, which allows you to define your operations, can be used in these situations. You can specify the stored procedure to be executed at design time. These operation definitions are stored in the repository in a way that allows easy integration with other applications.

One-way and Two-way Invocation

An RPC operation uses two-way invocation by default, in which it expects a reply from the server. You can make an RPC operation use one-way invocation, where no reply is expected by the server.


RPC Mode | 135

Standard RPC OperationThe repository contains descriptions of the classes and operations provided for MOperation support.

The SQL_OPS class describes the operations that the adapter instance, acting as an RPC server, can handle.

The structure of the server object is:

<servers> <rvCmRpcServer name = "agentNamereqreprvcmRPCServer" session = "agentNamepubreqreprvcmRvCmSession" subject = "ADB.SDK.OPERATION" classRef = "SQL_OPS" /></servers>

Two types of operations are supported, SQL_EXECUTE and SQL_BATCHEXECUTE. SQL_EXECUTE takes a single SQL statement and processes it. SQL_BATCHEXECUTE takes a sequence of SQL statements and processes them. The other classes, such as SQL_STATEMENT, SQL_BIND, and SQL_RETURN, describe metadata for the input and output parameters to the operations.

• Use SQL_BATCHEXECUTE to send one or more SQL statements in a request.

• SQL_BATCHRETRUN is the return class.

• If an error occurs while executing a statement, the adapter returns the error immediately without executing the remaining statements.

• On Oracle only, all statements are executed within the same transaction.

SQL_OPS Class

The structure of the SQL_OPS class is:

<class name = "SQL_OPS"> <operation name = "SQL_EXECUTE" returnClass = "SQL_RETURN"> <parameter name = "STATEMENT" classRef = "SQL_STATEMENT"

direction = "in"> </parameter> </operation> <operation name = "SQL_BATCHEXECUTE" returnClass = "SQL_BATCHRETURN">



<parameter name = "STATEMENTS" classRef = "sequence[SQL_STATEMENT]"direction = "in" />

</operation> </class>

SQL_RESULTSET Class

The structure of the SQL_RESULTSET class is:

<class name = "SQL_RESULTSET"> <attribute name = "HEADER" class = "sequence[string]"> </attribute> <attribute name = "ROWVALUES" class = "sequence[SQL_ROW]"> </attribute> <attribute name = "OUTBINDS" class = "sequence[SQL_BIND]"> </attribute></class>

SQL_ROW Class

The structure of the SQL_ROW class is:

<class name = "SQL_ROW"> <attribute name = "ROW" class = "sequence[any]"> </attribute> </class>

SQL_RETURN Class

The structure of the SQL_RETURN class is:

<object name="SQL_RETURN" lastModified="1046487158293" id="202"><assoc name="attribute">

<string name="name" value="STATUS"/><ref name="attributeType" value="/tibco/public/scalar/ae/string"/><string name="isKey" value="false"/><string name="isReadable" value="true"/><string name="isWriteable" value="true"/>

</assoc><assoc name="attribute">

<string name="name" value="SQL"/><ref name="attributeType" value="/tibco/public/scalar/ae/string"/><string name="isKey" value="false"/><string name="isReadable" value="true"/>


RPC Mode | 137

<string name="isWriteable" value="true"/></assoc><assoc name="attribute">

<string name="name" value="ERROR_DESC"/><ref name="attributeType" value="/tibco/public/scalar/ae/string"/><string name="isKey" value="false"/><string name="isReadable" value="true"/><string name="isWriteable" value="true"/>


<string name="name" value="CLOSURE"/><ref name="attributeType" value="/tibco/public/scalar/ae/any"/><string name="isKey" value="false"/><string name="isReadable" value="true"/><string name="isWriteable" value="true"/>


<string name="name" value="RETURNVALUE"/><ref name="attributeType"

value="/tibco/public/class/ae/ADB/adbmetadata/SQL_RETURNVALUE"/><string name="isKey" value="false"/><string name="isReadable" value="true"/><string name="isWriteable" value="true"/>

</assoc><string name="family" value="ae"/><string name="objectType" value="class"/>

</object>

SQL_RETURNVALUE

<object name="SQL_RETURNVALUE" lastModified="1046487158293" id="204"><assoc name="attribute">

<string name="name" value="OUTBINDS"/><ref name="attributeType"

value="/tibco/public/sequence/ae/class/ae/ADB/adbmetadata/sequence[SQL_BIND]"/><string name="isKey" value="false"/><string name="isReadable" value="true"/><string name="isWriteable" value="true"/>


<string name="name" value="RESULTSETS"/><ref name="attributeType"

value="/tibco/public/sequence/ae/class/ae/ADB/adbmetadata/sequence[SQL_RESULTSET]"/>

<string name="isKey" value="false"/><string name="isReadable" value="true"/><string name="isWriteable" value="true"/>

</assoc><string name="family" value="ae"/><string name="objectType" value="class"/>

</object>



SQL_ BATCHRETURN Class

The structure of the SQL_BATCHRETURN class is:

<class name = "SQL_BATCHRETURN"> <attribute name = "STATUS" class = "string"> </attribute> <attribute name = "RESULTSETS" class = "sequence[SQL_RESULTSET]"> </attribute> <attribute name = "SQL" class = "string"> </attribute> <attribute name = "ERROR_DESC" class = "string"> </attribute> <attribute name = "CLOSURE" class = "any" </attribute></class>

SQL_BIND Class

The structure of the SQL_BIND class is:

<class name = "SQL_BIND"> <attribute name = "POSITION" class = "i4"> </attribute> <attribute name = "TYPE" class = "string"> </attribute> <attribute name = "DATA" class = "any"> </attribute> <attribute name = "NAME" class = "string"> </attribute></class>

SQL_STATEMENT Class

The structure of the SQL_STATEMENT class is:

<class name = "SQL_STATEMENT"> <attribute name = "SQL_STRING" class = "string"> </attribute> <attribute name = "BINDS" class = "sequence[SQL_BIND]"> </attribute> <attribute name = "CLOSURE" class = "any" </attribute> <attribute name = "MAXROWS" class = "i4"> </attribute></class>


RPC Mode | 139

Custom RPC OperationWhen you select Use Custom Operations in the Configuration tab of the Request-Response dialog, you can create custom schema for your operations. You can define the call operations to be executed by the RPC server in advance. The palette automatically generates the input and output schema for the call operation.

The operation takes one input class, REQUEST, which describes the input parameters and other input options for the stored procedure execution. The operation returns a class called REPLY, which describe the output schema of the stored procedure or any error message the agent returns.

Table 6 REQUEST Schema Description

Input Item

Data Type or Classname Description

INBINDS INPUT_BINDS The input parameters of the stored procedure.

OPTIONS INPUT_OPTIONS The input options:

Option Data Type Description

MAXROWS Integer The maximum number of rows to retrieve.

SQL String The SQL string use to execute the stored procedure. This string is automatically generated by the palette.

CACHE Boolean True if user want the agent to cache the statement for performance optimization

PACKAGE String read only, uses the call operation form to modify the package of the stored procedure.

SCHEMA String read only, uses the call operation form to modify the package of the stored procedure.

CLOSURE Any Closure argument. The reply returns this closure argument untouched.



Table 7 REPLY Schema Description

Output Item Data Type or Classname Description

OUTBINDS OUTPUT_BINDS The output parameters of the stored procedure

RESULTSET[1..n]/RESULTSETS

OUTPUT_ROWS/ SQL_RESULTSET

The result set(s) returned by the stored procedure, see section 'Result Set Support' for more description

STATUS String "SUCCESS" if the stored procedure is executed successfully. "FAILURE" if there is an error. Error details are stored in the OPTIONS class

OPTIONS CUSTOM_OP_OUTPUT_OPTION

Contains the error description and the SQL statement if an error occurred.

CLOSURE Any Closure argument obtained from the request


RPC Mode | 141

Resultset Support

TIBCO ActiveMatrix Adapter for Database supports multiple resultsets returned for a stored procedure. The class generated for the resultset output depends on the schema information provided by the JDBC driver.

• For drivers that return valid information of a resultset schema, ADB creates an output class, named OUTPUT_ROW[1..n], for each result set. This is shown in the screen below.

• If a driver does not return valid information for the resultset schema, TIBCO ActiveMatrix Adapter for Database uses the generic class SQL_RESULTSETS as the output schema, as described in Standard RPC Operation on page 135.

DataDirect JDBC 3.3 drivers only provide valid resultset schema information for Oracle Server version 8.1.7 or higher.



Implementing Client RPC ProgramsTwo sample clients are provided in the install-path\demo\operation directory:

• a sample C++ client program that you can build and change

• a sample RPC client

Creating a Custom RPC Client

There are two tasks in creating a custom RPC client:

1. Write an RPC client program using the TIBCO Adapter SDK. This task is described in the TIBCO Adapter SDK Programmer’s Guide.

2. Create a new entry for the RPC client in the repository. This task is described below.

For more information on RPC server and client endpoint type fields, see TIBCO Designer Palette Reference.

To create a new RPC client entry in the repository:

Perform these steps after writing the client program and configuring the adapter, as described above. Detailed instructions for configuring a generic adapter are in the TIBCO Designer User? Guide.

1. In TIBCO Designer, drag a Generic Adapter Configuration icon to the design panel.

2. Enter an Instance Name and an SDK AppName for the adapter.


RPC Mode | 143

3. Click Apply.

4. In the project tree panel, expand the GenericAdapterConfiguration icon and click the Adapter Services folder.

5. Drag a Request-Response Invocation Service icon to the designer panel.

6. In the Configuration tab, set the name of the service and transport type of your client service.

7. In the Transport tab, set the message subject, quality of service, and wire format of your client service.

If using the TIBCO Rendezvous message transport, you must use the TIBCO ActiveEnterprise wire format.

8. In the Schema tab, set the classReference to point to the correct class schema.

— For a standard RPC schema, use the following class schema:

AESchemas/ae/ADB/adbmetadata/Classes/SQL_OPS

— For custom RPC schema, use the class schema that corresponds to the server endpoint:

AESchemas/ae/ADB/serviceName/Classes/ADBServer_OPS

9. Click Apply.



Load Balancing

The adapter balances the load of requests coming in from applications in two ways, within a single adapter instance and across multiple adapter instances.

Load Balancing in a Single Instance Using ThreadsAn adapter instance performs load balancing within itself by allowing you to specify the number of threads that will be responsible for processing application requests. The number of threads to be spawned is specified in TIBCO Designer for an adapter instance.

To set load balancing parameters for a configuration:

1. In the project tree panel, click the corresponding ActiveDatabase Adapter resource to select it.

The TIBCO Designer window should look similar to the following:

2. Click Show All Tabs.


Load Balancing | 145

3. Display the Adapter Services tab.

4. Change the value of Number of Request-Response Service Threads as needed.

5. Click Apply.

Each request-response service thread is dedicated to listening on an agreed request subject.

1. When a request message is received by the request-response communication thread, it enqueues it into an in-memory queue. Each database request-response thread has a connection to the database using the user name-password combination that was provided to the adapter.

2. When an item is enqueued by the communication thread, one of the database request-response threads picks it up and executes the one or more SQL statements in the message as an atomic transaction.



3. It then composes a TIBCO Rendezvous self-describing response message and sends back a response on a response subject. If the original request was sending using rv_SendWithReply() or rv_Rpc(), the response is sent back on the subject that was specified by the requesting application. Otherwise, the response is sent back on the subject that was set up at configuration time.

Load Balancing Across Adapter InstancesTo achieve load balancing across adapter instances, TIBCO Rendezvous Distributed Queueing can be used to assign each application request to exactly one adapter instance. For example, the following diagram shows three adapter instances each connected to a database server (not shown) that contains replicated data.

Figure 6 Request-Response Load Balancing

The adapter instances have been set up to use distributed queueing with one instance acting as the scheduler. Only one of the three adapter instances will receive an incoming request from an application. The scheduler will assign certified requests to the least loaded member of the queue.

While the request message can be sent with either reliable or certified quality of service, the response message is always sent back using the reliable quality of service.

TIBCO Messaging

Application

Adapter 4

Adapter 3

Adapter 2

Adapter 1(Scheduler)

Request

Request

Request

RequestResponses

Distributed Queue



A queue member’s load is determined by the number of pending processes that are waiting to be processed by the member’s database request-response thread. After the request is processed, the processing configuration sends a response back to the requesting application.

Configuring an Adapter Instance to use Distributed Queues

You can configure the adapter with a subscription service to use a distributed queue, then use TIBCO Administrator Enterprise Edition to deploy multiple instances of the adapter. The instances can be deployed on the same machine, or multiple machines. In this scenario, each adapter instance uses the same (default) values for the distributed queue.

1. Start TIBCO Designer and create a new project.

2. Drag the ActiveDatabase Adapter Configuration icon to the design panel.

3. Configure the adapter instance with a design-time connection and ODBC DSN value.

4. Drag the ADBSubscriber icon to the design panel.

Reliable requests in a certified environment use a simpler type of load balancing. In this example, the scheduler distributes the first reliable request to Adapter 2, the second to Adapter 3, and the third to Adapter 4 without evaluating a queue member’s load.



5. Under the Configuration tab, set Transport Type to Rendezvous and Quality of Service to Distributed Queue. For example:

6. Configure the service with a destination table name and other settings for your environment.

7. Click Project>Save.

8. Click Tools>Create Project EAR. Under Configuration, provide a name and file location for the EAR file.

9. Click Build Archive.

After creating the EAR file, you are ready to import it into TIBCO Administrator Enterprise Edition and add the adapter instance multiple times to the same machine or multiple machines.

1. Start TIBCO Administrator Enterprise Edition.

2. Select Application Management, then click the New Folder button. In Name, provide a name for the folder and click Save.

3. Select the folder, then click New Application.

4. Click Browse and navigate to the EAR file you created in Designer. Click OK.

5. In the dialog that appears, click Save.



6. Expand the configuration, then double-click the <name>.aar file.

7. Click the Add to Additional Machine button and select the machine to bind to and click OK.



8. Repeat to create the number of adapter instances required for your environment. For example, the next screen shows that three adapter instances have been added.

9. Click Save.

The adapter instances are ready to be deployed. After the instances are deployed, when a message is sent to the adapter, the instance that is not busy will handle the request.


Oracle REF Data Type Support | 151

Oracle REF Data Type Support

The adapter supports the use of the REF data type (cursor) as an OUT parameter only in an Oracle stored procedure. The adapter returns a result in the same way as it does a result set from a SELECT query. The exact usage depends on the driver used (DataDirect Connect or Oracle ODBC), as described below.

For example, consider the following PL/SQL procedure:

-- Create a REF cursor procedureCREATE OR REPLACE PACKAGE ORDER_DEMO AS TYPE OrderCurTyp IS REF CURSOR RETURN SUB_ORDER%ROWTYPE; PROCEDURE FindOrders (price IN NUMBER, order_cv OUT OrderCurTyp);END ORDER_DEMO;

CREATE OR REPLACE PACKAGE BODY ORDER_DEMO AS PROCEDURE FindOrders (price IN NUMBER, order_cv OUT OrderCurTyp) IS BEGIN OPEN order_cv FOR SELECT * FROM SUB_ORDER where ORDER_PRICE > price; END FindOrders;END ORDER_DEMO;

Using DataDirect Connect Drivers

When using DataDirect Connect® (Merant) drivers, the procedure is called by leaving the REF cursor parameter out of the procedure's call list. For example, the procedure ORDER_DEMO.FindOrders with an input price of $1.00 can be called through the request program with the following request:

{stmt={sql="{call ORDER_DEMO.FindOrders(?)}" bind={position=1 type="IN" data=1}}}

The REF cursor's data will be returned in the results section of the reply.

Also, ProcedureRetResults=1 must be added to the data source entry in the odbc.ini file.

Using Oracle ODBC Drivers

When using the Oracle ODBC Driver, the procedure is called by including the REF cursor parameters in the call list but not binding them. For example, the procedure ORDER_DEMO.FindOrders with an input price of $1.00 can be called through the request program with the following request:

{stmt={sql="{call ORDER_DEMO.FindOrders(?, ?)}" bind={position=1 type="IN" data=1}}}

The REF cursor's data will be returned in the results section of the reply.




| 153

Chapter 8 Advanced Features

This chapter describes advanced features such as the format of a publishing table, stored procedures and the triggers used to populate publishing tables and adapter instance options.

Topics

• Publishing Table, page 154

• Source Table, page 160

• Exception Table, page 161

• Opaque Exception Table, page 164

• Incremental Parent-child Operation, page 166

• Database Types, page 168

• Subject Names, page 171

• Preregistering a Certified Subscriber, page 173

• Changing the Location of the Ledger File, page 175

• Sending Messages to an Adapter Instance, page 177

• Publishing by Reference Object, page 179

• User Callout Library, page 187

• Publishing in a Commitment Controlled Environment, page 194

• Using the Adapter with a Revision Control System, page 196

• Using a Log File for an Adapter Instance, page 198

• Using Database Cleanup Scripts, page 203

• Change the SQL Statement Terminator (DB2 for z/OS Only), page 205

• Fault Tolerance, page 206

• SSL Data Encryption using DataDirect ODBC Drivers, page 207


154 | Chapter 8 Advanced Features

Publishing Table

Publishing tables mirror tables that you have identified for monitoring. They contain additional columns, primarily a sequence number and delivery status, which are needed by the adapter to detect new rows. You create a publishing table for each table you want to activate using TIBCO Designer.

The trigger generated at designtime will automatically populate the values for these additional columns. It is not reccomended that you modify these values.

In addition to a copy of the source table’s columns, the publishing table has the following additional columns.

Table 8 Publishing Table Additional Columns

Column Name Type Description

ADB_SUBJECT VARCHAR2(255)

Used to specify the subject to publish the current row. Length is 255.

ADB_SEQUENCE NUMBER(38) Stores the monotonically increasing sequence number that represents new rows in the publishing table. If a column with this name exists, the number is generated automatically.

By default, the schema type is string. The adapter treats this number as a string. This number can be larger than an integer if the database supports it.

This number should be greater than 0 (zero).

ADB_SET_SEQUENCE NUMBER(38) Currently not used.


Publishing Table | 155

ADB_TIMESTAMP DATE Time of row insertion in publishing table that is used to calculate expiration of rows. The timestamp is generated automatically.

For Oracle databases, the timestamp includes the time zone information.

For all new services:

— Microsoft Windows platforms, the timestamp with time zone option should be selected in the Advanced Settings tab of the ODBC connection.

— Unix platforms, the LocalTimeZoneOffset parameter in the odbc.ini should be set to the local time zone. For example, LocalTimeZoneOffset=+10:00.

Also the EnableTimestampWithTimeZone parameter in the odbc.ini file should be set to 1.

ADB_TRACKINGID VARCHAR2(40)

Tracking ID of the message.

This is automatically added to the publishing table and the publishing schema.

If you do not want to monitor the tracking id, you can manually remove this field from the project schema and the publishing table

ADB_OPCODE NUMBER(38) Operation code used by an adapter instance:

1 indicates INSERT.

2 indicates UPDATE.

3 indicates DELETE.

4 indicates UPSERT. UPDATE if row exists, otherwise INSERT.

10 indicates BYPASS (See Incremental Parent-child Operation on page 166)

If an incoming TIBCO Rendezvous message does not have an operation code, an INSERT occurs.

Table 8 Publishing Table Additional Columns (Cont’d)

Column Name (Cont’d) Type Description (Cont’d)



The publication table cannot contain any user-created columns where the column name starts with ADB_. These characters are reserved for adapter use.

Removing Records from the Publication TableAn adapter instance does not remove the records from the publication table automatically, because often that information must be retained for auditing reasons. You can use the following command at the SQL prompt to manually remove records from the given publishing table. The command deletes all rows where ADB_L_DELIVERY_STATUS is C or F.

SQL> delete from publication_table_name where ADB_L_DELIVERY_STATUS = "C" or ADB_L_DELIVERY_STATUS = "F";

To automate the process, you could include this statement in a trigger.

Publication ExampleThe following is a listing of the SQL commands used to create a publication. The listing can be found in the demo\demo1\demo_database.sql file. The file shows the stored procedure, and insert, update and delete triggers that are generated for a publication.

-- *******************************************-- TIBCO ActiveMatrix Adapter for Database

ADB_UPDATE_ALL NUMBER(38) Currently not used.

ADB_REF_OBJECT VARCHAR2(64)

When publish by reference object is used, contains the name of the reference object that provides source data.

ADB_L_DELIVERY_STATUS CHAR(1) Delivery status of a TIBCO Rendezvous message:

P indicates pending acknowledgement.

N indicates that a new message has arrived, but has not yet been published.

C indicates complete.

F indicates failed.

ADB_L_CMSEQUENCE NUMBER(38) Certified messaging sequence number associated with this message.

Table 8 Publishing Table Additional Columns (Cont’d)

Column Name (Cont’d) Type Description (Cont’d)



-- Demo SQL script -- *******************************************

-- Spool the output of sql commands to a log file spool demo.log

-- *******************************************-- Create sample tables-- *******************************************

-- Publication source tableCREATE TABLE ORDER_TABLE ( ORDER_ID NUMBER PRIMARY KEY, ORDER_DESCRIPTION VARCHAR2(128), ORDER_PRICE NUMBER(10,3));

-- Create the destination tableCREATE TABLE SUB_ORDER ( ORDER_ID NUMBER PRIMARY KEY, ORDER_DESCRIPTION VARCHAR2(128), ORDER_PRICE NUMBER(10,3));

-- *******************************************-- Setup a publication -- *******************************************

CREATE TABLE PUB_ORDER (ORDER_ID NUMBER,ORDER_DESCRIPTION VARCHAR2(128),ORDER_PRICE NUMBER(10,3),ADB_SUBJECT VARCHAR2(255) NULL,ADB_SEQUENCE INTEGER NOT NULL,ADB_SET_SEQUENCE INTEGER NULL,ADB_TIMESTAMP DATE NULL,ADB_OPCODE INTEGER NOT NULL,ADB_UPDATE_ALL INTEGER NULL,ADB_REF_OBJECT VARCHAR2(64) NULL,ADB_L_DELIVERY_STATUS CHAR NULL,ADB_L_CMSEQUENCE NUMBER(38, 0) NULL

)/

CREATE UNIQUE INDEX IDX1_PUB_ORDER ON PUB_ORDER (ADB_SEQUENCE)/

CREATE UNIQUE INDEX IDX2_PUB_ORDER ON PUB_ORDER (ADB_L_DELIVERY_STATUS, ADB_SEQUENCE)/

CREATE INDEX IDX3_PUB_ORDER ON PUB_ORDER (ADB_L_CMSEQUENCE)/

CREATE SEQUENCE PUB_ORDER_SEQSTART WITH 1



INCREMENT BY 1NOMAXVALUENOCYCLECACHE 10

/

CREATE OR REPLACE TRIGGER TRI_PUB_ORDER AFTER INSERT OR DELETE OR UPDATE ON

ORDER_TABLE FOR EACH ROW DECLAREupdating_key_fields EXCEPTION;

BEGINIF INSERTING THENINSERT INTO PUB_ORDER VALUES (

:NEW.ORDER_ID,:NEW.ORDER_DESCRIPTION,:NEW.ORDER_PRICE,NULL,PUB_ORDER_SEQ.NEXTVAL,0,SYSDATE,1,NULL,NULL,'N',-1);

END IF;IF UPDATING THEN

IF UPDATING('ORDER_ID') THENRAISE updating_key_fields;

END IF;INSERT INTO PUB_ORDER VALUES (

:OLD.ORDER_ID,:NEW.ORDER_DESCRIPTION,:NEW.ORDER_PRICE,

NULL,PUB_ORDER_SEQ.NEXTVAL,

0,SYSDATE,

2,NULL,NULL,'N',-1);

END IF;IF DELETING THENINSERT INTO PUB_ORDER VALUES (

:OLD.ORDER_ID,:OLD.ORDER_DESCRIPTION,:OLD.ORDER_PRICE,NULL,PUB_ORDER_SEQ.NEXTVAL,0,SYSDATE,3,NULL,



NULL,'N',-1);

END IF;EXCEPTION

WHEN updating_key_fields THENraise_application_error(-20300, 'ActiveDB Error: cannot update key fields of source table.');

END TRI_PUB_ORDER;/

-- *******************************************-- Setup a subscription-- *******************************************

CREATE TABLE SUB_ORDER_EXCEP (ORDER_ID NUMBER,ORDER_DESCRIPTION VARCHAR2(128),ORDER_PRICE NUMBER(10,3),ADB_OPCODE INTEGER NULL,ADB_UPDATE_ALL INTEGER NULL,ADB_ERROR_TEXT VARCHAR(4000) NULL,ADB_ERROR_TIME DATE DEFAULT SYSDATE

)/

COMMIT;

Changing the Publication Trigger to Publish a Subset of RowsYou can set up a publication to publish only a subset of the rows within a table. That is, there may be scenarios where it is desirable to publish only if the inserted, deleted, or updated row satisfies certain conditions. While it is possible to use the callout library to do this additional filtering, it is simpler and more efficient to use a trigger that directly tests whether publishing conditions are met.

For Oracle, add a when clause to a row level trigger to test for the desired conditions. For Sybase and Microsoft SQL Server, use the if statement within a trigger body to test for the desired conditions.

For example, you could change the demo.sql trigger described in the previous section to fire only if the ORDER_PRICE value is $1.00 or greater:

CREATE OR REPLACE TRIGGER TRI_P_ORDER_TABLE AFTER INSERT OR UPDATE OR DELETE ON ORDER_TABLEFOR EACH ROW WHEN (new.ORDER_PRICE >= 1.00)



Source Table

If loop detection is enabled, the following column will be added to the source table. Loop detection is explained in Loop Detection on page 11.

If the source table contains a field ADB_TRACKINGID, it will be updated using the tracking id of the message. If this field is not present, the tracking id field will be ignored..

The source table cannot contain any user-created columns where the column name starts with ADB_. These characters are reserved for adapter use.


ADB_SOURCE CHAR Used for loop detection. Denotes whether the row was inserted or updated as the result of a TIBCO Rendezvous message, rather than user intervention. See Loop Detection on page 11 for more information.

Valid values are T or NULL.

• T indicates the row is not to be published.

• NULL indicates the row can be published.

ADB_TRACKINGID VARCHAR2(40)

Tracking ID of the message.


Exception Table | 161

Exception Table

If a database restriction or failure occurs, an exception table can be configured to receive the message. If insertion into an exception table also fails, an error message will display and the adapter instance will terminate. You can build a TIBCO Hawk rulebase that detects when the configuration is down and automatically restarts it when the database is up. See the TIBCO Hawk Administrator? Guide for details about creating a rulebase.

For bulk insert operations, the agent will insert the unsuccessful rows in the exception table only when an exception occurs. For example, if the agent attempts to perform a bulk insert of 500 rows to the destination table, and the first 300 rows are inserted successfully but the last 200 rows are unsuccessful, the agent will insert only the last 200 rows to the exception table. It will then commit the operation. (The bulk insert operation is described in Adapter Services Tab on page 53.)

In addition to destination table columns, the following columns are added to the exception table.


ADB_OPCODE NUMBER(38) Operation code used by the adapter:

1 indicates INSERT

2 indicates UPDATE

3 indicates DELETE

4 indicates UPDATE if row exists, otherwise INSERT.

If an incoming TIBCO Rendezvous message does not have an operation code, an INSERT occurs.

ADB_UPDATE_ALL NUMBER(38) Currently not used.

ADB_TRACKINGID VARCHAR2(40) Tracking ID of the message. This column is the primary key. Each exception table that is mapped to a child table is connected to the parent’s exception table by this column

ADB_ERROR_TEXT VARCHAR2(4000) Text of the error from the database server, ODBC driver, or other source that caused the exception.



The exception table cannot contain any user-created columns where the column name starts with ADB_. These characters are reserved for adapter use.

Child Exception TableIn addition to child table columns, the following columns are added to a child exception table.

ADB_ERROR_TIME DATE Timestamp of the inserted record.




— Unix platforms, the LocalTimeZoneOffset parameter in the odbc.ini should be set to the local time zone. For example, LocalTimeZoneOffset=+10:00


ADB_JOIN_ID VARCHAR Joined column used to link a parent record with its child record. ADB_JOIN_ID is generated from ADB_TRACKINGID and concatenated with the record number in the group. A child table’s exception table is connected to its parent’s table exception table by ADB_JOIN_ID.



ADB_ERROR_TEXT VARCHAR Text of the error from the database server, ODBC driver, or other source that caused the exception.

ADB_ERROR_VTIME DATE Timestamp of the inserted record.

ADB_TRACKINGID VARCHAR TrackingID of the message


Exception Table | 163

Using an Exception Table as a Source TableIf you want to publish from an exception table and also want to use that exception table as the source table, do not use the ADB_ERROR_TEXT or ADB_OPCODE column names. Instead, follow these guidelines:

• Create a database view that mirrors the exception table, but rename the ADB_ERROR_TEXT and ADB_OPCODE columns so that they do not begin with "ADB_."

• After renaming the columns, use Publish By Reference Object (see Publishing by Reference Object, page 179) and choose your view as the reference object.

ADB_JOIN_ID VARCHAR Tracking ID of the Message




Opaque Exception Table

The subscription service uses two logical layers when processing a message. The first layer decodes data from the message and the second layer provides the database transaction. If an exception occurs in the first layer, the adapter logs the message to the opaque exception table. In the second layer, if any DML command fails at any level, the adapter rolls back this transaction and starts another transaction, inserting into exception tables. If the insert into exception table transaction fails, the adapter then logs the message to the opaque exception table.

The opaque exception table records the entire message into a column along with the error message. The opaque exception table has the following columns:


ADB_TRACKINGID VARCHAR Tracking ID of the message.

ADB_ERROR_TEXT VARCHAR Text of the error from the database server, ODBC driver, Adapter SDK or other source that caused the exception.

ADB_ERROR_TIME DATE Timestamp of the inserted record.




— Unix platforms, the LocalTimeZoneOffset parameter in the odbc.ini should be set to the local time zone. For example, LocalTimeZoneOffset=+10:00


ADB_MSG BLOB Raw bytes of the message.


Opaque Exception Table | 165

The ADB_SUBTAB column allows you to configure several subscription services using only one opaque exception table in the same database schema.

ADB_SUBTAB VARCHAR Destination table name.

ADB_SUBJECT VARCHAR Subscription service destination or subject.

ADB_TRANSPORT INT Subscription service transport type:

• 0 unknown

• 1 Rendezvous

• 2 JMS




Incremental Parent-child Operation

To support incremental parent-child operations, each child row has an opcode, that is, an extra ADB_OPCODE field that is added to the child schema. The opcode ADB_OPCODE_BYPASS is used to bypass the current table operation. The adapter determines if the operation is an incremental parent-child operation by checking the first level child opcode. If the first level child opcode is not set, the adapter treats it as a complete operation. For the subsequence child level, if the child opcode is not set, it will inherit the parent opcode.

Mixed parent-child operations are also supported. You can send a message to insert new child rows, update other child rows, and delete other child rows for an existing parent-child object. Following is an example of a mixed parent-child operation:

adb.key{ RVMSG_INT 2 ^type^ 1 RVMSG_INT 2 ^pfmt^ 10 RVMSG_INT 2 ^ver^ 30 RVMSG_INT 2 êncoding^ 1 RVMSG_RVMSG 110 ^prefixList^ { RVMSG_STRING 49 1 "/tibco/public/sequence/ae/class/ae/ADB/abc" RVMSG_STRING 37 default "/tibco/public/class/ae/ADB/abc" } RVMSG_RVMSG 77 ^tracking^ { RVMSG_STRING 30 îd^ "Gi2--4--DGMSk--s-064zzw8L-zzw" RVMSG_STRING 22 ^1^ "adb.key" } RVMSG_RVMSG 1200 ^data^ { RVMSG_STRING 8 ^class^ "S_KEYP1" RVMSG_INT 4 ADB_OPCODE 10 RVMSG_RVMSG 480 ADB_SEQUENCE_S_KEYP2 { RVMSG_STRING 18 ^class^ "sequence[S_KEYP2]" RVMSG_INT 4 îdx^ 1 RVMSG_RVMSG 210 ^1^ { RVMSG_STRING 8 ^class^ "S_KEYP2" RVMSG_INT 4 ADB_OPCODE 10 RVMSG_RVMSG 119 ADB_SEQUENCE_S_KEYP3 { RVMSG_STRING 18 ^class^ "sequence[S_KEYP3]" RVMSG_INT 4 îdx^ 1 RVMSG_RVMSG 58 ^1^ { RVMSG_STRING 8 ^class^ "S_KEYP3"


Incremental Parent-child Operation | 167

RVMSG_INT 4 ADB_OPCODE 4 RVMSG_REAL 8 A 1 RVMSG_STRING 2 B "a" RVMSG_REAL 8 C 4 } RVMSG_RVMSG 58 ^2^ { RVMSG_STRING 8 ^class^ "S_KEYP3" RVMSG_INT 4 ADB_OPCODE 3 RVMSG_REAL 8 A 1 RVMSG_STRING 2 B "a" RVMSG_REAL 8 C 4 } } } RVMSG_RVMSG 210 ^2^ { RVMSG_STRING 8 ^class^ "S_KEYP2" RVMSG_INT 4 ADB_OPCODE 1 RVMSG_REAL 8 A 2 RVMSG_STRING 2 B "a" RVMSG_REAL 8 C 4 RVMSG_RVMSG 119 ADB_SEQUENCE_S_KEYP3 { RVMSG_STRING 18 ^class^ "sequence[S_KEYP3]" RVMSG_INT 4 îdx^ 1 RVMSG_RVMSG 58 ^1^ { RVMSG_STRING 8 ^class^ "S_KEYP3" RVMSG_REAL 8 A 2 RVMSG_STRING 2 B "a" RVMSG_REAL 8 C 4 } } } } }}

When the adapter receives this message, it performs the following database operations in sequence:

1. Bypass the parent table operation

2. Bypass first row operation on child table S_KEYP2

3. UPDATE S_KEYP3 set B = 'a', C = 4 where A = 1;

4. DELETE FROM S_KEYP3 where A = 1;

5. INSERT INTO S_KEYP2 (A, B, C) values (2, 'a', 4);

6. INSERT INTO S_KEYP3 (A, B, C) values (2, 'a', 4);



Database Types

The database data types listed in this section are supported.

OracleThe following Oracle data types are supported:

The following ANSI data types in Oracle are supported:

SQL ServerThe following SQL Server data types are supported:

• CHAR • TIMESTAMP

• DATE • RAW

• LONG • VARCHAR2

• LONG RAW • CLOB

• NUMBER • BLOB

• DOUBLE PRECISION • INTEGER

• FLOAT • SMALLINT

• INT • REAL

• BINARY

• BIT

• CHAR

• DATETIME

• DECIMAL

• FLOAT

• IMAGE (publish by reference only)

• INT

• MONEY

• NTEXT

• NUMERIC

• NVARCHAR

• REAL

• SMALLDATETIME

• SMALLINT

• SMALLMONEY

• TEXT (publish by reference only)

• TINYINT

• UNIQUEIDENTIFIER

• VARBINARY

• VARCHAR


Database Types | 169

Sybase and Sybase Adaptive Server AnywhereThe following Sybase data types are supported:

DB2The following DB2 data types are supported:

• BINARY • INT • TEXT

• BIT • MONEY • TINYINT

• CHAR • NUMERIC • VARCHAR

• DATETIME • REAL • VARBINARY

• DECIMAL • SMALLDATETIME

• FLOAT • SMALLINT

• IMAGE Publish by reference only)

• SMALLMONEY

• CHARACTER • NUMERIC

• DECIMAL • REAL

• DOUBLE PRECISION • SMALLINT

• FLOAT • TIMESTAMP

• INTEGER • VARCHAR

• LONG VARGRAPHIC



Mapping to Database TypesThe adapter converts values fetched from the database to types supported by the wire format. Similarly, data fields received in a message must be converted to meaningful database types before being stored.

The following table shows the mapping used by the adapter for such conversion.

Table 9 Mapping to Database Type

ODBC Type TIBCO ActiveEnterprise Type

TIBCO Rendezvous Message Type

SQL_CHAR

SQL_VARCHAR

SQL_LONGVARCHAR

string RVMSG_STRING

SQL_SMALLINT

SQL_INTEGER

i4 RVMSG_INT

SQL_REAL

SQL_DOUBLE

SQL_FLOAT

SQL_DECIMAL

SQL_NUMERIC

r8 RVMSG_REAL

SQL_BINARY

SQL_VARBINARY

SQL_LONGVARBINARY

binary RVMSG_OPAQUE

SQL_TIMESTAMP dateTime RVMSG_DATETIME


Subject Names | 171

Subject Names

An adapter instance sends and receives messages using TIBCO Rendezvous subject names. Subject names are structured strings of alphanumeric characters. Each subject name is a string of characters divided into elements by the dot (.) character. For example, Tibco.Tsi and sales.accounting.order are valid subject names.

Default subject names are provided in TIBCO Designer. You can change these default values to specify a custom subject name. A default subject is created each time an adapter service is created.

The following are restrictions on subject names:

• Subject names are limited to a total length of 255 characters (including the dot delimiters). Even though TIBCO Rendezvous has a 255-character subject length limit, it is very unlikely that applications will need to use the full length. If you want to run an adapter instance against Microsoft SQL Server, you must design your subject space to conform to the length restrictions.

• There can be at most 100 elements in a subject name. The maximum element length is 255 characters.

• The asterisk (*) and greater than (>) are wildcard characters that can be used by subscriber configurations.

For example, if messages are published on the tsi.sales.sports and tsi.sales.clothing subjects and your subscriber adapter listens on the tsi.sales.* subject, the configuration will get messages sent on both subjects.

• The underscore (_) character is reserved. An underscore and any other unprintable character should not be used in your subject name elements. It is also recommended that you do not use tabs or spaces in subject names.

• It is illegal to incorporate the dot character into an element by using an escape sequence.

• Subject names are case sensitive.

See the TIBCO Rendezvous Concepts manual for a full explanation of subject names.



Parameterized Subject NamesA TIBCO Rendezvous subject name can be created from one or more columns in a publishing table. A subject created this way is known as a parameterized subject. For example, you can use TIBCO Designer to define a publication subject to MYSUBJECT.$COLUMN1.$COLUMN2

where COLUMN1 and COLUMN2 are names of columns in the publishing table. For example:TIBCO.ORDER.$ORDER_ID.$ORDER_DESCRIPTION

On publication, a subject is created from the contents of those columns. This allows receiving applications to filter publications based on value of certain fields. Parameterized subject names are supported with certified or reliable delivery.

Parameterized Subject Name Restrictions

• A TIBCO Rendezvous parameterized subject name can be at most 255 characters in length, with each component at most 127 characters. In the previous example, if the ORDER_DESCRIPTION column contains a description value that is longer than 127 characters, the parameter name $ORDER_DESCRIPTION is used as the subject in place of the description value.

• The asterisk (*) or greater than (>) characters or empty strings should not be used as attribute values (varchar, char) in a column that is part of a parameterized subject. Doing so makes the subject invalid.

For example, if the TIBCO.ORDER.$ORDER_ID.$ORDER_DESCRIPTION parameterized subject name is defined and the following values are inserted into ORDER_TABLE:

insert into ORDER_TABLE values(111,'*',88.81);

and the following values are inserted into the ORDER_DESCRIPTION table:insert into ORDER_TABLE values(112,'>',99.91);

the publisher would return an error.

• Do not use columns that contain Date types as part of a parameterized subject. Date values are likely to contain characters such as dashes and spaces that can cause problems when used as part of a subject.

• Do not use columns that contain binary types as part of a parameterized subject.

• Do not insert data of Float type as an attribute value in a column that is part of a parameterized subject. Doing so makes the subject invalid.

• If a column should not be published, do not include the column name as part of the parameterized subject name.


Preregistering a Certified Subscriber | 173

Preregistering a Certified Subscriber

You can add one or more certified subscribers, or listeners, for a publisher adapter that sends messages on subjects using the RVCM quality of service. The name of the subscriber is added to the publisher adapter’s preregistered list. The subscriber can be a TIBCO ActiveMatrix Adapter for Database subscriber, or any other TIBCO Rendezvous subscriber.

A subscriber adapter on the preregister list is certified to receive all messages sent on the specified subject using the RVCM quality of service. If a subscriber is not preregistered, it may miss one or more of the initial messages sent by the publisher adapter.

To preregister a certified subscriber in TIBCO Designer:

1. In the project tree panel, under the publisher adapter, open the Advanced>Sessions>adbadapterInstanceNamervcmRvCmSession endpoint object.

Wildcard TIBCO Rendezvous subject names are not supported for preregistered listeners.



2. Click the publisher endpoint in the design panel.

The Configuration Tab for this endpoint is displayed, as shown below.

3. Verify the name of the certified subscriber in the Preregistered Listeners field: %%Domain%%.%%Deployment%%.adb.%%Adapter_name%%.CM. In this example, the preregistered listener is the subscriber adapter named rvsub.

4. Click Apply.


Changing the Location of the Ledger File | 175

Changing the Location of the Ledger File

The ledger file provides temporary storage for published messages that are sent by a publisher adapter using the TIBCO Rendezvous certified messaging quality of service. Each message is held in the ledger until an acknowledgement is received from the subscriber that the message has been consumed. If there are many subscribers, the ledger file can grow to a substantial size. In this case, it might be preferable to store the ledger file in a directory other than the default directory (which is install-path/ledger).

To change the location of the ledger file in TIBCO Designer:

1. In the project tree panel, navigate to the open the Advanced>Sessions> adbadapterInstanceNamervcmRvCmSession object, as shown in the previous section, Preregistering a Certified Subscriber.

2. Edit the Ledger File value to point to the new ledger file location.

In the following example the name of the ledger file has been changed to adbdemo2pubrvcm.ldg, and it will be generated in the directory in which you run the adapter instance.



When the adapter instance is restarted, it writes the ledger file to the new location. For more information on using TIBCO Designer, see the TIBCO Designer User? Guide.


Sending Messages to an Adapter Instance | 177

Sending Messages to an Adapter Instance

When TIBCO ActiveMatrix Adapter for Database publishes messages including datetime data, the wire format is handled implicitly since both publisher and subscriber use the same format.

In many scenarios, however, a separate product acts as the publisher, which means that you must explicitly specify the wire format. For example, an adapter instance might subscribe to messages published by another TIBCO ActiveEnterprise product, such as TIBCO ActiveMatrix BusinessWorks, after some transformation has been performed.

Messages containing datetime data sent to TIBCO ActiveMatrix Adapter for Database adapters must have that data in a string format. If it is in a different format, the TIBCO ActiveMatrix Adapter for Database adapter will not be able to process the date values. Formatting dates as strings allows the adapter to handle date values outside the range allowed by RVMSG_DATETIME, January 1, 1970 to January 1, 2034.

The format used by the adbDateTime metadata class is described below.

adbDateTime Object

To view the structure of the adbDateTime class:

1. In the project tree panel, navigate to AESchemas>ae>ADB>adbmetadata>Classes>adbDateTime>dateTime

Do not change any adbDateTime object information in this screen. Use this information to determine how to format the third-party subscriber date value.




2. Click the adbDateTime entry.

3. Click the dateTime entry.

The format of the c lass is displayed. The input is a string value with one of the following formats:

yyyy-mm-ddyyyy-mm-dd hh:mm:ss.xxx

where xxx represents the millisecond value.


Publishing by Reference Object | 179

Publishing by Reference Object

When source data is stored in a view or other database object, you can publish this data using the publish by reference object feature. Publishing by reference object is an extension of the publish by reference feature. For a description of publishing by reference, see Publish by Reference on page 15.

In both cases, only key values from the source table are stored in the publishing table. The difference when publishing by reference object is when a row changes in the source table and the associated trigger fires, the adapter fetches data from the reference object, rather than the source table. The name of the reference object is stored in an ADB_REF_OBJECT column in the publishing table. For a description of this column, see Publishing Table on page 154. Publishing by reference object is recommended when a view provides the most efficient access to source data, for example when many levels of nesting exist between a parent and child table.

To configure a publisher adapter to publish data by reference object:

1. Drag the ActiveDatabase Adapter Configuration icon to the design panel or select the resource in the project panel.

2. Drag a Publication Service icon to the design panel.

3. To add a source table for a publisher adapter, do the following in the Tables tab:

a. Click the Add Table icon. The Add Table dialog displays.

b. Select a source table from the list and click OK.

If parent-child relationships are used, follow the steps under Adding Child Tables on page 76 to add child tables.

You must explicitly designate a key column or substitute key column for the reference object, since a reference object has no external designation of the key column.

4. Specify the following options in the Publisher Options tab:

Storage Mode — From the drop-down list, select Publish by Reference.

Publishing Table — Type the table name to use for storage. A common practice is to use the source table name prefixed by P_.

Referred Object — Type the name of the view or other database object to select source data from. To select from a list of tables in the current user schema, click Add. To select an object from a different schema, select Add From.



Update Mode — Select the method for updating tables from the drop-down list. Choose Update to update a row in the destination table only if the row exists. Choose Upsert to update a row in the destination table if the row exists and, if no such row exists, perform an insert.

5. Click Apply.

Designating the Key Column

To designate a column of the reference object as the key:

1. In the TIBCO Designer project tree panel, navigate to AESchemas\ae\ADB\adapter_instance\classes\ref_object\key_column. The TIBCO Designer window should look similar to the following:

2. In the Configuration tab, click the Key Field checkbox to select it.

3. Click Apply.

Changing Repository Objects for Parent-Child RelationshipsIf your publication service includes parent-child relationships, you must also add a sequence and association to the metadata stored in the repository. These tasks are described in the following sections.



Adding a Sequence

To add a sequence:

1. In the TIBCO Designer project tree panel, navigate to AESchemas\ae\Classes\ae\ADB\adapter_instance\Classes\reference_object.

2. Drag a Generic Sequence icon from the palette panel to the design panel to add it to the reference object class. The TIBCO Designer window should look similar to the following:

3. In the Name field, type ADB_SEQUENCE_childTableName.



4. In the Select Sequence field, click Browse. The Select Resource dialog displays:

5. Click the name of the sequence that represents the child table for the publication service, for example, sequence[ORDER_DETAILS].

6. Click OK.

7. In TIBCO Designer, click Apply. The sequence is added to the metadata for the reference object.

Adding an Association

To add an association:

1. In the TIBCO Designer project tree panel, navigate to AESchemas\ae\ADB\adapter_instance_name\Associations.

Several associations could display, at least one for the parent table to the child table on the publisher side, and another for the destination parent table to the child table on the subscriber side.



2. Click each of the resource icons that represent the parent-child relationship on the publisher side. For example:

3. In the Configuration tab, click in the Name field and replace the name of the publishing table with the name of the reference object. For example,P_CUSTOMERÔRDER_DETAILS becomes ORDER_VIEWÔRDER_DETAILS.

4. Click Apply.

5. In the project tree panel, double-click the new association name to expand it.



6. In the project tree panel, click Left Association to select it.

7. Enter a Role Name value that contains the reference object name. For example, roleORDER_TABLE.



8. In the Endpoint Class field, click Browse. Navigate to the class that represents the reference object, in this case ORDER_TABLE, and click OK.

9. In TIBCO Designer, click Apply.

ExampleIn the following example, the publisher adapter is configured to publish source data from the CUSTOMER table and its child table, EXTERNAL_ORDER_DETAILS. The relevant key columns are CUST_ID and ORDER_ID.

The publishing table is created with necessary adapter fields, as well as CUST_ID and ORDER_ID fields. When a row in CUSTOMER is modified, the trigger fires, populating fields and copying the CUST_ID and ORDER_ID values, as well as the name of the reference object, ORDER_VIEW, to the publishing table. When the



adapter polls the publishing table, it detects the new row. The adapter then selects the order and customer data from ORDER_VIEW, using the CUST_ID and ORDER_ID values along with the view name found in the publishing table. Then the message is published.

Table to Record Sequence Numbers (DB2 on iSeries)The palette will attempt to create the following table (if it does not exist) to record sequence numbers:

CREATE TABLE library.ADB_SEQTAB (PUB_TABLE VARCHAR(64) NOT NULL,

ADB_SEQ NUMERIC(20), constraint ADB.ADB_SEQTAB_KEY primary key (PUB_TABLE))

The table must be journalled.

You can create this table or if journalling is automatically turned on, the palette will create the table automatically.


User Callout Library | 187

User Callout Library

The user callout shared library allows the transformation of a message that the adapter publishes into a structure that you want to be published. The shared library can be customized to apply limited transformations to outgoing and incoming messages.

To use the callout library, the TIBCO Adapter SDK must be installed because the callout library uses SDK include files and links with SDK libraries. Additionally, you must use the same TIBCO Adapter SDK version that your adapter instance is using.

If a message is modified before it is sent by a publisher adapter, all subscribers will receive the modified message. If a message is modified before it is received by a subscriber adapter, only that adapter will get the modified message.

For example, at the publisher adapter, a message can be modified to have a field added, or publish an empty message if a certain criteria is met. At the subscriber adapter a message can be modified to be written to the database or not, based on a filter, or have the subject name inserted into the table.

Before a message is sent or received, the adapter checks the callout library. If the callout library has not been modified, the message is passed back to the adapter unchanged. If the callout library has been modified, the message is passed back to the adapter with the modification. After receiving a modified message from the callout library, the adapter instance sends or consumes the message like any other message.

• The altermsg.cpp source file must be edited with your custom code.

• The callout library must be rebuilt for your platform.

The source code is included in your installation media in the install-path\demo\altermsg directory.

If you send or receive messages in the TIBCO Rendezvous Message wire format, the TIBCO Rendezvous Message APIs cannot be used to directly alter those messages. Instead, you will receive an MTree to which various functions can be applied to view or traverse the TIBCO Rendezvous message. If you must change the message, create a new message and pass it back to the adapter using the newmsg argument.



Using the alterMsgPub and alterMsgSub FunctionsTwo functions are available in the altermsg.cpp source file for creating callouts, alterMsgPub() to create a callout for a publisher and alterMsgSub() to create a callout for a subscriber. Two examples are included in the altermsg.cpp file. You can change the examples for your needs, or write new code in the file. The first example appends a new field to a message. The second example works in conjunction with the demo and returns an empty message if the message meets a certain criteria.

The alterMsgPub() and alterMsgSub() functions are the entry points into the callout library. These two function names and their signatures must not be changed.

The alterMsgPub and alterMsgSub functions allow you to alter messages in the TIBCO ActiveEnterprise Message, TIBCO Rendezvous Message, and XML wire formats. The function signatures are:

alterMsgPub(char * subj,const MTree *msg,MTree **newmsg,MInstance *pInst,MApp *pMapp,MMessageFormat dataformat);

alterMsgSub(char * subj,const MTree *msg,MTree **newmsg,MInstance *pInst,MApp *pMapp,MMessageFormat dataformat);

Both functions share the same set of parameters. However, several parameters are specific to a particular wire format.

The following table contains parameter descriptions and a column for each format. If the R column contains an X, use the parameter for TIBCO Rendezvous Message wire format. If the M column contains an X, use the parameter for TIBCO ActiveEnterprise Message wire format. Otherwise, leave the value as NULL.

Table 10 Wire Format Parameters and Descriptions

Parameter R M Description

subj X X Subject on which the message is sent or received.

msg X MTree containing the message to send or receive.

newmsg X New MTree user creates if msg must be altered.



When sending or receiving messages in the TIBCO ActiveEnterprise Message wire format, you must use the pInst argument to change the message.

Using the adbPreCommit FunctionThe adbPreCommit function allows you to perform a custom operation, such as invoking a stored procedure or sending a TIBCO Rendezvous message, just before the transaction is committed. This function is called by the subscriber after all inserts, updates and delete operations have been performed on both parent and child tables.

The function signature is:

adbPreCommit(SQLHDBC connectHandle,MApp *pMapp,MInstance *pInst,MTree *pTree,MMessageFormat dataformat);

The following table contains parameter descriptions.

pInst X MInstance of the message. Alter directly if necessary.

pMapp X TIBCO Adapter SDK MApp structure for this MInstance.

dataformat X X The data format of the message.

Table 10 Wire Format Parameters and Descriptions (Cont’d)

Parameter R M Description

Table 11 adbPreCommit Parameter Descriptions

ParameterTIBCO Rendezvous Format

TIBCO ActiveEnterprise Format

Description

connectHandle X X The database connection handle.

pMapp X X TIBCO Adapter SDK MApp structure for this MInstance.

pInst X MInstance of the message.

pTree X MTree containing the message.

dataformat X X The data format of the message.



Building the Callout LibraryThis section explains how to build the callout library on each supported platform. The next table lists required compilers for each platform.

Building the Callout Library on Microsoft Windows

1. Using Visual C++, create a new workspace and insert a new project into the workspace.

2. In the new project, open the file:

install-path\demo\altermsg\altermsg.dsp

If an adapter is configured to use the adb.subBatchCommitTimeout or adb.subBatchCommitSize options, the custom operation is performed once for each message received. However, the entire transaction is not committed until the size or time-out values are reached. Adapter instances that use the bulk-insert-size option cannot use adbPreCommit. For more information on these options, see Appendix D, Adapter Properties File, on page 321.

Table 12 Compilers for Platforms

Platform Compiler

Microsoft Windows

MSVC++ 6.0 with Service Pack 5, 7.0 or 7.1 (.NET)

Solaris 8, 9, 10 CC 5.0 (SPARCworks) or

Sun Forte Developer 6 Update 2 compiler

HP-UX 11i aCC 3.13

AIX 5.2 and 5.3 XLC_r v5.1.0 (You must use the xlC_r compiler)

Linux GNU gcc v2.96.112



3. Select Project>Settings. In the Project Settings dialog box:

a. As shown next, select the C/C++ tab. In Category, select Preprocessor. In Additional include directories type the path to the TIBCO Rendezvous include directory.

b. As shown next, select the Link tab. In Category, select Input. In Additional library path type the path to the TIBCO Rendezvous library directory.

4. Select Build>Build adbaltermsg.dll to create the library.



5. After building the library, change to the <install-path>/tibco/adapter/adadb/5.5/lib directory.

a. Make a backup copy of the existing adbaltermsg.dll library.

b. Copy the new adbaltermsg.dll library into the bin directory.

Building the Callout Library on UNIX

1. Using a text editor open the MakeFile and set the RV_Home environment variable. For example, for Solaris 8:RV_HOME=/usr/tibco/tibrv

2. Using a text editor open the MakeFile and set the SDK_Home environment variable. For example:

SDK_HOME=/local/activedb/sdk50

3. Set the command to invoke your compiler. For example:CC= CC

4. Set the options to be passed to your compiler. For example:DEBUG=-gOPT=-O4

5. Save the MakeFile and execute:% make libadbaltermsg.so

6. After building the library, change to the install-path\lib directory.

a. Make a backup copy of the existing libadbaltermsg.so library.

b. Copy the new libadbaltermsg.so library into the lib directory.


Unique Connection Identifier for Oracle Databases | 193

Unique Connection Identifier for Oracle Databases

Within each adapter instance, the adapter typically creates multiple database connections or sessions. Each connection has a different responsibility (such as selecting from the publishing table or inserting into the destination table) and each is identified using a unique connection identifier.

The unique identifier added to each session allows a database administrator to monitor the adapter’s database activity from Oracle by being able to identify what each database connection is doing. The Oracle DBA can query the CLIENT_INFO column of V$SESSION to identify each Oracle connection the adapter is using. The identifiers for each service are:

• InstanceId.ADBPubPolling (publication service polling connection)

• InstanceId.ADBPubUpdate (publication service update connection)

• InstanceId.ADBPubConfirm (publication service confirm connection)

• InstanceId.ADBSubscriber (subscription service connection)

• InstanceId.ADBRequestReply.n (request-response service connection where n is the thread ID)



Publishing in a Commitment Controlled Environment

You can enable the adapter to publish messages in a commitment controlled environment.

In its standard functionality the adapter generates and attaches database triggers to an application table that is the source for publishing data. A trigger program is activated on a database event on the source table (Insert, Update, Delete) and the trigger program uses the source table data to reflect the database event change on a publishing table.

Therefore, if the source table is being accessed by an application under commitment control, it is critical from a database integrity standpoint that the trigger program be part of the same commitment cycle.

To publish messages in a commitment controlled environment:

These steps assume that the trigger program is being generated by the palette in the RPG ILE language.

1. Once the adapter is configured and saved in TIBCO Designer, the trigger programs will be compiled and attached to the source table. Log on to the iSeries system using a profile that has *PGMR rights to the library and source file where the trigger program is created. The trigger program is always created in a source file called TIBCOSRC that is in a library representing the DB2 collection.

2. Edit the trigger program member in the TIBCOSRC file by adding the keyword 'Commit' on the F-Spec for the publishing table. An example is given below.

FP_EMPL O E Disk Rename(P_EMPL:Pub_Record)F UsrOpnF Commit

Make sure that the application program that is accessing the source table is running with commitment control enabled. Otherwise, the trigger program will encounter a runtime error with this keyword.

3. After editing the trigger program member, save and compile it with options DFTACTGRP(*NO) and ACTGRP(*CALLER). These options ensure that the trigger program runs as part of the calling application program's commitment cycle. The trigger program will run in the application program's activation group. Any database updates done by the trigger program will be done or undone by a Commit or Rollback issued by the application program, thus ensuring database integrity.

This feature applies only to DB2 on iSeries.


Publishing in a Commitment Controlled Environment | 195

4. To have the trigger program run under commitment control, the publishing table (listed in the trigger program) must be journalized. This can be done by using the iSeries command STRJRNPF. The journal that is being used to log changes to the source table can be used for the publishing table as well.



Using the Adapter with a Revision Control System

TIBCO Designer supports revision control systems such as MicroSoft Visual SourceSafe and Perforce. If you are using a revision control system, you must manually add some configured resources to the revision control system and check in the resources when completing the instance configuration.

As part of service configuration, the adapter creates schema files in root/AESchemas/ae/adadb. For example, if you configure a service in an adapter configuration Instance1, the following files are created:

Project_root/AESchemas/ae/adb/Instance1.aeschemaProject_root/AESchemas/ae/adb/Instance1/businessObjects.aeschema

When the project is saved and a revision control system has been specified, the adapter displays a warning that additional files were created and should be added to the revision control system. This warning appears only when the files are created for the first time. The warning displays a Go To Resource button that helps in navigating to the resource. You should use the Multi-User>Add Resources to RCS menu command to add these files to the revision control system.

For information about how to use the Multi-User feature in TIBCO Designer, refer to the TIBCO Designer User’s Guide.

Copy, Cut, Paste and Move Operations

To successfully copy and paste a service from adapter Instance1 to Instance2, the adapter configuration and schema files for the Instance2 must be checked out.

To successfully cut and paste a service from adapter Instance1 to Instance2, the adapter configuration and schema files for both Instance1 and Instance2 must be checked out.

To successfully move a service from adapter Instance1 to Instance2, the adapter configuration and schema files for both Instance1 and Instance2 must be checked out.

Regeneration When Moving, Copying and Pasting

• Default subjects are not regenerated to reflect the new instance name when a service is moved.

• Manually changed certified messaging and certified messaging queue ledger file names are regenerated to defaults when a service is moved, or copied and pasted to a new instance.


Using the Adapter with a Revision Control System | 197

• If a service associated with a custom session is moved, or copied and pasted, the custom session is not moved, or copied and pasted. The session is regenerated as a default session.



Using a Log File for an Adapter Instance

By default all error, warning and information messages are printed in the console window in which the configuration was started and to a default log file. The log file can be located anywhere on your file system.

The adapter will trace to whatever sink is present in your adapter instance configuration. If you have the default sinks of stdioSink and fileSink, the adapter will write to both the file and stdout.

Fine-Tuning Tracing AttributesThe tracing facility allows you to fine-tune where and what different types of information are sent.

To access tracing attributes for an adapter instance:

1. In the project tree panel, click the corresponding ActiveDatabase Adapter resource to select it.

Only errors that originate in the adapter instance itself are logged in the log file. Errors from other sources, for example TIBCO Rendezvous APIs or TIBCO Adapter SDK APIs, are not logged to a log file, but are printed to the console window where the configuration started.


Using a Log File for an Adapter Instance | 199


2. Display the Logging Tab, as shown below.

The Log to Standard I/O checkbox is selected so that trace information can be written to the console window where the adapter instance was started. The log file is generated in the start directory unless specified otherwise in the properties file.

To disable file tracing for the adapter instance, delete this file name and click Apply.



The following checkboxes control the level of trace information that is captured:

• Log Info Messages—Capture error and warning information.

• Log Debug Messages—Capture all available information.

• Log Warning Messages—Capture warning information only.

• Log Error Messages—Capture error information only.

Log File SizeWhen a log file name is specified in the Logging tab, the adapter does the following:

1. Creates a file with no extension, using the file name specified in TIBCO Designer.

2. Redirects all trace statements generated by this configuration to that file until it reaches the default file size of 30 KB.

3. When this file size is reached (that is, as soon as the file is greater than or equal to the limit), renames the current file to file.1 and creates a new file with no extension.

Note that the log file can be slightly larger than the limit because the new file is only created after the limit has been reached.

4. Repeats this process of rolling log files and renaming files each time a new file is generated, until 3 log files exist.

Using Advanced Logging FeaturesThe example in this section shows how to change default log file attributes in TIBCO Designer.

To access advanced logging features in TIBCO Designer:

1. With the Logging tab selected in the Configuration panel, click the Use Advanced Logging checkbox.

With these default settings, the adapter instance overwrites the oldest file after 3 files have been created and the last file reaches 3 KB. For instructions on changing the default settings, see the next section.


Using a Log File for an Adapter Instance | 201

2. Click Apply. The following information displays:

3. In the project tree panel, navigate to adapter_instance_name\Advanced\Log Sinks\fileSink.

The TIBCO Designer window looks similar to the following:

File Limit (bytes)

Sets the maximum size of a log file.



File Count

Specifies the maximum number of log files that are created for this adapter instance.

Append Mode

When this box is checked (the default), information is written to the log file in the Append mode, where trace information is appended to existing entries in the file.

Unchecking this box changes to Overwrite mode, where the adapter instance overwrites the current log file the next time the adapter starts.


Using Database Cleanup Scripts | 203

Using Database Cleanup Scripts

Changing an existing adapter instance typically generates legitimate changes in the database. During this change process, the adapter creates a SQL script for changing the database objects and an associated cleanup script and stores them in the install-path\sql directory. If the legitimate database changes result in error messages, you may need to run these scripts. The messages and procedure for running the scripts are given in this section.

If the scripts created for legitimate database changes are not generated successfully, a message such as the following sample may display.



If the scripts created for legitimate database changes are generated successfully but an error occurs while these changes are being saved to the database, a message such as the following sample may display.

To Run the Cleanup Scripts

1. Select each message in the top portion of the message window and read the explanation for it in the bottom portion of the message window.

2. In TIBCO Designer, close the project containing the adapter instance that you changed.

3. Fix the errors that caused the database changes to fail, as indicated by the messages.

4. If necessary, clean up the old database configuration by running the scripts created by the adapter. The scripts are in the sql directory and are named instanceId.sql or instanceId.undo.sql, where instanceId is the name of the adapter instance you changed.

5. Reopen the project in TIBCO Designer.

6. Select the adapter instance you were attempting to change when the errors occurred.

7. Save the project.


Change the SQL Statement Terminator (DB2 for z/OS Only) | 205

Change the SQL Statement Terminator (DB2 for z/OS Only)

The SQL statement terminator by default is a semicolon ( ; ), but DB2 trigger statements use embedded semicolons as part of their syntax. Because of this, adapter SQL statements use a pound sign ( # ) as the SQL statement terminator.

When using the adapter in the default (real-time) mode, where repoOnly=false, the SQL statement terminator is not used and no changes are necessary.

However, when running the adapter in batch mode (repoOnly=true), you must change the SQL statement terminator from the default semicolon to the pound sign ( # ) in order to avoid conflicts with DB2 trigger statements.

Common methods of changing the SQL statement terminator are:

• Use SPUFI to set the terminator to # in the default panel.

• Include the following control statement at the beginning of the script file, or immediately before the DDL statement to be executed:

--#SET TERMINATOR #

• Use DSNTEP2 or DSNTIAD to code the RUN command as follows:

RUN PROGRAM(DSNTEP2) PLAN(DSNTEP61) PARMS(’SQLTERM(#)’)

Task A Change the Statement Terminator

The following steps change the statement terminator to a '#', then executes the SQL from the script file.

1. In a Windows command prompt, go to the directory where the script file resides.

2. Connect to DB2/OS390. For example:

C:\your-directoryDB2 Connect to S390LOC user logon_id using password

3. Execute your script file. For example:

C:\your-directoryDB2 -td# -f filename

4. In the Window's Command Center, go to the Script menu and select options.

5. Check the use statement termination character box and enter '#'.

6. Connect to DB2/OS390 from the Command Center GUI. For example:

Connect to S390LOC user logon_id using password

7. Cut the SQL statements from the script file and paste them into the Command Center GUI, then execute.



Fault Tolerance

Within the context of the adapter, a primary instance is the adapter instance that processes messages between the TIBCO environment and the database. The secondary instance uses the same adapter configuration but runs in a stand-by mode and takes over when the primary instance goes down.

To run in the fault tolerance mode, the property adb.mutex has to be defined. The value of this property is the name of the table that an adapter instance will attempt to lock.

At start up, if the adb.mutex property is present, the adapter will try to issue an exclusive lock on the table defined by the value of this property before processing any messages. The named table is created at runtime if it does not exist. The adapter instance that is able to lock this table is considered a primary instance. All instances that are not able to issue an exclusive lock on this table are considered secondary instances. Instances that specifies the same mutex table will be within a fault tolerance group.

The primary instance releases the lock prior to shutdown. If the primary instance terminates for an unknown reason, the lock is collected by the database system. As soon as the lock is released, one of the secondary instances will hold the lock and become the primary.

If the primary instance looses connection to the database, a reconnection will not be attempted. The adapter instance will release the lock and gracefully shutdown letting any secondary instance take over as the primary.

The time interval that the secondary instance(s) wait before attempting to lock the table specified by the adb.mutex property is defined by the adb.heartbeat property.

On Oracle, you have to set the SQLNET.EXPIRE_TIME in sqlnet.ora script to a value greater than 0. This ensures that the server will collect the lock held by the invalid connection of the primary instance due to connection lost.


SSL Data Encryption using DataDirect ODBC Drivers | 207

SSL Data Encryption using DataDirect ODBC Drivers

SSL data encryptions provide secure transmission of data. The adapter supports SSL data encryption between the Oracle database and the adapter with DataDirect Oracle Wire Protocol ODBC Drivers.

You may want to use data encryption in the following scenarios:

• You have offices that share confidential information over an intranet.

• You send sensitive data, such as credit card numbers, over a database connection.

• You need to comply with government or industry privacy and security requirements.

Refer to ODBC Driver documentation for more information on the available features and the required configuration details..

Data encryption may adversely affect performance because of the additional overhead required to encrypt and decrypt data.




| 209

Chapter 9 Setting Encoding Options

This chapter describes how to use TIBCO ActiveMatrix Adapter for Database in an international environment using non-ASCII languages.

Topics


• Setting TIBCO Messaging Encoding, page 212

• Configuring the Adapter to Communicate with the Database, page 214

• Relevant Environment Settings, page 219


210 | Chapter 9 Setting Encoding Options

Overview

TIBCO ActiveMatrix Adapter for Database provides internationalization support by taking advantage of Unicode, which is supported by the TIBCO Adapter SDK. Currently the international data support is provided for text data only. Database name, database schema, metadata and error messages do not currently support international character sets, but these features will be available in a future release.

The following illustration shows an example of TIBCO ActiveMatrix Adapter for Database in a Japanese language environment serving databases of different encodings. The adapter converts the encoding between Unicode and the character sets of the databases.

Figure 7 Example of Unicode Conversion

In this example, Shift-JIS encoded Japanese data is retrieved from an Oracle database by the adapter’s publication service, which converts the data into UTF-8 and publishes the UTF-8 data to the TIBCO messaging environment. An adapter subscription service receives this message and converts it from UTF-8 to EUC-JP, and inserts the data into an Oracle EUC-JP database to which it is connected.

ActiveDatabase Publishing Service

ActiveDatabase Subscription Service

TIBCO Messaging

Shift-JIS

UTF-8 UTF-8

EUC-JP

Oracle DB in Shift-JIS

Oracle DB in EUC-JP

Driver Manager/ODBC Driver/Database Client


Overview | 211

By using UTF-8 as the TIBCO Messaging encoding between TIBCO ActiveEnterprise components, the two adapter services can serve databases of different encodings by exchanging data without data garbled. For detailed information about how TIBCO Adapter works in an international environment, please refer to TIBCO Adapter Concepts.

In circumstances where only ASCII and Latin-1 data (including English and Western European languages) are exchanged in the system, you can use Latin-1 encoding as the TIBCO Messaging encoding between TIBCO ActiveEnterprise components. It is not necessary to use UTF-8 as the TIBCO Messaging encoding in this case. UTF-8 is only required when the data to be exchanged is not included in the Latin-1 character set, as is the case with Asian languages and other European languages.

To configure the adapter to work correctly in transmitting non-ASCII data, you must set the TIBCO messaging encoding property. The following sections explain how to do this.

In certain cases, intermediate entities (such as an ODBC driver, ODBC driver manager, or the database vendor's proprietary database client) may also perform some encoding conversion. This is discussed later in this chapter.



Setting TIBCO Messaging Encoding

TIBCO ActiveEnterprise components (TIBCO applications and adapters) use TIBCO Messaging encoding when communicating with each other. In the case of TIBCO ActiveMatrix Adapter for Database, publication services use this encoding to publish data into TIBCO messaging system, and subscription services use it to receive data from the TIBCO messaging system.

There are currently two choices for TIBCO Messaging encoding: Latin-1 (ISO8859-1) for transmitting ASCII and Latin-1 data, and UTF-8 for transmitting data of other languages.

The encoding property is set on the project itself at design-time, and in the TIBCO administration server’s property file when creating a TIBCO administration domain.

Encoding in a Server-based Project

The TIBCO administration server setting is used when the project is exported to a server repository or deployed using TIBCO Administrator Enterprise Edition.

For a server based project, the TIBCO messaging encoding is set by the repo.encoding property in the server's tibcoadmin<domain-name>.tra configuration file (located in <install-path>/tibco/administrator/n.n/bin/).

The encoding is set when using the TIBCO Domain Utility to create the domain or by editing the repo.encoding property in the .tra configuration file.

Each adapter or TIBCO application that uses the same server for storing and retrieving configuration data uses this encoding setting when communicating to each other. This assures that all TIBCO components (including adapters and other TIBCO applications) that belong to the same project use the same encoding value to communicate.

Encoding in a Local Project File

If the project's configurations are saved in a local project file, which is the normal case when the project is in design stage before deployment, the TIBCO Messaging encoding is determined by the encoding property saved in the local project file.

The services do not use TIBCO Messaging encoding to connect to the backend database. See Configuring the Adapter to Communicate with the Database on page 214 for more information.


Setting TIBCO Messaging Encoding | 213

The encoding value is set on the root project folder in TIBCO Designer. By default, the value is set to ISO8859-1. You can change the value by selecting the folder and under the Project Settings tab, changing the value for the TIBCO Message Encoding field.

In order for different TIBCO components that each have their own local file repository to communicate with each other without data garbled, they must use the same TIBCO Messaging encoding, meaning that all these components must set their local project encoding to the same value.

After a project is deployed as a server-based project, the encoding property set at design-time is superseded by the encoding property set for the TIBCO administration server.

The encoding property of a local project file is only for determining the TIBCO Messaging encoding, not the encoding used for the persistent storage of the project files. The project files are always saved as UTF-8 encoded files.



Configuring the Adapter to Communicate with the Database

The TIBCO message encoding is a project setting used for data conversion among different components. However, the adapter uses a different encoding value for exchanging data with the database. Because there are many intermediate entities between the adapter and the database (such as the ODBC driver manager, ODBC driver, and possibly the database native client), and each of these entities may conduct encoding conversions to the data stream flow between the service and the database, the data encoding from the database might not be known, while the data encoding to the database must be compatible to that of the database.

To eliminate the complexity introduced by these intermediate entities, you can make some configuration changes that ensure the service’s encoding setting are compatible with the database instance's encoding. This also boosts the performance by eliminating unnecessary intermediate data conversions.

The configuration changes differ depending on the operating system that the service is running on, the database vendor, the ODBC driver used, and other factors. The following sections explain how to configure the adapter to communicate with the database.

Configuring the Adapter at Design-timeUse the following steps to set the adapter’s encoding.

1. Determine your database server encoding. This is the encoding you want to use for the adapter’s encoding. For example, for the Oracle database, the encoding can be found by querying the database with:

select * from nls_database_parameters;

2. Set the NLS_LANG environment variable for the client system.

Set the client machine's NLS_LANG environment variable to be consistent to the database instance's character set (NLS_CHARACTERSET) value. The client machine is where the adapter service is going to run. Consult your database administrator for the NLS_CHARACTERSET value of the database instance to connect to, and how to choose the right NLS_LANG value.

As an example, the system may use an Oracle database instance with the NLS_CHARACTERSET value set to JA16SJIS. In this case, the client machine's NLS_LANG environment variable must be set to JAPANESE_JAPAN.JA16SJIS.

The correct setting of the NLS_LANG environment variable eliminates the encoding conversion possibly conducted by the Oracle proprietary client.

3. Set the Adapter encoding in TIBCO Designer.


Configuring the Adapter to Communicate with the Database | 215

You must select or enter a valid encoding that the adapter understands. The adapter assumes the data coming from the database is in this encoding and ensures that data sent to the database is in this encoding. The encoding given here should match the encoding set for the database to which the adapter connects.

There are two ways to do this configuration:

— In TIBCO Designer, set this in the adapter resource. In the Configuration tab, click Show All Tabs. In the General tab, choose the encoding that is compatible with the encoding of the database instance to be connected to. Refer to Table 14, TIBCO ActiveMatrix Adapter for Database and Oracle NLS_Lang Values.

The adapter may support more encoding values than those shown in TIBCO Designer. You can type in these encoding values. For more information about these, refer to Appendix B in TIBCO Adapter Concepts.

— You can modify the adapter encoding property, adb.encoding, in the adapter’s.tra file. This setting supersedes the encoding option chosen in TIBCO Designer.

4. Set the ODBC DSN in TIBCO Designer

This field is set under the adapter’s Run-time Connection tab. You must also define the ODBC DSN on the machine on which the adapter is running.

— On Unix systems, modify the odbc.ini file <install-path>/tibco/adapter/adadb/5.5/odbc. In addition, you must set the correct IANAAppCodePage to match the database encoding, if you are using a wire protocol driver. See IANAAppCodePage on page 215 for a list of IANAAppCodePage values. The complete list is available in <install-path>/tibco/adapter/adadb/5.5/odbc/help/odbcref/odbcref.pdf

— On Windows systems, use the Windows ODBC GUI to define a system DSN. Note that if you are using a TIBCO wire protocol driver, the machine on which the adapter is running must be on the locale that matches the database encoding.

Table 13 IANAAppCodePage

Value Encoding (Language)

3 ASCII (English)

4 ISO_8859-1 (Western European)

17 Shift_JIS (Japanese)



Using a non-Oracle Database

When using a database other than Oracle, contact the database vendor for information about whether the database's proprietary client performs any encoding conversion to the data flow, and how to configure the database client to eliminate the conversion possibly conducted by it. Several possibilities exist:

• The database client does not perform any conversion to the data flow in any cases. In this case, you only need to set the adapter encoding to be compatible with the database instance's encoding.

• The database client relies on a local setting, such as an environment variable, to determine whether or not a conversion should to be conducted. This is similar to the above Oracle example, where you need to set both this local setting and the adapter’s adapter encoding property to be compatible with the database server instance's encoding.

• The database client relies on the local system encoding to determine the conversion.

— When the local system encoding is compatible with the database instance encoding, no conversion will be conducted by the database client. You only need to set the adapter encoding to be compatible with the database server instance encoding;

— When the local system encoding is different from the database instance's encoding, the proprietary database client may do the encoding conversion between the two different encodings. This conversion cannot be avoided by any settings you can make. In this case, set the adapter encoding to be consistent with the local system encoding, because the data from the

18 EUC_JP (Japanese)

38 EUC_KR (Korean)

113 GBK (Simplified Chinese)

2026 BIG5 (Traditional Chinese)

Table 13 IANAAppCodePage

Value Encoding (Language)


Configuring the Adapter to Communicate with the Database | 217

database has been converted to local system encoding by the database client.

Special Configurations for Supporting a UTF-8 DatabaseThis section applies only to TIBCO ODBC drivers.

Companies doing business globally may require their database to be multilingual, that is, a single database instance supporting multiple languages so the data from the company's global business can be stored in it. Unicode is a natural choice for the encoding of this kind of database instance.

This requires the adapter to be able to use Unicode, or specifically, UTF-8 encoding, to communicate with the database. While this is already a supported feature, an exception exists when using TIBCO ODBC drivers.

TIBCO ActiveMatrix Adapter for Database contains a binary mode for use with an embedded TIBCO driver when UTF-8 is the encoding to communicate with the backend database instance. The configurations described below are applicable to both Windows and UNIX platforms, and both non-wire and wire protocol drivers.

1. Set the adapter’s encoding option to UTF8.

2. In the adapter’s.tra file, uncomment the following line:

#ADB.WCHAR = SQL_C_BINARY

This forces the adapter to use binary mode to communicate with the database.

When using the DataDirect non-wire protocol ODBC driver and UTF-8 as the adapter encoding to support multilingual data exchange, the configuration is totally different. Please refer to the section Special Configurations for Supporting a UTF-8 Database.

These are for the internationalization configuration, which forces the adapter to use binary mode to communicate to the database instance. No further encoding related configurations are necessary. For example, on UNIX there is no need to set the AppCodePage attribute, and when using a non-wire driver there is no need to set the proprietary database client.



Run-time Environment VariablesIn general, after deploying the adapter, all information should be in the correct place, however you should verify the deployed .tra file and platform settings. The encoding conversion is dependent on the TIB_ICU_DATA environment variable. The deployed adapter .tra file should have a property called tibco.env.TIB_ICU_DATA. Make sure it points to the directory containing the tibicudata.dat file.

For the Oracle database, make sure the installed NLS_LANG attribute’s value matches the database character set. If not, set NLS_LANG to the correct value before starting the adapter. See TIBCO ActiveMatrix Adapter for Database and Oracle NLS_Lang Values on page 219 for a list of values.


Relevant Environment Settings | 219

Relevant Environment Settings

In order to support international languages, the adapter uses the tibicudata.dat file to convert between the original database encoding and the TIBCO Messaging encoding (usually UTF-8). The environment variable TIB_ICU_DATA is already configured to point to the directory containing this file. Do not change this setting.

This environment variable setting is required on both Windows and Unix platforms.

The following table lists the TIBCO ActiveMatrix Adapter for Database encoding values as shown in TIBCO Designer.

Table 14 TIBCO ActiveMatrix Adapter for Database and Oracle NLS_Lang Values

Adapter Encoding Oracle NLS_LANG (containing NLS_CHARACTERSET) Description

ASCII US7ASCII 7-bit ASCII

ISO8859-1 language_territory.WE8ISO8859P1 ISO8859-1 (Latin-1), West European

Shift_JIS (CP943) JAPANESE_JAPAN.JA16SJIS Japanese Shift-JIS, CP943

Shift_JIS (TIBCO) JAPANESE_JAPAN.JA16SJIS Variant of IBM-943 (created by TIBCO for flavoring some MS-932 conversions)

KSC-5601 KOREAN_KOREA.KO16KSC5601 Korean KSC-5601

Big5 TRADITIONAL CHINESE_TAIWAN.ZHT16BIG5

Traditional Chinese Big5 with Euro

GBK SIMPLIFIED CHINESE_CHINA. ZHS16CGB231280

Simplified Chinese GBK (superset of GB2312-80)

EUC-JP JAPANESE_JAPAN.JA16EUC Japanese EUC

UTF8 AMERICAN_AMERICA.UTF8 Unicode Transformation Format-8




| 221

Chapter 10 Monitoring the Adapter

Read this appendix if you have installed TIBCO Hawk and want to use TIBCO Hawk microagents through the Monitoring tab of an adapter instance. TIBCO Hawk is an optional tool from TIBCO Software Inc. that can be used in addition to the standard trace message logging method described under Logging Tab on page 60).

Topics


• Starting TIBCO Hawk Software, page 223

• The Auto-Discovery Process, page 224

• Invoking Microagent Methods, page 225

• Available Microagents, page 228


222 | Chapter 10 Monitoring the Adapter

Overview

TIBCO Hawk is a sophisticated tool for enterprise-wide monitoring and managing of all distributed applications and systems. System administrators can use it to monitor adapters in a wide area network of any size. TIBCO Hawk can be configured to monitor system and adapter parameters and to take actions when predefined conditions occur. These actions include: sending alarms that are graphically displayed in the TIBCO Hawk display, sending email, paging, running executables, or modifying the behavior of a managed adapter.

Unlike other monitoring applications, TIBCO Hawk relies on a purely distributed intelligent agent architecture using publish or subscribe to distribute alerts. TIBCO Hawk uses TIBCO Rendezvous for all messaging and thus gains the benefits and scalability from the TIBCO Rendezvous features of publish/subscribe, subject name addressing, interest-based routing, and reliable multicast.

TIBCO Hawk is a purely event-based system that uses alerts. The agents are configured with rules that instruct them on everything from what and how to monitor to what actions to take when problems are discovered. Thus the workload is fully distributed throughout the enterprise. Every agent is autonomous in that it does not depend on other components to perform its functions.

The TIBCO Hawk Enterprise Monitor consists of these components:

• Display—GUI front end that displays alarms and provides editors to create rule bases, create tests, view messages, and invoke microagents to request information or initiate an action.

• Agents—Intelligent processes that perform monitoring and take actions as defined in rules.

• Rulebases—Rules that are loaded by agents to determine agent behavior.

• Application Management Interface (AMI)—Manages network applications via TIBCO Rendezvous and supports communication between a network application and monitoring TIBCO Hawk agents, including the ability to examine application variables, invoke methods, and monitor system performance.

• Microagents—Feed information back to TIBCO Hawk and expose action methods to rulebases.

For more information, see the TIBCO Hawk documentation.


Starting TIBCO Hawk Software | 223

Starting TIBCO Hawk Software

The TIBCO Hawk agent can be configured to start automatically during the system boot cycle. See the TIBCO Hawk Installation and Configuration guide for information about starting TIBCO Hawk.

The TIBCO Hawk Administrator’s Guide explains how to start the TIBCO Hawk Display.

The guides are included in your TIBCO Hawk software installation area.



The Auto-Discovery Process

After you start an instance of TIBCO Hawk Display, it continually discovers machines running TIBCO Hawk Agents on your network. Container icons are created for each agent, and arranged hierarchically in clusters. By default, agent icons are clustered according to subnets.

At first, the Agents container is empty. Its counter displays a value of zero and, on the right, the Discovered counter is also at zero. Both icons are initially green in color to show that no alerts, or warning messages, are in effect. As agents are discovered, the counters increment to reflect the current number of discovered agents:

Monitored network nodes are arranged in a hierarchical tree of containers. Clicking a container in the left panel displays nested items on the right.

Icon colors change to reflect the highest level of alert found on discovered agents. For explanations of icon elements and characteristics, see your TIBCO Hawk Administrator? Guide.


Invoking Microagent Methods | 225

Invoking Microagent Methods

A set of default microagents is loaded when a TIBCO Hawk Agent is started. When you install and start TIBCO ActiveMatrix Adapter for Database, its microagents are dynamically added to the local agent.

To invoke a microagent method:

1. In TIBCO Hawk Display, right-click on the agent icon and select Get Microagents.

If TIBCO Hawk security is implemented on your system and you do not have access to microagents on this agent, an error dialog displays. Select another agent, or contact your system administrator to obtain access.

The Microagents, Methods and Arguments dialog displays. The panel on the upper left lists microagents you can access on the current agent.

This dialog has two modes, Invoke and Subscribe. Invoking a method immediately returns a single set of current results. Subscribing provides updates of current results at regular intervals. Radio buttons at the bottom of the dialog control these modes.

2. Click a microagent name, such as Self, to display a list of associated methods and text descriptions in the panels below.



3. Click the name of the method to invoke, such as getComponentInfo.

If the method accepts arguments, fields for each argument display in the upper right panel. Detailed help text displays in the lower panel.

4. Specify any arguments for the method invocation.

5. Verify that the Invoke radio button is selected.


Invoking Microagent Methods | 227

6. Click Invoke to invoke the selected method.

The Invocation Results dialog displays the results returned by the method.

7. Click Done to close the dialog.

These steps describe how to interactively invoke a microagent method and receive a single set of results in TIBCO Hawk Display. You can also use a microagent method as the data source of a TIBCO Hawk rule. Rules automatically receive method results, apply tests to evaluate them, then take action if necessary. For more information on building TIBCO Hawk rules and rule bases, see your TIBCO Hawk Administrator’s Guide.



Available Microagents

Each adapter has two microagents, a standard TIBCO Hawk microagent named COM.TIBCO.ADAPTER.xyz (where xyz is the adapter configuration name) and a custom microagent. These microagents provide:

• Business level statistics—statistics that report the progress of the adapter as it interacts with the database. For example, in a database adapter such statistics might indicate whether objects were successfully or unsuccessfully inserted, updated, or deleted in the database.

• Queries that return information about the state of the adapter. This can be an important tool for seeing the internals of an adapter and debugging it if something appears wrong. For example, methods can return information about threads, internal queues, or connections to the target system. Using these methods, one might be able to identify certain bottlenecks or gauge how successfully an adapter is scaling with respect to the current environment.

• Updates of the adapter runtime parameters. This includes retrieving the current runtime parameters and setting new runtime parameters without restarting the adapter. An example of this is getting and setting the polling interval. Updating a runtime parameter through the Hawk microagent only affects the setting of the instance that is running. It does not make a permanent change of the setting in either the repository or the .tra file.

The following table lists each method available for the adapter and page on which the method is explained. Although the Microagents, Methods and Arguments dialog in TIBCO Hawk Display lists more methods than are documented here, only the following methods are supported.

Table 15 Standard Microagent Methods

Standard Method Description Page

activateTraceRole() Activates a mapping of a role to a sink at runtime.

232

deactivateTraceRole() Deactivates a mapping of a roles to sinks at runtime.

233

getAdapterServiceInformation()

Returns information about the services implemented by this adapter.

234

getComponents() Returns information about the publisher, subscriber and IODescriptor.

235


Available Microagents | 229

getConfig() Returns basic configuration information. More specific information is accessed by the more specific methods.

236

getConfigProperties() Returns a list of publishers and subscribers.

237

getHostInformation() Returns standard and extended application information.

238

getRvConfig() Returns information about all TIBCO Rendezvous sessions defined.

239

getStatus() Returns general status information, such as the number of TIBCO Rendezvous messages received and published, the number of errors since the last call, the PID of the application, and more.

240

getTraceSinks() Returns information about sinks to which traces currently go.

241

getVersion() Returns the configuration ID, application name, version, and date for this adapter instance.

242

_onUnsolictedMsg() Displays alert messages sent to the current adapter.

243

preRegisterListener() Preregisters an anticipated listener. 244

reviewLedger() Returns information retrieved from the ledger file of a certified messaging session for a publisher adapter.

245

setTraceSinks() Adds a role or changes the file limit of a previously specified sink.

246

stopApplicationInstance()

Stops the running adapter instance. 247

unRegisterListener() Unregisters a currently preregistered listener.

248

Table 15 Standard Microagent Methods (Cont’d)

Standard Method Description (Cont’d) Page



Table 16 Custom Microagent Methods

Custom Method Description Page

getEventQueueSize() Retrieves the size of the TIBCO Rendezvous event queue for the specified adapter instance.

249

setDebugLevel() Sets the debug level for the current adapter instance.

250

toggleVerboseFlag() Changes the value of the verbose flag from on to off, or the reverse.

251

showConfiguration() Shows the configuration defined for the current agent.

252

terminateADBagent() Stops the current TIBCO ActiveMatrix Adapter for Database agent.

253

getPollingInterval() Returns the current polling interval setting.

254

setPollingInterval() Set the polling interval for the publication service.

255

getPollingBatchSize() Get the polling batch size for the publication service.

256

setPollingBatchSize() Set the polling batch size for the publication service.

257

Table 17 Custom Microagent Methods with adb.perfMon

Custom Method Description Page

getActivityStatistics() Returns the total number of objects processed for all the schemas, based on the request type. Also, returns the number of success and error objects.

258


Available Microagents | 231

getActivityStatisticsByService

Returns the total number of objects processed for each of the schemas associated with the specified service.

259

getActivityStatisticsByOperation()

Returns the total number of objects processed for all the schemas by each service that is associated with a specified operation.

260

getThreadStatistics() Return the operation counts of the current threads.

261

getQueueStatistics() Return the current count of elements in any internal queue used by the adapter.

262

getConnectionStatistics() Returns the state and statistics for all the current connections used by the adapter.

263

resetActivityStatistics() Resets all the counts for the activity statistics.

264

resetThreadStatistics() Resets all the counts for the thread statistics.

265

resetQueueStatistics() Resets all the counts for the queue statistics.

266

resetConnectionStatistics() Resets all the counts for the connection statistics.

267

Table 17 Custom Microagent Methods with adb.perfMon (Cont’d)

Custom Method (Cont’d) Description (Cont’d) Page



activateTraceRole()

Purpose (Standard) Activates a mapping of a role to a sink at runtime. This replaces the now-deprecated setTraceSink() TIBCO Hawk method.

Parameters Parameters Type Description

roleName string Name of the role to activate.

sinkName string Name of a sink for which to activate the role.


deactivateTraceRole() | 233

deactivateTraceRole()

Purpose (Standard) Deactivates a mapping of a roles to sinks at runtime.

Parameters Parameters Type Description

roleName string Name of the role to activate.

sinkName string Name of a sink for which to activate the role.



getAdapterServiceInformation()

Purpose (Standard) Returns information about the services implemented by this adapter. The information is a summary of available adapter services.

Parameters

Returns

Parameter Type Description

serviceName string Name of the service from which to get information. Default is ALL.

Returns Type Description

Line Integer Sequential row number.

Name string Name of the Service.

Endpoint string Name of the endpoint used for this service.

Type string Type of the endpoint, for example, Publisher, Subscriber.

Quality of Service

string Quality of service for the endpoint, for example, RV, RVCM.

Adapter Name string Name of the application for this sink.

Session Name string Name of the TIBCO Rendezvous session.

Subject string Subject defined for this endpoint

Number of Messages

Integer Number of messages processed for this endpoint.


getComponents() | 235

getComponents()

Purpose (Standard) Returns information about the currently active TIBCO Hawk components such as publishers, subscribers, or timers.

Parameters

Returns

Parameters Type Description

Component Name string Name of the TIBCO Hawk component. Default is all components.

Component Type string Any of Publisher, Subscriber, Timer, or IODescriptor. The default value is All.


Component Name string Name of the TIBCO Hawk component.

Instance ID string Name of this adapter instance.

Adapter Name string Name of the adapter.

Session Name string Name of the TIBCO Rendezvous session.

Component Type string The name of the TIBCO Adapter SDK class for this TIBCO Hawk component, such as MPublisher, MSubscriber, or MIODescriptorSource. For more information, see your TIBCO Adapter SDK documentation.

Description string Information about this TIBCO Hawk component, for example, time interval, signal type, validating publisher (or subscriber) etc.



getConfig()

Purpose (Standard) Retrieves generic configuration information. More specific configuration information is accessed through separate methods.

Returns Returns Type Description

Instance ID string Configuration ID of this adapter.


Repository Connection

string URL of the repository used for adapter instance.

Configuration URL

string Location of the adapter project; either a file name or configuration URL.

Command string Command line arguments used to start the adapter.


getConfigProperties() | 237

getConfigProperties()

Purpose (Standard) Returns all attributes and elements for the given repository object.

Parameters

Returns


Property string Name of the property for which elements (tags) and attributes are desired. For example, rvpub/startup.

If no value is given, all properties are returned.


Element Name string Repository directory for the property.

Attribute Name string Name of the repository object attribute.

Attribute Value string Value of the repository object attribute.

Line integer Line number in which this property is defined in the project file.



getHostInformation()

Purpose (Standard) Returns standard and extended application information.


Name string Name of the property.

Value string Value of the property.


getRvConfig() | 239

getRvConfig()

Purpose (Standard) Returns information about the TIBCO Rendezvous session defined by this adapter. Information about all currently defined sessions is returned if no sessionName is provided.

Parameters

Returns


Session Name string Name of the TIBCO Rendezvous session for which configuration is required (default is all).


Instance ID string The configuration ID of this adapter.


Session Name string Name of the session.

Service string Service parameter for this session.

Daemon string Daemon parameter for this session.

Network string Network parameter for this session.

Synchronous? boolean Returns 1 if this is a synchronous session, 0 otherwise.

Session Type string Type of session; one of M_RV, M_RVCM, or M_RVCMQ.

Certified Name string Name of this certified session.

Ledger File string Ledger file for this certified messaging session. Returns the empty string for sessions that are not certified messaging sessions.

CM Timeout string Timeout for this certified messaging session. Returns the empty string for sessions that are not certified messaging sessions.



getStatus()

Purpose (Standard) Retrieves basic status information about the adapter.

This information is fairly limited; for more detail, additional methods are provided: getConfig() on page 236 and getRvConfig() on page 239.


Instance ID string Configuration ID for this adapter instance.


Uptime integer Number of seconds since startup.

Messages Received integer Number of TIBCO Rendezvous messages received.

Messages Sent integer Number of TIBCO Rendezvous messages published.

New Errors integer Number of errors since the last call to this method.

Total Errors integer Total number of errors since startup.

Process ID integer Process ID of the application.

Host string Name of host machine on which this adapter is running.


getTraceSinks() | 241

getTraceSinks()

Purpose (Standard) Returns information about sinks to which traces currently go.

Parameters

Returns


Sink Name string Name of the sink for which you need information. If no name is specified, information about all sinks is returned. Default is all.

Role Name string Name of the role for which you need information for the specified sink or sinks. Default is all.


Instance ID string Name of this adapter instance as a string.

Adapter Name string Name of the application for this sink.

Sink Name string Name of the sink

Sink Type string Type of this sink. One of fileSink, rvSink, hawkSink, stderrSink.

Roles string Roles this sink supports, as a string. For example “warning, error, debug”.



getVersion()

Purpose (Standard) Retrieves version information for the current application. Two lines may be returned, one for the TIBCO Adapter SDK, one for the adapter.

Returns Returns Description

Instance ID The configuration ID as a string, for example SDK.

Adapter Name Name of the adapter as a string, for example rvpub.

Version Version number as a string, for example 1.1.


_onUnsolictedMsg() | 243

_onUnsolictedMsg()

Purpose (Standard) Displays all alert messages sent from the adapter or an error if not successful.



preRegisterListener()

Purpose (Standard) Preregister an anticipated listener. Some sending applications can anticipate requests for certified delivery even before the listening applications start running. In such situations, the sender can preregister listeners, so TIBCO Rendezvous software begins storing outbound messages in the sender’s ledger. If the listening correspondent requires old messages, it receives the backlogged messages when it requests certified delivery.

Parameters

Returns OK if the listener was preregistered successfully, false otherwise.


Session Name string Name of the session that anticipates the listener.

Publisher Name string Name of the component for which the listener should be preregistered.

Listener Session Name

string Name of the listener to preregister.


reviewLedger() | 245

reviewLedger()

Purpose (Standard) Returns information retrieved from the ledger file of a TIBCO Rendezvous certified messaging session. Before invoking this method, ensure that the certified messaging publisher adapter has established a certified delivery agreement with its subscriber agents.

Parameters

Returns


Session Name string Name of the TIBCO Rendezvous session for which ledger information is desired (default is all).

Subject string Name of the subject for which ledger information is desired.


Session Name string Name of the TIBCO Rendezvous CM session to which this information applies.

Subject string Subject name for this session.

Last Sent Message integer Sequence number of the most recently sent message with this subject name.

Total Messages string Total number of pending messages with this subject name.

Total Size integer Total storage (in bytes) occupied by all pending messages with this subject name. If the ledger contains ten messages with this subject name, then this field sums the storage space over all of them.


string Within each listener submessage, the Listener Session Name field contains the name of the delivery-tracking listener session.

Last Confirmed string Within each listener submessage, the Last Confirmed field contains the sequence number of the last message for which this listener session confirmed delivery.

Line integer Row number in ledger file.

UnacknowledgedMessages

integer Number of RVCM messages pending for this listener. The value is computed by subtracting the last sent sequence number from the last acknowledged sequence number.



setTraceSinks()

Purpose (Standard) Adds a role or changes the file limit of a previously specified sink.

Parameters

Returns OK if successful or an error if not successful.


Sink Name string Name of the sink for which you want to add a role or change the file limit.

Role Name string Name of the role you want to add to this sink (warning, error, debug, or user defined). Default is all.

File Size integer Maximum file size for this sink.

This parameter is ignored if the sink specified by sinkName is not a file sink.


stopApplicationInstance() | 247

stopApplicationInstance()

Purpose (Standard) Stops the specified adapter by calling the internal stop() method.

Returns OK if successful or an error if not successful.



unRegisterListener()

Purpose (Standard) Unregisters a currently preregistered listener.

Parameters

Returns True if the listener was successfully unregistered, false otherwise.


Session Name string Name of the session that anticipates the listener.

Publisher Name string Name of the component for which the listener should be preregistered.


string Name of the listener to unregister.


getEventQueueSize() | 249

getEventQueueSize()

Purpose (Custom) Retrieves the size of the TIBCO Rendezvous event queue for the specified adapter instance. For more information on events and event queues, see TIBCO Rendezvous Concepts.

Parameters

Returns


Session string Type of listener for this adapter instance. Possible values are Subscriber or Request/Reply. The default value is Subscriber.

Return Type Description

QueueCount integer The number of events currently in the event queue.

QueueLimit integer The maximum number of events this adapter instance can have in the event queue. This value is set using the adb.rvMaxQueueSize option in the adapter’s properties file. The default value is 0, which means the event queue has unlimited size.



setDebugLevel()

Purpose (Custom) Sets the debug level for the current adapter instance.

Parameters

Returns Returns OK if successful or an error if not successful.


DebugLevel integer Sets the debug level to 0 (off), 1, 2, or 3.

0— No debug information displayed.

1—SQL commands executed against the database shown.

2—ODBC data source for each SQL command shown.

3—All debug information displayed.


toggleVerboseFlag() | 251

toggleVerboseFlag()

Purpose (Custom) Changes the value of the verbose flag from on to off, or the reverse.

Returns This method returns OK if successful or an error if not successful.



showConfiguration()

Purpose (Custom) Shows the configuration defined for the current agent.

Parameters Parameter Type Description

VerboseInfo string The current setting for the verbose mode, on or off.

DebugLevelInfo string The debug level setting:

0— No debug information displayed.

1—SQL commands executed against the database shown.

2—ODBC data source for each SQL command shown.

PollIntervalInfo string The poll interval in milliseconds.


terminateADBagent() | 253

terminateADBagent()

Purpose (Custom) Stops the current TIBCO ActiveMatrix Adapter for Database agent.

Remarks Invoking this method displays a warning message with the text MicroAgent returned error. AMI Method invocation timed out. The message can be ignored.

Returns This method returns OK if successful or an error if not successful.



getPollingInterval()

Purpose (Custom) Returns the current polling interval setting.


PollingInterval integer Polling interval in milliseconds.


setPollingInterval() | 255

setPollingInterval()

Purpose (Custom) Set the polling interval for the publication service.


PollingInterval integer Polling interval in milliseconds.

ServiceName string (Optional) Name of service where the polling interval is set.



getPollingBatchSize()

Purpose (Custom) Get the polling batch size for the publication service.


BatchSize integer The batch size for the publication service.


setPollingBatchSize() | 257

setPollingBatchSize()

Purpose (Custom) Set the polling batch size for the publication service.


BatchSize integer The new batch size for the publication service.



getActivityStatistics()

Purpose (Custom with adb.perfMon) Returns the total number of objects processed for all the schemas, based on the request type. Also, returns the number of success and error objects.

Parameters

Returns


Get Subtotal By integer Statistics categorized by service or by operation.


Name string Name of the service or operation.

Total integer Total number of objects processed for this schema for a publication service.

Total number of objects received for this schema for a subscription service.

Success integer The number of objects that were successfully identified for this schema which will be published or written to a file.

Failure integer The number of objects that were identified for this schema but were not published because the header of the schema failed validation for the publication service, or was written to a file because the schema was not associated with the subscriber for a subscription service.


getActivityStatisticsByService | 259

getActivityStatisticsByService

Purpose (Custom with adb.perfMon) Returns the total number of objects processed for each of the schemas associated with the specified service. Also, returns the number of success and error objects.

Parameters

Returns


Service Name string Name of the service.


Operation string Type of operation that the service performs.

Schema Name string Name of the schema that is associated with the service.

Total integer Number of objects processed for this schema for a publication service.

Number of objects received for this schema for a subscription service.





getActivityStatisticsByOperation()

Purpose (Custom with adb.perfMon) Returns the total number of objects processed for all the schemas by each service that is associated with a specified operation. Also, returns the number of success and error objects.

Parameters

Returns


Operation string Name of operation. Pick from list.


Service Name string Name of the service that is associated with the specified operation.

Total integer Total number of objects processed for this schema for a publication service.

Total number of objects received for this schema for a subscription service.




getThreadStatistics() | 261

getThreadStatistics()

Purpose (Custom with adb.perfMon) Return the operation counts of the current threads.


ThreadID string A unique identification of a particular thread.

ThreadType string A type or key that will match this thread to a queue or connection.

TaskType string Short description of the tasks this thread processes.

TaskCount integer Number of tasks processed by this thread.



getQueueStatistics()

Purpose (Custom with adb.perfMon) Return the current count of elements in any internal queue used by the adapter. This includes the TIBCO Rendezvous event queues automatically spawned by TIBCO Rendezvous for each adapter.


QueueID string A unique identification of a particular queue.

QueueType string A type or key that will match this queue to a thread or connection.

QueueCount integer Current number of elements in the queue.

MaxQueueSize integer Maximum number of elements in the queue.


getConnectionStatistics() | 263

getConnectionStatistics()

Purpose (Custom with adb.perfMon) Returns the state and statistics for all the current connections used by the adapter.


Connection ID string A unique identification of a particular connection.

Connection Type string A type or key that will match this connection to a thread or queue.

State string Current state: CONNECTED or DISCONNECTED

NumRetries integer Total number of times this connection had to be reestablished.

TotalNumOperations

integer Total number of operations processed by this connection since the adapter started.

CurrentNumOperations

integer Total number of operations processed by this connection since the last reconnection.



resetActivityStatistics()

Purpose (Custom with adb.perfMon) Resets all the counts for the activity statistics.


resetThreadStatistics() | 265

resetThreadStatistics()

Purpose (Custom with adb.perfMon) Resets all the counts for the thread statistics.



resetQueueStatistics()

Purpose (Custom with adb.perfMon) Resets all the counts for the queue statistics.


resetConnectionStatistics() | 267

resetConnectionStatistics()

Purpose (Custom with adb.perfMon) Resets all the counts for the connection statistics.




| 269

Appendix A Using TIBCO ActiveMatrix Adapter Service Engine for Database

Topics


• Installation, page 271

• Configuration, page 275

• Deployment, page 280


270 | Appendix A Using TIBCO ActiveMatrix Adapter Service Engine for Database

Overview

This section describes the TIBCO ActiveMatrix Adapter Service Engine for Database (Adapter SE) component. This component enables the adapter to run in the ActiveMatrix environment as a container and to provide services. The Adapter SE supports both the AE and ActiveMatrix endpoints. Adapter AE endpoints, which are adapter services configured with RV or JMS transports, can be converted as ActiveMatrix endpoints that let the Adapter SE communicate seamlessly with other ActiveMatrix services. All adapter endpoint types are converted to WSDLs and can be used for configuration in the ActiveMatrix environment.

The Adapter SE supports communication with TIBCO ActiveMatrix BusinessWorks and TIBCO ActiveMatrix BusinessWorks Service Engine using only the AE transport.

This section describes how to install the Adapter SE component and the steps involved in importing your Designer-based adapter projects into Business Studio and configuring for use in the ActiveMatrix environment.

To get familiar with the ActiveMatrix concepts and the ActiveMatrix development tool, TIBCO Business Studio, refer TIBCO ActiveMatrix Concepts and TIBCO ActiveMatrix Composite Editor User’s Guide.


Installation | 271

Installation

This section describes the installation process for Adapter SE.

Required Software

The following software and their dependant software must be installed before installing Adapter SE:

• TIBCO ActiveMatrix Adapter for Database

• TIBCO ActiveMatrix Foundation

Supported Platforms

Adapter SE is supported on the following platforms:

• Microsoft Windows 2003 Server, Microsoft Windows XP Professional (x86)

• Solaris 10 (SPARC)

• RedHat Enterprise Linux 4 (x86)

• SUSE Enterprise Server Linux 10.1 (x86)

• HP-UX 11i v3 (IA64 Itanium)

Installation Packages

Make sure the installation package matches your ActiveMatrix installation environment.

The following table lists the available installation packages.

Platform Package Names Hardware Platform

Microsoft Windows XP Professional

Microsoft Windows Server 2003

TIB_adadbse_5.5.0_win_x86.zip x86

Solaris TIB_adadbse_5.5.0_sol_sparc.zip SPARC

HP-UX TIB_adadbse_5.5.0_hpux_ia64.zip Itanium

Linux TIB_adadbse_5.5.0_linux_x86.zip x86



Installation Steps

Install the Adapter SE product using one of the modes described in the sections:

• Installing in GUI Mode on page 272

• Installing in Console Mode on page 272

• Installing in Silent Mode on page 273

Installing in GUI ModeTo install Adapter SE in the GUI mode:

1. Open the physical media or download the Adapter SE package.

2. Extract the Adapter SE archive file to a temporary directory.

3. Navigate to the temporary directory that contains the universal installer.

4. Run TIBCOUniversalInstaller.

5. The Welcome screen appears. Click Next.

6. Read and accept the license agreement. Click Next.

7. The Destination Info screen appears. TIBCO_HOME was defined when the pre-requisite software was installed. Click Next.

8. Install all features by clicking Typical or choose the features to install by clicking Custom. After making your choice, click Next.

9. If you selected Custom, in step 8, uncheck the checkbox next to the component you don’t want installed and click Next.

If you selected Typical in step 8, click Next.

10. The installer prepares the components for installation. A pre-install summary screen appears. Click Install.

11. Read the post-install summary and click Finish.

Installing in Console ModeTo install the software in console mode:




Installation Steps | 273

3. Using a console window, navigate to the temporary directory that contains the universal installer.

4. Run TIBCOUniversalInstaller -console.

5. Complete the installation by responding to the console window prompts.

Installing in Silent ModeThe following procedure explains how to install a Adapter SE product in silent mode. The TIBCOUniversalInstaller.silent file is packaged in the directory that contains the universal installer. You must edit the file with information for your environment before launching the silent installation. The file includes comments that describe the installation properties you can set.

While you can use the TIBCOUniversalInstaller.silent file, it’s good practice to copy the file to a different name and use that file for the silent install.

If errors occur during installation, they will be listed in the installation log file (see the User_Home/.TIBCO directory).



3. Using a console window, navigate to the temporary directory that contains the universal installer.

4. Copy the TIBCOUniversalInstaller.silent file and name the file.

5. Using a text editor, open the copied file and update the install location, and features to install.

6. Run: TIBCOUniversalInstaller -silent -V responseFile="myfilename.silent"

If you are using the TIBCOUniversalInstaller.silent file, you need not supply the file name and can run: TIBCOUniversalInstaller -silent.

A line similar to the following is written to the installer log file when installation completes:Install, com.tibco.installer.util.TIBCOInstaller, dbg.Debug, The installation has completed. Please check the log file for additional information.

Uninstallation Before uninstalling the Adapter SE component, make sure you deactivate the container for the node where the Adapter SE component is installed.

Follow these steps to uninstall the Adapter SE component:



1. Navigate to TIBCO_HOME\_uninstall directory and invoke universal_uninstall.

2. Choose the Custom Uninstall option

3. Check the checkbox for the Adapter SE component and uncheck the checkboxes for the other listed products.


Configuration | 275

Configuration

This section describes the steps involved in importing and configuring a Designer project into the ActiveMatrix environment using TIBCO Business Studio.

Import a Designer Project

To import files for your Designer project:

1. Launch TIBCO Business Studio.

2. Select File > New > Project> General > Project.

3. Type a name in the Project name field and click Finish.

4. In the Project Explorer view, right click the folder and select Import....

5. Expand the General node and click File System.

6. Click Next.

7. To the right of the From directory field, type a directory name or click

Browse... to locate a directory from which to import the files. The Import from

directory dialog displays.

a. Browse to and choose a directory. This should be the root directory of your Designer project.

b. Click OK.

8. In the panel on the right, check the checkboxes next to the files you want to import.

9. Click Finish.

An additional project, this time an ActiveMatrix SOA project should be created. This project will contain files that for the corresponding ActiveMatrix project.

Create an ActiveMatrix SOA Project

To create a new ActiveMatrix SOA project:

10. Click File -> New > Project>TIBCO ActiveMatrix> ActiveMatrix SOA Project.

11. Type a name for the project and click Finish.



Generate the WSDL and Deployment Artifacts

To generate the WSDL file:

12. Right click the adapter configuration file and select Generate Adapter WSDL.

13. Choose the project created in step 10 and click OK.

The following folders are created:

• adapter configuration filename_resources, and

This folder contains:

adapter configuration filename_number.dat This file contains information used by the Adapter container.

adapter configuration filename.adendpoints The adapter service endpoints file contains the services supported by the adapter.

adapter configuration filename.wsdl The generated WSDL file which references the schema and .dat files contains the service endpoints.

adapter configuration filename.substvar The substitution variables are imported from the adapter configuration file.

The shared resource that defines the transport used to communicate with the adapter is created. If the Adapter project referenced global variables, the shared resource will reference substitution variables defined in adapter configuration filename_number.substvar.

• schema

This folder contains the Adapter XSD schema.

14. Double click the adapter configuration filename.substvar file and add the variables InstanceId and AppName of type String with any value.

Adding Composite Elements

15. Double click the adapter configuration filename.composite file that is created.

The Palette displays the Adapter component.

16. Drag and drop the adapter component into the canvas.

17. Type in a name for the component.

18. Select the Services tab and click Add.

19. Expand the tree displayed in the Resource Picker dialog box and choose a service and click Add.

20. Click OK.


Configuration | 277

The following table lists the fields displayed in the Component tab

Add Services Elements

21. Drag and drop a service, say SOAP, from the palette onto the canvas.

22. Give it a name.

23. Save the project.

Add Shared Resources

In order for the service elements created in the previous step to be accessible from external client, an HTTP connection resource has to be created.

24. In the Project Explorer panel, right click Shared Resources and choose New > HTTP Server.

25. Type in a name for the shared resource or accept the default.

To add this shared resource to the composite:

26. Open the composite by double clicking the composite in the Project Explorer.

Property Names Description

Name The Adapter component name.

Description The Adapter component description.

Version The Adapter component version number.

Type (Read-only) The Adapter type.

SDK Application Name (Read-only) The instance name of your Adapter configuration.

Configuration URI (Read-only) The Adapter configuration that the service uses.

Repository URI (Read-only) This file contains information used by the container to communicate with the Adapter.

TRA File Name This file contains the value for the application.ase.library or application.library properties that points to the dynamic library of the Adapter Service Engine component.

Skip the steps involved in the Add Services Elements and Add Shared Resources sections if you do not want to convert your existing AE endpoints to ActiveMatrix endpoints.



27. Click the canvas.

28. Click the Properties tab below the canvas.

29. Click on the field. A row is created in the table.

30. Type in a name or accept the default.

31. In the type column, choose HTTP.

32. In the Target column, click to bring up the Select Shared Resource dialog.

33. Choose the shared resource and click OK.

34. Open the service and from the Properties view click the Target tab.

35. In the table, click the radio button next to the component service.

A wire will be drawn between the service and the component service.

Creating the Service Assembly

In order to deploy the services and components that were created in previous steps, they have to be packaged in a service assembly.

36. In the Project Explorer panel, right click the composite and choose Service Assembly.

The deployment package is created.

37. Right click the deployment package and choose Refresh Service Assembly.

38. Right click the Service Assembly and choose Build Archive.

This service assembly archive can now be deployed and started using TIBCO ActiveMatrix Administrator.

Using Substitution VariablesAt design-time, you can assign default values to all the custom properties (substitution variables) from the .tra file using TIBCO Business Studio:


Configuration | 279

1. In Project Explore panel>adapter configuration filename_resources folder, find the substitution variable file adapter configuration filename.substvar.

2. Double click the .substvar file to bring up the Defined Substitution Variable editor.

3. Modify the default properties for the substitution variables.

See TIBCO ActiveMatrix Composite Editor User’s Guide for details on working with substitution variables.

At run-time, you can update the values of the substitution variables using TIBCO ActiveMatrix Administrator.

The following figure shows the Substitution Variables editor in TIBCO ActiveMatrix Administrator, where you can specify runtime values for substitution variables.

In In the ActiveMatrix environment, values for the adapter properties are assigned in TIBCO ActiveMatrix Administrator as Substitution Variables.

Do not assign values for the adapter properties in your.tra file.



Deployment

The service assembly created in the previous section is deployed using TIBCO ActiveMatrix Administrator.

Refer to TIBCO ActiveMatrix Administration for detailed information on deploying service assemblies.

Deployment Considerations

• Before starting the node on UNIX platforms, the required environment variables for the system path have to be set by running the scripts packaged with the adapters. The node can then be started from either of the following ways:

— Start the node from the command line: Run the scripts packaged with the adapters to make sure the proper environment variables are set. Start the node on which the adapter service unit will be deployed.

— Start the node using TIBCO ActiveMatrix Administrator: Before starting the TIBCO Management Daemon, run the scripts packaged with the adapter to make sure the proper environment variables are set. Start the node using TIBCO ActiveMatrix Administrator.

• The values for the following properties have to be specified.

— application.ase.library: This parameter specifies the name of the dynamic library of the adapter service engine instance. This property is available only on UNIX platforms.

— adb.nodupdetection: This parameter disables detection of duplicate configurations. Detection of duplicate configurations has to be disabled in order to run the adapter in the TIBCO ActiveMatrix load balanced nodes.


| 281

Appendix B Frequently Asked Questions

This appendix lists answers to frequently asked questions.

Topics

• General Questions, page 282

• Request-Response Questions, page 288


282 | Appendix B Frequently Asked Questions

General Questions

How can I determine the source of a problem or an error?

In some cases it is helpful to turn on ODBC tracing. Activate tracing in Control Panel>ODBC Data Sources on the Tracing tab. For details, see your Microsoft Windows documentation.

How can I find the version number of an adapter instance?

A banner displays when an adapter instance starts. The banner lists component versions for the adapter and for TIBCO Adapter SDK software. You can use this information to diagnose compatibility issues, or to report any problem details to Customer Support.

You can also display version information in TIBCO Designer by clicking the Help>Runtime Environment menu command.

Why is a database trigger error not logged in the exception table?

When using an adapter instance as a publisher, if an error occurs in the database trigger that is used to copy data from the source table to the publishing table, the database trigger error will not be logged in the exception table for the subscriber adapter.

How should the adapter react if the database connection is lost and the database is later restarted? Does it automatically try to reconnect?

If TIBCO ActiveMatrix Adapter for Database detects it has lost its database connection, it shuts down. You can configure the adapter for automatic reconnection (see Run-time Connection Tab on page 51). Alternatively, TIBCO Hawk rules can be written to detect this and restart the adapter whenever this occurs.

Must an incoming message contain all the columns that are defined for the destination table?

The incoming message need not contain all the columns defined in the destination database table. You can configure the adapter to expect only a subset of the columns, defined in the repository. The adapter is driven from the subscribing class description and will iterate through the attributes in the class definition for the subscribing table and specifically look for those attributes in the incoming


General Questions | 283

messages. It inserts NULLs for the attributes that it is expecting but does not find in the message. If there are more columns in the subscribing table than are listed in the subscribing class (set when adding a subscription), those extra columns will get whatever default values were specified during the table creation.

Can an existing table be used as the publishing table?

No. TIBCO ActiveMatrix Adapter for Database requires additional columns in the publishing table. Even when every field in a table is published, a separate publishing table is required.

Does the TIBCO Rendezvous message that is published have to contain every field in the publishing table?

Yes. You can control which fields are copied to the publishing table by configuring the adapter and by changing the publication trigger to publish a subset of rows. You can also append additional fields to a message or drop a message based on some criteria using the user callout library. For more information on the user callout library, see User Callout Library on page 187.

If multiple updates occur between polling intervals, are updates published in multiple TIBCO Rendezvous messages or a single large message?

If you are using publish by value, a TIBCO Rendezvous message is created for each individual update. If you are using publish by reference, that operation will get the last update.

Is it possible to delete older entries in the publishing table?

Yes. When a row is published, the value of the ADB_L_DELIVERY_STATUS field in the publishing table changes to either C (complete) or F (failed). You can write a trigger in your publishing table that deletes the row when the delivery status changes to C or F.

You can also publish data directly from the source table by configuring the adapter instance to publish by reference. A publishing table is created, but it contains only required fields and key fields of the source table.



How does the exception table work?

Before starting an adapter instance, you must set the adb.useExceptTable option in the adapter’s properties file to on and specify an exception table when configuring the adapter instance. If an error occurs when inserting data into the destination table, it will be inserted into the exception table. The transaction will be committed and a confirmation sent back for the message (RVCM delivery). If the insertion into the exception table also fails, an error message will display and the adapter instance will terminate.

Can an adapter instance be used to replicate binary types, such as BLOB?

There is only limited support for binary large object (BLOB) data types. Oracle LONG and LONG RAW types are supported in top-level tables when the adapter is configured to publish by reference. Oracle BLOB and CLOB data types are supported.

Can an adapter instance write to tables that belong to a database account different from that used by the adapter?

Yes. A source table or destination table can belong to a different database user than the default account created in create_user.sql. For more information, see Referencing External Schemas on page 10.

Can a publisher adapter and a subscriber adapter use different projects?

Yes, unless the publisher adapter is configured to use parent-child relationships.

Can TIBCO ActiveMatrix Adapter for Database guarantee exactly once delivery of messages over RVCM?

Exactly once delivery of messages over RVCM is not currently guaranteed. The same quality of service that RVCM provides is supported, which is at least once. To get exactly once delivery requires combining the messaging operations and the database operations in a single atomic transaction, which is not supported in RVCM.

What guarantees does the adbagent make with regards to the order of database operations? For instance, is it guaranteed that for a given table, modifications are made in the same order that they were made to the source database? What guarantees are made for operations across



different tables?

TIBCO ActiveMatrix Adapter for Database guarantees that for database operations that are published on the same subject, the order of the operations is preserved. Usually, this applies to database operations made to one table. It does not usually apply to database operations across different tables.

When using RVCM for delivery, at what point does the subscribing adapter acknowledge an incoming message?

A TIBCO ActiveMatrix Adapter for Database subscriber confirms the message only after the database operation is committed. If there is an error and no exception table is used, the database operation is rolled back and no confirmation is sent. If there is an error and an exception table is used, the insert to the exception table is committed and the message is then confirmed.

By maintaining the publishing tables, all changes to the source table can be captured. If, however, there is a failure between the point of publishing a message using RVCM and updating the publishing table, will the adapter republish the message that has already been sent?

Yes, the message will be republished and the subscriber would have to deal with the duplicate message.

For the subscriber adapter, if failure occurs after doing a database update and before sending an acknowledgement back to the publisher adapter, can the restarted configuration be prevented from redoing the update operation?

No. This would cause a duplicate insert.

For a certified subscriber adapter, if an insert exception occurs and no exception table is specified, what happens?

If the tibco.clientVar.DirTrace option is specified in the adapter’s properties file when the adapter is started, exception handling information is written to the log file and the configuration continues to run. Since the insert could not be performed, the ADB_L_DELIVERY_STATUS publishing table column has a value of P for the message.

How do I pre-register non-TIBCO ActiveMatrix Adapter for Database subscribers, like custom adapters, to ensure no messages are lost?

Specify the CM name of the listener’s RVCM session

The session names are automatically created. Can these be changed, without confusing TIBCO



ActiveMatrix Adapter for Database, so that we can use a standard naming convention throughout the entire integration process?

No, the session names are fixed and used by TIBCO ActiveMatrix Adapter for Database internally. They cannot be altered.

What’s the proper way to permanently remove an adbagent subscriber when using CM? One way is to completely REMOVE the publisher’s ledger and to change all ’P’ records back to ’N’, then restart the publisher. Is there a better, more correct or automated method?

There is a TIBCO Hawk method, unRegisterListener(), which unregisters a CM subscription. This is the proper way to remove the adbagent subscriber as a CM listener.

Is it possible to run an adapter instance using a remote TIBCO Rendezvous daemon?

Yes. Change the default settings for network, service, and daemon parameters for the adapter using TIBCO Designer.

Is it possible to run two configurations of the adapter on the same machine?

It's possible to run multiple configurations of the adapter on the same machine if each adapter instance has a unique name. If both configurations use TIBCO Rendezvous certified messaging, each must use a different RVCM session.

Can an adapter instance collate information from several database tables to send as a single TIBCO Rendezvous message, or can it only publish data from a single table, in the format defined by that table?

There are two ways to publish related tables:

• Set the adb.publishChildData option to on when configuring the adapter’s properties file. When there are insertions into parent table, the adapter will publish parent rows and the corresponding child rows using TIBCO ActiveEnterprise or XML format.

Note that the adapter currently does not support publishing child data in the TIBCO Rendezvous Message format. Update and delete on parent-child relationship when publishing is also not supported. See Publisher Options Tab on page 81.

• Combine several tables into one table using a trigger and then publish from the combined table.

When publishing or subscribing, an adapter instance allows you to change the message using the callout library. See User Callout Library on page 187 for details.



Why does TIBCO Designer display a Java exception error while I’m trying to use TIBCO Designer through Exceed?

When using Exceed to simulate an X Windows environment, start the X Windows server in the Exceed Session Startup Application. In this application, make sure the Run X server checkbox is selected, then try TIBCO Designer again.

When my Sybase transaction log becomes full, the adapter hangs. How can I fix this problem?

When a Sybase transaction log becomes full, if the database setting abort trans on log full is set to false, your application will hang instead of printing a transaction log full error. Execute the following command:

sp_dboption dbname, "abort tran on log full", true



Request-Response Questions

When using request-response, can an INSERT statement with values only (without field names) be sent to the improve application's performance.?

Yes this is allowed. Your application can also send INSERT statements without the binds.

Can an application send UPDATE statements to a subscriber adapter with only those fields which are being updated? That is, if a table has ten records and only two should be updated, can an UPDATE statement be constructed and sent only for those fields?

Yes, this is supported.

Does an adapter instance send responses back as one large message with all rows in it or is the message sent in chunks?

The adapter sends results back to an application as one large message.

If a failure occurs when doing an insert or update, what is returned to the application?

If an error occurs while the adapter is processing a request, an error code and description is returned to the application. In the case of success, a result set and row count is returned to the application.


| 289

Appendix C Trace Messages

This appendix explains the trace messages in TIBCO ActiveMatrix Adapter for Database. These include status messages, informational messages, error messages, and other types of messages.

Topics


• Status Messages, page 295


290 | Appendix C Trace Messages

Overview

Trace messages provide information about adapter activities. The messages are logged to the console where the runtime adapter was started and to a log file in a location specified during configuration. Trace messages can also be redirected to the TIBCO Hawk Display application, or sent to other applications using the TIBCO Rendezvous transport.

Each trace message can include the following fields in the order shown:

Timestamp Adapter_Identifier Role Category Status_Code Tracking_Identifier Application_Information

The above fields are explained in Trace Message Fields on page 291. The following example shows a trace message containing some of these fields and then identifies the fields.

Example Trace Messages

The following trace messages were written during a session where TIBCO Adapter for Files received an object from TIBCO Adapter for R/3, then processed the object.

Example 1:Adapter Started

The following message indicates that TIBCO Adapter for Files has started. The timestamp indicates when the adapter started, and the role indicates that the trace message is informational, which means the activity is normal for the adapter. The category is identified, and the corresponding status code is displayed. The status code indicates that the adapter started successfully.

2003 Feb 22 20:14:51:718 GMT -8 FileAdapter.FileAdapterConfiguration Info [Configuration] AEFA-000058 TIBCO Adapter for Files successfully initialized

2003 Feb 22 20:15:12:937 GMT -8 FileAdapter.FileAdapterConfiguration Info [Adapter] AEFA-000067 Message containing class /tibco/public/class/ae/Customer received on subject FROM.SAP tracking=#MU3oTJ/WWCV1MU96J0zzwA9kzzw#

2003 Feb 22 20:15:12:937 GMT -8

FileAdapter.FileAdapterConfiguration

Info

[Adapter]

AEFA-000067 Message containing class /tibco/public/class/ae/Customer received on subject FROM.SAP

tracking=#MU3oTJ/WWCV1MU96J0zzwA9kzzw#

Timestamp:

Adapter ID:

Role:

Category:

Status Code:

Tracking ID:


Overview | 291

Example 2:MessageReceived

The next set of trace messages indicates the adapter received an object that was sent on the TIBCO Rendezvous subject, FROM.SAP. The #MU3oTJ/WWCV1MU96J0zzwA9kzzw# tracking identifier included in the trace message uniquely identifies the message. The adapter (TIBCO Adapter for R/3) from which the message originated provided the identifier.

2003 Feb 22 20:15:12:937 GMT -8 FileAdapter.FileAdapterConfiguration Info [Adapter] AEFA-000067 Message containing class /tibco/public/class/ae/Customer received on subject FROM.SAP tracking=#MU3oTJ/WWCV1MU96J0zzwA9kzzw#

2003 Feb 22 20:15:12:937 GMT -8 FileAdapter.FileAdapterConfiguration Info [Adapter] AEFA-000068 Message containing class /tibco/public/class/ae/Customer written to working file customers.txt in Working Directory F:\ca\integration\001\data_sets\files\wip tracking=#MU3oTJ/WWCV1MU96J0zzwA9kzzw#

Example 3:Object Moved

The final trace message indicates the object has been moved to the output directory, which completes the adapter’s interaction with the object. Because the trace message is the termination point, the tracking identifier is not displayed.

2003 Feb 22 20:15:42:812 GMT -8 FileAdapter.FileAdapterConfiguration Info [Adapter] AEFA-000070 File customers.txt is moved to the Output Directory F:\ca\integration\001\data_sets\files\solutions\output

Trace Message FieldsEach trace message includes the following fields:

Table 18 Trace Message Fields

Field Name Description

Timestamp Timestamp of occurrence. For example, 2003 Feb 22 20:14:51:718 GMT -8.

Adapter Identifier

Name of the adapter that wrote the trace message. This is a combination of the adapter acronym and adapter configuration name. For example, the application identifier, ADB.publisher1 identifies a TIBCO ActiveMatrix Adapter for Database service named publisher1.



Role A role can be:

• Info. Indicates normal adapter operation. No action is necessary. A tracing message tagged with Info indicates that a significant processing step was reached and has been logged for tracking or auditing purposes. Only info messages preceding a tracking identifier are considered significant steps.

• Warn. An abnormal condition was found. Processing will continue, but special attention from an administrator is recommended.

• Error. An unrecoverable error occurred. Depending on the error severity, the adapter may continue with the next operation or may stop altogether.

• Debug. A developer-defined tracing message. In normal operating conditions, debug messages should not display.

When configuring the adapter you define what roles should or should not be logged. For example, you may decide not to log Info roles to increase performance.

Table 18 Trace Message Fields (Cont’d)

Field Name Description (Cont’d)


Overview | 293

Category One of the following:

• Adapter. The adapter is processing an event.

• Application. The adapter is interacting with the database.

• Configuration. The adapter is reading configuration information.

• Database. The adapter is interacting with a database.

• Metadata. The adapter is retrieving metadata from the database.

• Palette. The adapter is interacting with the palette.

• Publisher Service. The publication service is reporting this trace message.

• Request-Response Client Service. The request-response invocation service is reporting this trace message.

• Request-Response Server. The request-response service is reporting this trace message.

• Shutdown. The adapter is shutting down.

• Startup. The adapter is starting.

• Subscription Service. The subscription service is reporting this trace message.

• System. This category is not linked to a specific event process. The trace message may be related to a Windows service related messages, memory allocation, file system error, and so on.

• TibRvComm. The adapter is communicating with TIBCO Rendezvous.

• XML. The adapter is parsing XML documents.

Status Code Unique code for the message and description. Status codes are identified by a unique number and description. If a trace message includes an error or warn role, the status code documentation includes a resolution. See Status Messages on page 295 for details.





Tracking Identifier

A unique identifier that is "stamped" on each message by the originating adapter. The tracking identifier remains in effect from a message’s beginning to its completion as it is exchanged by TIBCO applications. If the adapter is the termination point of the message, the tracking identifier is not displayed in the trace message.

You cannot modify the tracking identifier format or configure what information is displayed.

Application Information

Application-specific information added to the tracking info to trace the message back to its source. Set initially by the originating adapter and carried forward. It is augmented by each intermediate component.




Status Messages | 295

Status Messages

Status Code Role Category Resolution

AEADB-100001 %1Database connection succeeded.

infoRole Database Normal operation; no action is necessary.

AEADB-100002 %1Database connection failed.

errorRole Database Check and test the ODBC configuration and restart the TIBCO Runtime Agent.

AEADB-100003 odbc connection as: DSN: %1, User: %2


AEADB-100004 Database driver code: %1Database vendor message: %2

errorRole Database Refer to an ODBC manual and to the manual of the native database by using the error code.

AEADB-100005 Unknown database system.

warnRole Database The adapter is unable to detect the DBMS type of the database it connects to. Make sure a supported DBMS is used. This condition sometimes occurs because a native character set is used. If you are using a supported DBMS, ignore this warning.

AEADB-100006 %1SQL statement preparation succeeded.


AEADB-100007 %1SQL statement preparation failed.

errorRole Database Preparation of a SQL statement failed. Check the configuration for this adapter service and restart the adapter.



AEADB-100008 Problem reading values for %1 from Database

errorRole Database The adapter cannot retrieve data from tables. In the adapter properties file, set the adb.verbose option to on and the adb.debug option to 3 to print the SQL statement to the log file then report this error to TIBCO support.

AEADB-100009 Database Connection Lost, Terminating....

errorRole Database Check the connection to the underlying database and restart the adapter.

AEADB-100010 jdbc connection as : Driver %1, URL %2, User %3


AEADB-100011 %1Database connection closed


AEADB-100012 %1Invalid statement

errorRole Database The adapter detected an invalid SQL statement. For request-response service, an invalid request is processed; otherwise, check your adapter service configuration and restart the adapter instance.

AEADB-100013 Attribute %1 (class %2): No column data ignoring


AEADB-100014 Inserted %1into database

infoRole Adapter Normal operation; no action is necessary.

AEADB-100015 Deleted from database


AEADB-100016 No rows deleted


Status Code Role Category Resolution (Cont’d)



AEADB-100017 Updated database


AEADB-100018 No rows updated


AEADB-100019 No columns specified for WHERE clause in %1

errorRole Adapter The adapter could not find key fields to construct the WHERE clause. Check the repository definition of this class and make sure there is at least one field with isKey=true.

AEADB-100020 Building %1 criteria


AEADB-200001 Foreign key attribute %1 not found in the publishing class.

errorRole Configuration Check your parent-child relationship in publisher configuration and make sure that the child table is joined with the parent table by a common key column, then restart the adapter.

AEADB-200002 %1Class registry not found: %2.

errorRole Configuration Check the configuration of that service to make sure that the class registry exists.

AEADB-200003 %1Class description not found: %2.

errorRole Configuration Check the configuration of that service to make sure that the class description exists. Sometimes, if the loadUrl is not set correctly, this problem occurs.

AEADB-200004 Problem encountered when reading from ADB_PREREGLISTENER

errorRole Configuration The adapter has problems reading from the ADB_PREREGLISTENER table. Check to see that access to that table is allowed and the table is not corrupted.




AEADB-200005 Could not register Hawk method %1

warnRole Configuration A description of a TIBCO Hawk method was not found in the repository. This can happen if the adapter instance is not the latest. Configure the adapter with the latest TIBCO Designer version.

AEADB-200006 Setting %1 = %2.

infoRole Configuration Normal operation; no action is necessary.

AEADB-200007 %1 cannot be used with %2

warnRole Configuration Incomplete options or conflicting options are used. Check the log file for the detailed information and restart the adapter with correct options.

AEADB-200008 Preregistering %1 on %2

infoRole Configuration Normal operation; no action is necessary.

AEADB-200009 %1 creation failed

errorRole Configuration There was an error initializing the publication. Check your project and re-configure the publication.

AEADB-200010 %1 cannot be larger than %2

errorRole Configuration Conflicting options are used. Check the log file and restart the adapter with the correct options.

AEADB-200011 Unsupported encoding type: %1, Use default ASCII Encoding.

errorRole Configuration An unknown character set is detected. Check the language configuration in the OS environment and restart the adapter.




AEADB-200012 AEADB-POLLINGINTERVAL-NOVALUE-ERR

errorRole Configuration Unable to retrieve polling interval value from Hawk method. Retaining original value of %1 milliseconds. Please make sure a value is specified before invoking the setPollingInterval method.

AEADB-200013 AEADB-POLLINGINTERVAL-NEGATIVE-ERR

errorRole Configuration Negative value of %1 for polling interval not allowed. Retaining original value of %2 milliseconds. Please make sure to specify a non-negative value for the polling interval.

AEADB-200014 AEADB-POLLINGINTERVAL-SAME-WARN

warnRole Configuration Polling interval already set to %1 milliseconds.

AEADB-200015 AEADB-POLLINGINTERVAL-CHANGED

infoRole Configuration Polling interval changed to %1 milliseconds.

AEADB-200016 AEADB-POLLINGBATCHSIZE-NOVALUE-ERR

errorRole Configuration Unable to retrieve polling batch size value from Hawk method. Retaining original value of %1. Please make sure a value is specified before invoking the setPollingBatchSize method.

AEADB-200017 AEADB-POLLINGBATCHSIZE-NEGATIVE-ERR

errorRole Configuration Negative value of %1 for polling batch size not allowed. Retaining original value of %2. Please make sure to specify a non-negative value for the polling batch size.

AEADB-200018 AEADB-POLLINGBATCHSIZE-SAME-WARN

warnRole Configuration Polling batch size already set to %1.

AEADB-200019 AEADB-POLLINGBATCHSIZE-CHANGED

infoRole Configuration Polling batch size changed to %1.




AEADB-200020 AEADB-POLLINGBATCHSIZE-CHANGED

warnRole Configuration The repository encoding should be set to UTF-8 if your agent encoding is not ASCII or LATIN-1. Set the repository encoding to UTF-8.

AEADB-300001 %1exception thrown in %2.

errorRole System An adapter internal error. Contact TIBCO Support.

AEADB-300002 Not enough memory available.

errorRole System Check the virtual memory of the system and restart the adapter. If it keeps happening, contact TIBCO Support.

AEADB-300003 MException throw in %1: %2.

errorRole System An Adapter SDK exception was thrown. Report the SDK error code and text to TIBCO Support.

AEADB-300004 %1 on connect socket in thread failed

errorRole System A TCP system call failed. Check your computer’s network ports and TCP libraries. Rebooting the computer may solve the problem.

AEADB-300005 Binding address to socket failed. Listenport = %1


AEADB-300006 Listen call on socket failed

errorRole System A TCP system call failed. Check your computer’s network ports and TCP libraries. Rebooting the machine may solve the problem.

AEADB-300007 Accept on socket failed





AEADB-300008 Socket creation in thread failed


AEADB-300009 Out of memory building the WHERE clause

errorRole Adapter An adapter process has run out of memory. Increase the amount of memory allowed to the adapter on your system.

AEADB-400001 %1Unknown column name %2

errorRole Metadata Request is missing a column field in the Bind data. Add a column, tableName.columnName field, to your bind data.

AEADB-400002 %1Unknown column type: %2 for field: %3.

errorRole Metadata The type %2 is an invalid AE class type. Check in the repository for attribute %3 and change the class type to a valid AE class type.

AEADB-400003 Unknown column type, %1, %2

infoRole Metadata Normal operation; no action is necessary.

AEADB-400004 %1schema %2, tables: %3

debugRole Metadata

AEADB-400005 %1table %2, columns: %3

debugRole Metadata

AEADB-400006 Ill-formatted table name %1

errorRole Metadata A table name of this format is not supported by the adapter. Use a different table name.




AEADB-600001 %1Failed to set limit policy for event queue %2 on session %3: %4

errorRole TibRvComm TIBCO Rendezvous failed to set the limit policy according to the -rv-max-queue-size specification. Please report this error text to TIBCO customer support.

AEADB-600002 %1Failed to get queue count for event queue %2 on session %3: %4

errorRole TibRvComm TIBCO Rendezvous failed to get the queue count for the given session. Please report this error text to TIBCO customer support.

AEADB-600003 %1Failed to get queue limit for event queue %2 on session %3: %4

errorRole TibRvComm TIBCO Rendezvous failed to get the limit policy for the given session. Report the this error text to TIBCO Support.

AEADB-600004 OnEvent: eventname %1

infoRole TibRvComm Normal operation; no action is necessary.

AEADB-600005 MException thrown trying to publish message on "%1".

errorRole TibRvComm Unable to publish the message, check error description for details. Check your daemon connection.

AEADB-600006 %1Received ADVISORY %2

errorRole TibRvComm An unexpected TIBCO Rendezvous advisory was received. Refer to your TIBCO Rendezvous documentation for resolution information.

AEADB-700001 SQL: %1


AEADB-700002 %1Building SQL statement succeeded%2.





AEADB-700003 %1Binding data to placeholders: unknown column type, %1 for field %2. Ignoring

errorRole Adapter The type %1 is an invalid AE class type. Check in the repository for attribute %2 and change the class type to a valid AE class type.

AEADB-700004 %1database operation%2 succeeded


AEADB-700005 %1database operation%2 failed

errorRole Adapter The execution of this database statement failed. Use the error returned by the database to determine how to allow this SQL statement to succeed.

AEADB-700006 %1database transaction begin


AEADB-700007 %1database transaction commit


AEADB-700008 %1database transaction rollback.


AEADB-700009 %1Cannot bulk insert for variable binary type in column %2.%3.

warnRole Adapter The subscriber bulk insert option does not support a variable binary data type. Do not use bulk insert for tables with a variable binary data type.

AEADB-700010 %1Database transaction commit failed.

warnRole Adapter The database was unable to commit the transaction. Use the error returned by the database to determine how to allow this transaction to succeed.




AEADB-700011 Encoding error from %1 to %2.

errorRole Adapter An Adapter SDK error converting data to the specified encoding. Make sure the encoding is a supported one.

AEADB-700012 %1Cold start completed.


AEADB-700013 %1Cold start failed.

errorRole Adapter Run the adapter using debug 3 and verbose mode.

AEADB-700014 %1Confirming message.


AEADB-700015 %1 batch commit timeout.


AEADB-700016 %1Bulk insert begins.


AEADB-700017 %1Bulk insert ends.


AEADB-700018 Inserted %1into the corresponding exception table.


AEADB-700019 %1Exception table is not specified.

warnRole Adapter Reconfigure the subscriber by specifying an exception table, or turn off the adb.useExceptTable option (in the adapter properties file) when starting the adapter.

AEADB-700020 %1Cannot find bound data for %2.





AEADB-700021 %1 = %2

debugRole Adapter

AEADB-700022 %1 = NULL

debugRole Adapter

AEADB-700023 %1Bulk insert turning off.


AEADB-700024 Polling publishing table %1


AEADB-700025 Updating delivery status of publishing table %1.


AEADB-700026 %1 = %2

debugRole Adapter

AEADB-700027 %1Publishing by reference and selecting from reference object %2


AEADB-700029 %1


AEADB-700030 Received request %1.


AEADB-700031 Built reply %1.


AEADB-700032 %1sent to %2





AEADB-700033 %1


AEADB-700034 Bad date encountered: %1

errorRole Adapter Check your dateTime-related message data and make sure it is in the correct format.

AEADB-700035 Parameterized subject %1 exceeds TibRv maximum subject length of %2

errorRole Adapter Shorten a parameterized subject string length.

AEADB-700036 Unrecognized event

errorRole Adapter An unrecognized event received. Disable it if possible.

AEADB-700037 %1

errorRole Adapter An unexpected error happened internally. Contact TIBCO Support.

AEADB-700038 Poll for change failed

errorRole Adapter Run the adapter using debug 3 and verbose mode. Check printed database error code.

AEADB-700039 Update sequence Failed

errorRole Adapter Run the adapter using debug 3 and verbose mode. Check the update statement database error code.

AEADB-700041 %1, data = %2





AEADB-700042 Maximum of %1 thread allowed for %2 assuming %3

warnRole Adapter Maximum threads have been exceeded; however, the adapter runs by assuming the default thread count. Currently a publisher manager, subscriber manager and request-response communication can only have one thread each.

AEADB-700044 Main thread (name: %1, id: %2)


AEADB-700045 %1endpoints %2


AEADB-700046 Confirming request


AEADB-700048 %1Error creating reply message

errorRole Adapter Cannot create an Adapter SDK MTree object for a reply message. Contact TIBCO Support.

AEADB-700049 %1Request is being handled by thread %2


AEADB-700050 %1Could not start transaction

errorRole Adapter Cannot set autoCommit to false. Check the status of the database you are trying to connect to and try again. If it still fails, contact TIBCO Support.

AEADB-700051 Request Reply: Error retrieving %1

errorRole Adapter Missing required "sql" field when sending a request (rvMsg) to a request-response service for execution. Reconstruct the request message to include a "sql" field.




AEADB-700052 %1Error binding variables

errorRole Adapter Check the request message format. For a given field, make sure the data type in the message is the same as defined in the database.

AEADB-700053 %1Reply sent to %2


AEADB-700054 %1Error sending reply to %2 - %3

errorRole Adapter TIBCO Rendezvous failed to send a reply message to the request client. Report this error text to TIBCO Support.

AEADB-700055 %1Failed to create sender for reply

errorRole Adapter TIBCO Rendezvous failed to create a reply sender. Please report this error text to TIBCO customer support.

AEADB-700056 RequestReply: No reply name. Not sending reply

errorRole Adapter Specify a reply-subject when defining an adapter request-response service.

AEADB-700057 %1Error retrieving bind position

errorRole Adapter Check the request message format. For a given field in the message, make sure it is present in the database.

AEADB-700058 Stopped thread (%1)


AEADB-700059 Starting thread (name: %1, id: %2)


AEADB-700063 %1starts





AEADB-700064 %1Request: %2

debugRole Adapter

AEADB-700065 %1Reply: %2

debugRole Adapter

AEADB-700066 Load catalog table %1


AEADB-700071 Duplicate agent found %1

errorRole Adapter The adapter is terminated because a same-named configuration is started on the same TIBCO Rendezvous network. For more information, see to Notes on Configuring an Adapter on page 36.

AEADB-700072 %1MExceptionEvent - Not an RVMSG_RVMSG: %2

errorRole Adapter The adapter received an MExceptionEvent where an MDataEvent is expected. Check other running components’ configurations to determine the source of this message.

AEADB-700073 %1Unknown message format

errorRole Adapter The adapter received a TIBCO Rendezvous message of unknown format. Check other running components’ configurations to determine the source of this message.

AEADB-700074 %1Not processing message

errorRole Adapter The message is not being processed because an error has occurred. Check for related error messages to determine what the problem is.




AEADB-700075 %1Error converting event into an Mtree

errorRole Adapter A conversion error occurred in the Adapter SDK layer. The message may be corrupted or of a wrong format. Check the Adapter SDK error message if any to determine what the problem is.

AEADB-700076 %1CM sender: %2, CM sequence number: %3

debugRole Adapter

AEADB-700077 Received message altered:


AEADB-700078 Received message discarded by alterMsgSub()


AEADB-700079 %1Unable to create AE object from the message: %2

warnRole Adapter An unpacking error occurred in the Adapter SDK layer. The message may be corrupted or of a wrong format. Check the Adapter SDK error message if any to determine what the problem is.

AEADB-700080 Received its own message. Discarding.


AEADB-700081 Batch confirming %1 messages


AEADB-700082 RequestReply: Bind argument %1 must include a table name

errorRole Adapter The format of the Bind argument name should be tablename.columnname. This message will not be processed.

AEADB-700083 Published message on %1





AEADB-700084 ADB_SEQUENCE class %1 not supported

errorRole Adapter The class type of ADB_SEQUENCE can only be either string or i4. Make sure the class type of the ADB_SEQUENCE column of the publishing table is specified correctly in the repository.

AEADB-700085 Received ADVISORY %1


AEADB-700086 Publisher batch confirm timeout


AEADB-700087 %1 method invoked


AEADB-700088 Unknown driver %1

errorRole Adapter The adapter currently only supports ODBC drivers (including the IBM client access driver).

AEADB-700089 %1Request received from %2


AEADB-700090 Database error updating %1

errorRole Adapter An error occurred while updating the publishing table. This is a fatal error and the adapter will terminate. Use the error returned by the database to determine how to solve the problem.

AEADB-700091 Publishing ADBOPAQUE format not supported

errorRole Adapter The adapter currently does not support publishing in opaque format.

AEADB-700092 Keeping on original subject token %1





AEADB-700093 Binding data to insert placeholders


AEADB-700094 Polling for database change


AEADB-700095 Terminating agent


AEADB-700096 Unable to insert %1 into list

errorRole Adapter Wildcard TIBCO Rendezvous subject names are not currently supported for preregistered listeners.

AEADB-700097 Updating CM sequence failed

errorRole Adapter Currently not used.

AEADB-700098 %1Error binding parameter %2

errorRole Adapter There was a database error that occurred while binding parameters to a SQL statement. The message will not be processed. Use the database error if any to determine what the problem is.

AEADB-700099 RequestReply: Error retrieving value for parameter %1

errorRole Adapter A request has no value specified for input parameter %1. Add a (data, rvmsgData) field to the [Bind Data] of the request.

AEADB-700100 Received message:


AEADB-700101 Transaction rollback failed

errorRole Adapter The database was unable to rollback the transaction. Use the error returned by the database to determine how to allow this transaction to roll back.




AEADB-700102 Database object %1 not found

errorRole Adapter The database object was not found. Check to see if the object really exists (perhaps in another schema) and if you have permission to access it.

AEADB-700103 %1Unknown column type: %2 for field: %3. Ignoring

warnRole Adapter The type %2 is an invalid AE class type. For this particular situation, the adapter will skip this attribute and continue processing. Check in the repository for attribute %3 and change the class type to a valid AE class type.

AEADB-700104 Insert into the exception table failed

errorRole Adapter There was an error inserting into the exception table. This is a fatal error and the adapter will terminate. Use the error returned by the database to fix the inserts into the exception table.

AEADB-700105 %1%2 are not supported.

errorRole Adapter These AE/TIBCO Rendezvous data types are currently not supported by the adapter. Use only supported data types to represent this field instead; for example, a string or binary.

AEADB-700106 No data fetched for %1


AEADB-700107 New SQL: %1


AEADB-700108 Bulk insert of %1 rows executed for %2





AEADB-700109 Error executing bulk insert of %1 rows for %2

errorRole Adapter There was an error doing the bulk insert into the table. Use the error returned by the database to determine how to fix inserts into the table.

AEADB-700110 Statement changed, flush all previous bulk insert rows


AEADB-700111 Reassign data value to insert statement:


AEADB-700112 Cannot find publication entry. Not publishing message.

errorRole Adapter Internal error. Contact TIBCO Support.

AEADB-700113 Error retrieving column %1

errorRole Adapter The adapter could not retrieve this attribute from the message. Make sure the incoming message contains a value for this attribute.

AEADB-700114 %1

errorRole Adapter Internal error. Contact TIBCO Support.

AEADB-AEADB_700301

Error executing listen_alert procedure call

errorRole Adapter Check the corresponding database error.

AEADB-AEADB_700302

Error preparing listen_alert procedure call, stop alerter thread

errorRole Adapter Check if the listen_alert procedure exists in the database and if the user has authorization to execute it.

AEADB-AEADB_700303

Listen_alert procedure call not found, stop alerter thread





AEADB-AEADB_700304

Cannot stop alerter successfully, stop_alerter_agent call failed.


AEADB-AEADB_700305

Error registering alerter, config_alerter procedure call failed, stop alerter thread

errorRole Adapter The alerter failed to add the AQ subscriber. Check if the AQ database objects are configured properly in the database and if the user has authorization to access it

AEADB-AEADB_700306

Error unregistering alerter, cleanup_alerter procedure call failed.

errorRole Adapter The alerter failed to remove the AQ subscriber. Check if the AQ database objects are configured properly in the database and if the user has authorization to access it

AEADB-AEADB_700307

Stopped ALerter


AEADB-890001 AEADB-CONN-RETRY

infoRole Adapter Reconnect attempt %1 for service %2.

AEADB-890002 AEADB-CONN-RETRYSUCCESS

infoRole Adapter Reconnect succeeded on attempt %1 for service %2.

AEADB-890003 AEADB-CONN-RETRYFAIL

warnRole Adapter Reconnect failed on attempt %1 for service %2 -- will retry in %3 milliseconds. Make sure it is possible to establish an ODBC connection to the database.

AEADB-890004 AEADB-CONN-PUBDUP




warnRole Adapter Connection reestablished for the publisher - message may be a duplicate of a previously published message.

AEADB-890005 AEADB-CONN-REQERR

errorRole Adapter The request received could not be processed due to connection errors. Make sure it is possible to establish an ODBC connection to the database.

AEADB-890006 AEADB-CONN-STOP

errorRole Adapter Adapter stopping due to persistent connection errors. Please check your database and restart adapter. Make sure it is possible to establish an ODBC connection to the database.

AEADB-890007 AEADB-CONN-SVCSUSPEND

infoRole Adapter Adapter suspending service %1 due to persistent connection errors.

AEADB-910004 AEADB-STARTUP-4

errorRole Startup Startup Error. SDK Exception %1 occurred in the adapter initialization while creating the MAppProperties object. The Repository URL is %2 and the Configuration URL is %3. Please refer to SDK documentation for Repository URL and Configuration URL specification. Please cut and paste from SDK documentation for the above.


errorRole Startup Startup Error. SDK Exception %1 received on starting the adapter after initialization. The Repository URL is %2 and the Configuration URL is %3. Please verify your repository and environment settings. Refer to the User Guide documentation.





errorRole Startup Startup Error. Unable to create a connection with the target database using the dsn %1. The database error is: %2 Terminating adapter. Please verify your repository settings for the validity of connection parameters. Refer to the User Guide documentation.

AEADB-920001 AEADB-SUB-1

errorRole Subscriber Subscription error. Subscription service %1 listening on %2 received an unexpected event: %3. The Repository URL is %4 and the Configuration URL is %5. Check the configuration of the application that is publishing the event and make sure that it matches the inbound event definition for the above subscription service. Please refer to User Guide for details on configuration of subscription service.


errorRole Subscriber Subscription error. Subscription service %1 failed to deserialize the event received on subject %2 and SDK exception thrown is %3. The event is: %4. The Repository URL is %5 and the Configuration URL is %6. Check the configuration of the application that is publishing the event and make sure that it matches the inbound event definition for the above subscription service. Please refer to User Guide for details on configuration of subscription service.


errorRole Subscriber Subscription error. Subscription service %1 listening on subject %2 received error %3 in SDK message level UserExit %4. Make sure the UserExit parameters are valid and the user exit is invokable from SDK.





errorRole Subscriber Subscription error. Subscription service %1 listening on subject %2 could not get the class description of %3. The Repository URL is %4 and the Configuration URL is %5. Please check the repository configuration for this service. Please make sure the classes to be used by this service are accessible in the repository. Please refer to User Guide for details on how to configure, run and test the subscription service.


errorRole Subscriber Subscription error. Subscription service %1 listening on subject %2 failed due to database error: %3 Database is %4. The database commands and parameters are %5. Please look at the database error, note the database error code and consult with your database documentation.


errorRole Subscriber Subscription error. Subscription service %1 listening on %2 could not send response %3 on reply subject %4. The SDK error is %5. Please check your repository settings for the publish endpoint of this subscription service. Please refer to user guide on how to configure the subscription service.

AEADB-930002 AEADB-PUB-2

errorRole Publisher Publication error. Publication service %1 with publication subject %2 encountered database error: %3 while trying to create publish event with schema %4. Database is %5. The database command is %6. Make sure that the publication service is configured properly. Please look at the database error, note the database error code and consult with your database documentation.




AEADB-930003 AEADB-PUB-3

errorRole Publisher Publication error. Publication service %1 with publishing subject as %2 received event database %3. It failed while converting the event to "Minstance" as it could not get the class description for %4. Repository URL is %5 and the Configuration URL is %6. Please verify the configuration of the publication service and check that the schema/class definitions are present in the repository. Please refer to User Guide for details on how to configure a Publication service.

AEADB-940001 AEADB-REQRESP-1

errorRole Request_Response_Server

Request-Response error. Request-Response service %1 listening on %2 received unexpected null data in incoming request. Expects object %3. The Repository URL is %4 and the Configuration URL is %5. Please check the configuration of the application that is requesting the event and make sure that it matches the inbound event definition for the above RequestResponse service. Please refer to User Guide for details on configuration of RequestResponse service.



Request-Response error. Request-Response service %1 listening on subject %2 failed due to database error: %3Database is %4 and inbound event is %5. Please check the target application command and the parameters and make sure they are valid. Please look at the database error, note the database error code and consult with your database documentation.






Request-Response error. Request-Response service %1 listening on subject %2 failed to create Reply Business Object. Please check the database command and the parameters and make sure they are valid.



Request-Response error. Request-Response service %1 listening on subject %2 receive an error while sending Data on Reply Address %3. Error Message %4. Please check whether the request client is alive or message transport is functioning.

AEADB-990002 AEADB-SHUT-2

errorRole Shutdown Shutdown error. SDK cleanup exception = %1.



| 321

Appendix D Adapter Properties File

Topics


• Properties File Format, page 322

• Run-time Adapter Properties File Parameters, page 323

• Obfuscating or Encrypting a Password in the Properties File, page 337


322 | Appendix D Adapter Properties File

Overview

The run-time adapter parses a properties file at startup. The default run-time adapter properties file is named adbagent.tra.

The default properties file is located in the install-path\bin subdirectory.

Each line in a properties file is a single property. Each property consists of a key and a value. The key starts with the first non-whitespace character and ends at the first "=", ":", or whitespace character. The value starts at the first character after the equal sign (=). For example:

tibco.configurl=/tibco/private/adapter/test/config/config1tibco.repourl=tibcr://TEST_PROJECTtibco.username=admintibco.password=samplePasswordtibco.clientVar.service=7600tibco.clientVar.daemon=tcp:7600

Properties defined in the properties file override the same properties defined in the project.

Properties File FormatThe following restrictions apply to properties:

• The "!" character may not be used as a comment line indicator. Only the "#" character is recognized.

• The line continuation character is ignored (a value must fit on a line).

• The key may not contain any of the termination characters. Java allows termination characters by escaping the value with a preceding "\" character. TIBCO ActiveMatrix Adapter for Database does not support this syntax.

Tagging Values for Obfuscation

The presence of a "#" character as the first character in a value (not the key) indicates that the value has been obfuscated or is to be obfuscated. The obfuscation command-line tool prompts for values to be obfuscated when it encounters a value with "#" as the first character in the properties file.

When the obfuscate tool is run, it rewrites the properties file with the obfuscated value in place. See Obfuscating or Encrypting a Password in the Properties File on page 337 for more information.


Overview | 323

Run-time Adapter Properties File ParametersProperties are in two categories: Required Properties and Additional Properties. These are listed and described below.

Required Properties

In order for a properties file to be used with a run-time adapter, the following properties must be correct for the adapter’s configuration.

All paths inside a properties file, including Microsoft Windows directory names, must use forward slashes.

Table 19 Required Run-Time Adapter Properties File Parameters

Property Description

tibco.repourl repourl The absolute pathname to the local repository where the adapter instance is defined. For example: C:/TIBCO/LocalRepositories/repo.dat

For a remote project, the repourl value should use the form tibco.repourl tibcr@name where name is the repository name. For example:tibco.repourl tibcr@ADBRepoDefault

For Unix systems, the path separator should include a single slash, "/". For example: /local/tibco/repo/repo.dat

tibco.configurl

relative_path

or

tibco.configurl

absolute_path

The location of the adapter service inside the project file. If a relative path is specified, the adapter service is assumed to be under the default area in the project file (/tibco/private/adapter/). For example, the following value connects to an adapter service named adbpub in the /tibco/private/adapter/ directory:

tibco.configurl adbpub

If an absolute path is specified, the adapter instance is looked up in the repository as defined by the argument. For example:

tibco:configurl /tibco/private/adapter/adbpub



Additional Properties

You can modify the following parameters as necessary. Properties that start with ntservice are available only on Microsoft Windows platforms.

tibco.instanceid instance name

The name of the adapter instance.

The length of the name cannot be larger than 80 characters.

tibco.clientVar.adb.password

The password to connect to the targeted database. This can be obfuscated using the instructions in Obfuscating or Encrypting a Password in the Properties File on page 337.

application.args The properties (.tra) file to pass to TIBCO ActiveMatrix Adapter for Database. For example:

application.args adbagent --propFile C:/tibco/adapter/adadb/5.5/bin/adbsub.tra

application.start.dir The pathname of the adapter to start. For example:

application.start.dir C:/tibco/adapter/adadb/5.5/bin/

application.library The name of the dynamic library of the adapter instance.

application.ase.library

The name of the dynamic library of the adapter service engine instance.

This property is used only on the UNIX platforms.

Table 19 Required Run-Time Adapter Properties File Parameters (Cont’d)

Property (Cont’d) Description (Cont’d)

All paths inside a properties file, including Microsoft Windows directory names, must use forward slashes.


Overview | 325

Table 20 Additional Run-Time Adapter Properties File Parameters

Property Description

adb.addCustomHawkMethodsToClassMAgent <on/off>

The property is not set by default. Setting the property to off disallows adding custom methods to the adapter’s standard microagent. Setting to on allows custom methods to be added to the adapter’s standard microagent.

adb.batchPubStatusUpdates on|off -max_rows

When adb.PollingBatchSize is used, this optimizes publishing performance by batching message status updates to the publishing table. Do not use this option when messages are published using a parameterized subject name.

If an adapter instance stops before a batch update is performed, the status column is not updated. As a result, duplicate messages are published when the instance is restarted.

adb.debug The debug printing level. If not specified, the default, 0, is used. Possible values are:

0 - No debug information displayed.

1 - SQL commands executed against the database display.

2 - In addition to 1, the ODBC data source for each SQL command displays.

3 - Displays values for all ODBC placeholders.

adb.dsn ODBC system data source name. If not specified, the system data source name is read from the repository.

adb.encoding The encoding to use. Required when using a non-ASCII character set, as in globalization. See Overview, page 210, and Relevant Environment Settings, page 219, for more information. Valid values are:



Value

(no value)ibm-1370

tibx-eucJP

ibm-1386

LATIN_1

ibm-949

ibm-943

tibx-943

UTF8

Corresponding TIBCO Designer selectionASCII (default)

Big5

EUC-JP

GBK

ISO8859-1

KSC-5601

Shift_JIS (CP943)

Shift_JIS (TIBCO)

UTF8

adb.groupsize Specifies the number of rows to publish in a single message. Overrides the Group Size setting in the publication service configuration.

adb.password Password for the user account used to access your database. If not specified, the password stored in the repository is used.

adb.perfMon Either on or off. The default is off.

adb.PollingBatchSize Applies only to publisher instances. Limits the amount of messages to be picked up. The value indicates the number of parent rows to fetch for a poll interval. The default value, 0, indicates that all new rows should be fetched.

adb.PollingInterval Specific polling period. Applies only to publisher instances. If not specified the default, 10, 000 milliseconds, is used.

Table 20 Additional Run-Time Adapter Properties File Parameters (Cont’d)

Property (Cont’d) Description


Overview | 327

adb.pubBatchConfirmSize Applies only to publisher instances with publications that use certified message delivery. Optimizes performance by batching message status advisories to the publishing table. The value indicates the number of advisory messages to include in a single batch.

Do not use this option when messages are published using a parameterized subject name.

Note: If an adapter instance stops before a batch update is performed, the status column is not updated. As a result, messages that were successfully published might still have a status of P (pending) in the publishing table when the adapter instance is restarted. In this case, the ledger file contains the correct status information. A smaller value for this property decreases this risk.





adb.pubBatchConfirmTimeout Applies only to publisher instances with publications that use certified message delivery. This property specifies the number of milliseconds to wait before updating the status column. After this interval, an update is performed even if the batch size value is not reached. The default value is 10,000 (10 seconds). A value of 0 means that no timeout interval is used.

Do not use this option when messages are published using a parameterized subject name.

Note: If an adapter instance stops before a batch update is performed, the status column is not updated. As a result, messages that were successfully published might still have a status of P (pending) in the publishing table when the adapter instance is restarted. In this case, the ledger file contains the correct status information. A smaller value for this property decreases this risk.

adb.publishChildData Either on or off. Applies only to publisher instances. Enables publishing of table rows related to the source publishing table. By default, related table rows are not published.

adb.RetryTotal Total number of reconnection attempts.

adb.rvMaxQueueSize Applies only to subscriber instances. Maximum number of messages to allow in the TIBCO Rendezvous event queue. The default value is 0, which means no limit is placed on event queue size.

adb.SleepBetweenRetries Milliseconds of sleep between two reconnection attempts.




Overview | 329

adb.subBatchCommitSize Applies only to subscriber instances. The number of messages to batch before invoking a commit operation. The default is 1 (one).

Note: if messages greater than 32K are published, batching is automatically turned off.

adb.subBatchCommitTimeout Applies only to subscriber instances. The amount of time that can expire after which a batch commit operation is invoked. If not specified the default value of 10,000 milliseconds is used.

adb.subBulkInsertSize Applies only to subscriber instances. All incoming messages to insert are stored until this size is reached. Then a bulk insert operation is performed on the destination table. This value must be less than or equal to the value specified for adb.subBatchCommitSize, if used. The default value is 1.

If an update statement is published while messages are being batched, the bulk insert is performed regardless of whether the size value has been reached. After records have been inserted, the update operation is performed.

Note: Do not use this option if LONG, LONG RAW, image, or varbinary records are published.

adb.traceOldMessages Either on or off.





adb.unicode The unicode setting.

The default value is UTF8.

Valid values are UTF8 and UTF16

Additionally, the following options have to be set:

• On Microsoft Windows, select the Enable N-CHAR option from the Advanced tab when configuring the ODBC system data source.

• On UNIX platforms, set the EnableNcharSupport option to 1 in the ODBC initialization files.

Refer to the ODBC driver documentation for more details.

adb.useExceptTable on|off Enables the use of the exception table. The exception table is defined when a subscription is created.

adb.user Database account name used by the adapter to access your database. If not specified, the database username stored in the repository is used.

adb.verbose on|off Verbose mode. Print all available information to the console window or log file location. By default the verbose mode is off.

adb.wchar Default is SQL_C_BINARY. Required when using UTF-8 encoding with an adapter that uses a DataDirect driver to retrieve UTF-8 data from the database.




Overview | 331

adb.mutex This is a table that is used by the fault tolerance functionality. When an adapter instance starts and this property is present, it will try to issue an exclusive lock on the named table before processing any messages.

The adapter instance that is able to lock this table is considered a primary instance. All instances that are not able to issue an exclusive lock on this table are considered secondary instances.

The table name is shared within a fault tolerance group.

Example,adb.mutex FTtablename.

adb.heartbeat The time interval, in milliseconds, the secondary instance(s) waits before it tries to lock the table specified by the adb.mutex property.

adb.tablespace The DB2 tablespace name for OS390. The adapter uses the value of this property when creating the table used by the fault tolerance functionality.

adb.database The DB2 database name for OS390. The adapter uses the value of this property when creating the table used by the fault tolerance functionality.

adb.as400.defaultLibrary The default iSeries library to be accessed.

adb.as400.library The name of the library where the programs will be created.

adb.maxLongLen The buffer size. Used for long data types such as BLOB.





adb.requestResponseThreads The number of threads used by the Request/response Service. The default value is 1.

adb.stmtCache The number of cache statements for a generic RPC request/ reply service.

The number of statements that the adapter will cache so that it will execute the statement directly for repeated requests. If the cache is full, the adapter will remove the oldest message from the cache and will add the new statement. The default setting is off.

Example, adb.stmtCache 10

adb.terminateOnPubFail Specifies that if publication fails, the agent will terminate after the status has been updated to ’F’. The default value is off.

adb.ADBPublisher.preRegisteredListeners

Preregisters RVCM names for the specified subjects.

For example, adb.ADBPublisher.preRegisteredListeners=adb.sub1:rvcm1,adb.sub2:rvcm2

preregisters RVCM name rvcm1 to subject adb.sub1, and RVCM name rvcm2 to subject adb.sub2.




Overview | 333

adb.useSqlColumns Either on or off. Specifies whether the adapter should use SQLColumns() instead of a SELECT query to retrieve metadata for a table.

Note: Typically, a SELECT * FROM <table> WHERE 1>9 (which returns no results) is the fastest way to get metadata for that table. However, on some databases, such as DB2 OS390 v6, the database is not optimized for this query and a full table scan can result. Using an ODBC API call, SQLColumns(), can avoid this.

adb.rvAdvisoryNoLog This property specifies whether RV advisory messages are be logged in the adapter log files. The default value is off.

adb.setEmptyStringNullForRvMsg

Either on or off. Specifies whether the RVMSG fields of an empty string are treated as NULL or "" (empty strings). The default setting is off.

If the property value is set to on, empty strings are treated as NULL and if the property value is set to off, empty strings are treated as empty strings if the database allows it.

adb.<publisher_service_name>.lookback

Enables the adapter to look for remaining records if it did not fetch the number of records specified by the pollingBatchSize parameter for the publisher.

For example, if the publisher service name is ADBPublisher, the property name is adb.ADBPublisher.lookback.

adb.disableTerminationSubject

When this property is set to on, the adapter does not terminate on receiving a termination subject message.





adb.RequestResponseMaxRows This property specifies the maximum number of rows to fetch. This can be used limit the memory usage of the adapter. The unfetched rows will be ignored by the adapter.

adb.<table name>.poll.hint. <hint_value>

This feature helps improve the performance of your queries. This feature is only supported by Oracle databases.

See Using Oracle Hints on page 338 for more information.

adb.noDupDetection Disables detection of duplicate configurations.

Either on or off. The default value is off which indicates that duplicate configurations will be detected.

ntservice.account Username under which to run the Windows Service.

You can use this property to initially set the account for the service, but once the service is installed, use the Services control to change the user account of services.

ntservice.binary.path.absolute

Absolute path to the executable that is run when the service is started. For example:

ntservice.binary.path.absolute C:/tibco/adapter/adadb/5.5/bin/

adbagent.exe

ntservice.dependencies The number of dependencies.




Overview | 335

ntservice.displayname Name to display in the Services control for this Windows Service.

This property is useful if you wish to have multiple Windows Services for the same executable. That is, you may wish to have two adapter running on the same machine. By specifying different service names and display names for the adapters, you can accomplish this.

ntservice.interactive Either true or false. Specifies whether the Windows Service is interactive. Set to false if you are not using a system account.

ntservice.name Name for this Windows Service.

This property is useful if you wish to have multiple Windows Services for the same executable. That is, you may wish to have two adapters running on the same machine. By specifying different service names and display names for the adapters, you can accomplish this.

For example:

ntservice.name adapter_instance_name

ntservice.password Password for the username in the ntservice.account property.

You can use this property to initially set the password for the user account, but once the service is installed, use the Services control to change the password.





ntservice.starttype Start type for this Windows Service. Either manual or automatic. For example:

ntservice.starttype automatic

You can use this property to initially set the start type for the service, but once the service is installed, use the Windows Services control to change the start type of services.

tibco.clientVar Run-time values for global variables defined inside the repository. This value takes precedence over any global value set in the repository. Substitution takes place at run-time.

You append the global variable to tibco.clientVar, then give its value. For example, a global variable named DirLedger is specified as follows: tibco.clientVar.DirLedger

C:/tibco/adapter/adadb/5.5/myledger

Do not include the % character of substitution variables. For example, to set %%RvDaemon%%="tcp:7500", use tibco.clientVar.RvDaemon

"tcp:7500".

tibco.username User name and password used by the repository server to access the project. The password can be obfuscated using the instructions in Obfuscating or Encrypting a Password in the Properties File on page 337.

tibco.password

-version Displays a banner with version information, then exits. This option is for troubleshooting purposes only.




Overview | 337

Obfuscating or Encrypting a Password in the Properties File

Password Handling

At design-time, the adapter uses a password to connect to the backend application and fetch metadata. At run-time, the adapter uses a password to connect to the back-end application and interoperate with it. If you create a 4.x configuration using TIBCO Designer 5.5, and use the configuration against a 4.x adapter version, some special considerations are required for security.

If you plan to run the adapter locally, define the run-time password value to be a global variable. Before starting the adapter, include the run-time password as client variable in the adapter's .tra file and obfuscate it using obfuscate tool. For example, if the password value is defined as %%myPassword%%, create a global variable named myPassword in the global variables section with no value and include the following entry in the adapter's .tra file:

tibco.clientVar.myPassword

Obfuscating a Password

The run-time adapter needs a password to access the backend database. If you have unchecked the Remember Password checkbox in the Design-time Connection tab, the global variable %%adb.password%% is saved in the project. For run-time, you open the .tra file and add the password in clear text in the tibco.clientVar.adb.password=#password property. You can use the obfuscate tool to hide this password so it cannot be viewed by unauthorized users.

To obfuscate the password:

1. Using TIBCO Designer, open the adapter configuration and display the Design-time Connection tab.

2. Uncheck the Remember Password checkbox.

3. Apply the changes and save the project. This saves the global variable %%adb.password%% as the value of the password property in the project.

4. In the run-time adapter properties file, verify that the tibco.clientVar.adb.password=#password property is defined in clear text.

Do not set the password to type Password in the global variables section for adapter configurations that are set to AE Version 4.0 or AE Version 5.0 (in the Configuration tab Version field) or any intermediate version.

If you don’t want to obfuscate the password, remove the # at the beginning of the password.



5. Run the obfuscation tool supplied with the adapter against the properties file. This tool is named obfuscate.exe and resides in the TIBCO_home\tra\<version_number>\bin directory.

The command syntax is:

obfuscate tra_file_pathName

where tra_file_pathName is the absolute pathname of the adapter properties file that contains the tibco.clientVar.adb.password=#password property.

For example, on Microsoft Windows :

C:\tibco\tra\<version_number>\bin>obfuscate C:\tibco\adapter\adadb\5.5\bin\adbagent.tra

The password is now obfuscated and you can start the adapter with the changed properties file.

Encrypting a Password

Encryption is only supported for version 5.x adapters and higher. If you have a property in a properties file the needs to be encrypted, follow these steps:

1. In the property file, add the #! characters in front of the value you wish to encrypt. For example:

Repo.serverPassword = #!mysecret

2. Invoke the obfuscate utility from the command line:

<install-path>/tibco/tra/<version_number>/bin/obfuscate.exe --propertyfile=<property-file-name>

The next time you open the property file, mysecret will have been replaced with a random sequence of characters.

Using Oracle HintsOracle hints help improvce the performance of your queries. When the adapter publication service executes a poll operation to fetch data from a table, using hints greatly enhances the query.

Usage

To use hints add the following line to the .tra file:

adb.<table-name>.poll.hint <hint value>

where, <table-name> is the name of your table and<hint value> is the Oracle hint.


Overview | 339

Examples

a. To force an index scan when polling, set the property as follows:adb.p_p1.poll.hint /*+INDEX(P1,P1_INDX)*/

where P1 is the publisher table, and P1_INDX is is the index created on the publishing table.

The adapter will execute this select query on the publishing table: SELECT /*+INDEX(P1,P1_INDX)*/ * FROM P1 WHERE ID = ?"

b. To force a index scan when the adapter fetches records from a child table set the property as follows:

adb.p_p1.poll.hint /*+INDEX(C1,P1_INDX)*/

where C1 is the child table, and P1_INDX is the index created on the child table.

The adapter will execute this select query on the child table:SELECT /*+INDEX(C1,P1_INDX)*/ * FROM C1 WHERE ID = ?"




| 341

Index

Symbols

_ 36, 171. 171* 171> 171

A

access to DB2 AS/400 tables 76adapter

changing a configuration 37, 203checking for duplicate 18component (operation) information through TIBCO

Hawk 235error messages when changing a configuration 38naming restrictions 36size limitations 36stopping 36tracing 198

adapter configuration, changing 37, 203Adapter Service Engine 269

configuration 275deployment 280installation 271Overview 270

ADB_ERROR_TEXT 161ADB_ERROR_TIME 162

time zone 162, 164ADB_L_CMSEQUENCE 156ADB_L_DELIVERY_STATUS 156ADB_OPCODE 155, 161ADB_REF_OBJECT column in the publishing

table 179ADB_SEQUENCE 154ADB_SET_SEQUENCE 154ADB_SOURCE 11, 160, 160

ADB_SUBJECT 154ADB_TIMESTAMP 155

timezone 155ADB_UPDATE_ALL 156, 161adbDateTime 177adbPreCommit() 189adding

listeners 173adding a sequence for parent-child 181adding an association for parent-child 182agents 222alerter

on Sybase SQL Server 118alerts 222auto-discovery process, TIBCO Hawk 224

B

binary type 172bind statements 129

C

callout library 187callout library, compilers 190certified message delivery

exception handling 285moving ledger files 175preregistering listeners 173

Class Microagent Name field, adapter 66closure field 128column names in an exception table 163column names, reserved prefix 37columns, maximum number 36command line arguments 236


342 | Index

commandsadbagent 116, 120Connect 47, 51copy 115, 119cp 118isql 115, 116, 118, 119, 120select 28sqlplus 25, 113

commit_and_notify_table 111compilers for callout libraries 190configuration properties, retrieving through TIBCO

Hawk 236conversion, messages 6custom RPC operation 139customer support xx

D

dat file format 22data format 6data source, ODBC

using 325database errors, capturing 161database type mapping 171Date types 172date/time values 177dateTime 170DB2 on OS/390 trigger statements syntax 205debug, adapter option 325, 325delete trigger 74destination table

schema 10distributed queues 146dot character 171dsn, adapter option 325duplicate adapters 18

E

error messages about database cleanup 38

examplesREF cursor procedure 151

Exceed FAQ 287exception table

defined 10structure of 161

exception table, column names in 163

F

fault tolerance 13, 331file format

dat 22VC 22

Float data type 172format, messages 6

G

global variables 38setting for monitoring 65

I

install TIBCO ActiveMatrix 272

K

key columns, requirement for publish-by-reference 16


Index | 343

L

ledger fileslocation 175moving 175publication size status 327publication timeout status 328retrieving information through TIBCO Hawk 245

listeners, adding 173load balancing

across adapters 146in an adapter 144

Log File field, adapter 63log file options 198log files location 61Log Info field, adapter 63Log to Standard field, adapter 63logtest, sample program 19LONG data type 82loop detection

column used for 160defined 11using for replication 11

M

mapping database types 171messages 6

conversion 6data format 6

microagent methods supported 228Microagent Session field, adapter 66MInstance

data format 77MInstance data format 77multiple file project 12

N

naming restrictionsadapters 36database columns 37publishing table 84subjects 171

notify procedureon Microsoft SQL Server 115on Sybase 118

notifytable procedureon Microsoft SQL Server 115on Sybase 118

NULL attributes 90

O

obfuscation 322opaque format, using to publish 19

P

parameterized subject 172binary type 172Date types 172Float type 172illegal characters 172


344 | Index

parametersAE Type 75, 91Batch Publish Status Updates 54Child Table Mapping 92Class Reference 86, 96, 102Debug Level 59Deploy On Save 38Do Not Generate Triggers 85Enable Loop Detection 85Endpoint Reference 86, 96, 102Generate Verbose Output 59Join To 76, 92Polling Batch Size (Maximum Rows) 53Polling Interval 53Publish Child Data 55Publisher Batch Confirm Timeout 54Publishing Table 84Reply Subject 102Storage Mode 82Subscriber Batch Commit Size 55Subscriber Batch Commit Timeout 55Subscriber Bulk Insert Size 56Table Name 92Tables and Columns 75, 91Type 75, 91Update Mode 84Update Trigger? 75Use Exception Table 56Use? 75, 91User Key 75, 91

parent/child publication 16parent-child association 182parent-child sequence 181password property 336PL/SQL procedure for REF cursor 151preregistering listeners 173Publication Service typical flow diagram 4publications

changing subject 16size limitations 36

publish-child-data, adapter option 328publishing

by reference 15, 82by value 15source data 179

publishing tablecolumns in 154name size 84restrictions for 82

Q

quality of serviceand load balancing 146

queue member load 147

R

record size, limitations 36REF data type 151reference object location 179request

defined 126request/response

load balancing for 146using 124using multiple threads 144

response defined 127reviewLedger, TIBCO Hawk method 245RPC programs, client 142rv_Rpc() 130rv_Send() 130rv_SendWithReply() 130RVMSG_DATETIME 170RVMSG_INT 170RVMSG_OPAQUE 170RVMSG_REAL 170RVMSG_STRING 170

S

same-named adapters 18sample programs

logtest 19


Index | 345

schema changes affecting adapter configuration 38schemas

for related tables 77referencing external 10

sequence number 154source data publishing 179source data publishing example 185source table

loop detection column 160schema for 10

SQL statement terminator 205SQL_BATCHRETURN class, structure of 138, 138SQL_BIND class, structure of 138, 138SQL_OPS class, structure of 135SQL_RESULTSET class, structure of 136SQL_RETURN class, structure of 136SQL_ROW class, structure of 136SQL_STATEMENT class, structure of 138, 138Standard Microagent Name field, adapter 65standard RPC operation 135starting an alerter on Oracle 111stopping an adapter 36sub-batch-commit-timeout, adapter option 329sub-bulk-insert-size, adapter option 329subjects

changing dynamically 16naming restrictions 171parameterized 172subscriber 36using wildcards 171, 173

subscriptionssize limitations 36wire format for 177

substitution 38support, contacting xxsystem

configurl command-line argument 323, 323

T

table access problems, overcoming 76technical support xx

TIBCO Hawkbackground information 222enterprise monitor components 222methods supported 228Monitoring tab use with 65

TIBCO Hawk methodsgetComponents 235getConfig 236getRvConfig 239getStatus 240reviewLedger 245

TIBCO Rendezvous, retrieving configuration through TIBCO Hawk 239

TIBCO_HOME xviiTracing 290, 290tracing features 198Tracing Levels and Fields 291, 291tracking identifier 290, 294triggers

no primary key 74

U

underscore character 171update feature 16update trigger 74Use Advanced Logging field, adapter 61user authority 76user callout library 187user schema 110username property 336use-trace-file, adapter option 198

V

variable substitution 38variables 38version, adapter option 336


346 | Index

W

wildcard subject names 171, 173


Documents

TIBCO ActiveMatrix Adapter for Database ... - TIBCO Software