Chicago Inside Track Application Development Based on ABAP and SAP HANA (1)

Rich Heilman / SAP HANA Product Management

Thomas Jung/ SAP HANA Product Management

Application Development Based on ABAP and SAP HANA

© 2013 SAP AG or an SAP affiliate company. All rights reserved. 2

Full Day Workshop: ABAP on HANA

• Hands-on workshop – bring your

laptops and development skills

• Focus on ABAP on HANA for the

Business Suite

• Learn HANA Modeling, consuming

views from ABAP, SQLScript, and

OData services


Disclaimer

This presentation outlines our general product direction and should not be relied on in making a

purchase decision. This presentation is not subject to your license agreement or any other agreement

with SAP. SAP has no obligation to pursue any course of business outlined in this presentation or to

develop or release any functionality mentioned in this presentation. This presentation and SAP's

strategy and possible future developments are subject to change and may be changed by SAP at any

time for any reason without notice. This document is provided without a 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. SAP assumes no responsibility for errors or omissions in this

document, except if such damages were caused by SAP intentionally or grossly negligent.

Introduction


The Role of ABAP for Application Development on SAP HANA Applications based on ABAP and SAP HANA

SAP Business Suite powered by SAP

HANA available since January 2013

transition existing ABAP custom

developments to SAP HANA

optimize and develop new add-ons for

SAP Business Suite

SAP NetWeaver BW powered by SAP

HANA available since November 2011

significant new features for Enterprise

Data Warehousing / Big Data planned for

December 2013

ABAP custom developments in context of

SAP NetWeaver BW

develop completely new applications

(often running side-by-side with SAP

Business Suite)

specifically design the applications for

leveraging SAP HANA

SAP Business Suite SAP NetWeaver BW New applications

based on ABAP 7.4 = the “to-go” release for all SAP HANA based ABAP

applications


The Role of ABAP for Application Development on SAP HANA ABAP and the cloud

On-premise Private (managed cloud) Public

Cloud

SAP HANA Cloud Platform Application Development | Integration | Database and Analytics | Foundation

Any DB SAP HANA

ABAP development

Java development

SAP HANA Enterprise Cloud: managed

service offering for running

SAP Business Suite

SAP NetWeaver BW

custom SAP HANA applications

Line of Business Cloud Applications

Marketplace:

Customer applications + extensions

Partner / ISV applications + extensions

ABAP-based and non-ABAP business

applications running in your datacenter

based on

traditional database systems

SAP HANA database

native SAP HANA development

How can my ABAP code benefit

from SAP HANA?


What happens during the migration to SAP HANA? Best practices for custom code transition (1/2)

SAP HANA is a complete relational database

ABAP code only using database independent features

continue to run

general performance guidelines stay valid for SAP HANA

Priorities for custom code transition

avoid (functional) regression

detect (additional) performance optimization potential

Required and recommended adaptations

DB-migration related (e.g. usage of native SQL)

functional related (in particular code relying on implicit

sorting behavior)

performance related (usage of code patterns that can be easily optimized

for SAP HANA)


What happens during the migration to SAP HANA? Best practices for custom code transition (2/2)

Plan

Create project plan for migration project

including derive basic effort estimates

Housekeeping of custom code

Perform ABAP Code Analysis

Collect SQL profile of system

Proof of Concept activities

Prepare & Migrate

Regression-free migration (functional and

performance related)

Mandatory adaptations

Simple optimizations

Technical validation of migration

Optimize & Go-live

Performance tuning and harvesting of further

optimization potential / planning next steps

Analyze SQL profile of selected business

processes

Plan optimization projects

Full validation cycle

Primary goals Tasks


What happens during the migration to SAP HANA? Best practices for Open SQL (“5 golden rules”)

Existing “golden rules” for Open SQL remain valid as general recommendations. There are

some shifts of priorities (e.g. index definition (lower) and avoidance of nested selects (higher))

Keep the result sets small rule still valid

Minimize amount of transferred data rule still valid and even more important for SAP HANA

avoid selecting large numbers of not required columns

Minimize the number of data transfers rule still valid and even more important for SAP HANA

prefer array operations for INSERT, UPDATE und DELETE

when changing many records

avoid nested SELECT loops

Keep load away from the database keep unnecessary load away from the database

code pushdown for data-intensive calculations to benefit

from SAP HANA (see also slide 26 for details)

Minimize the search overhead rule still valid, but less important

in most cases SAP HANA does not require secondary

indices


How can my ABAP code benefit from SAP HANA? The new paradigm

AS ABAP

SAP HANA Database

Calculation

Calculation

“Data to Code” “Code to Data”

Code pushdown means delegating data intense calculations to the

database layer


How can my ABAP code benefit from SAP HANA? New features in ABAP 7.4

Benefits of In-Memory Architecture

DB-near programming

Transparent optimizations

Usage of HANA-specific

features

external views SP2

database procedure proxies SP2

HANA transport container SP2

SAP HANA content integration

* might also support HANA-specific features in the future (> SP5) / ** ‘>= SP5’ means SP5 or later

advanced Open SQL SP5

advanced ABAP view building

SP5*

ABAP-managed procedures SP5

Fulltext index in DDIC SP2

Advanced ABAP database

programming

Fast Data Access (new data exchange protocol) >= SP5** optimized SELECT... INTO ITAB and SELECT SINGLE >= SP5**

optimized FOR ALL ENTRIES-clause >= SP5**


How can my ABAP code benefit from SAP HANA? Demo scenario

Example: Calculation of shipping fee per

target country and period (simplified)

Shipping fee per target country = SUM shipping fee of

all delivered sales orders

Parameters to determine shipping fee per sales order

Target country

Relevant weight of sales order items

Relevant weight per item = MAX of

Volume weight (length x width x height / 5000)

Weight

January February

Jan

uary

Target country Shipping fee Currency

DE 150.000 EUR

FR 70.000 EUR

US 170.000 EUR

Feb

ruary

Target country Shipping fee Currency

DE 120.000 EUR

FR 90.000 EUR

US 180.000 EUR


Transparent optimizations Overview

Basic idea: optimize the communication between the AS ABAP

and the SAP HANA database ‘under the hood’ to accelerate existing

programs automatically

Fast Data Access (FDA)

new data exchange protocol between the AS ABAP and SAP HANA database

bulk operations instead of cell-wise data exchange

currently supported for Open SQL statements (but not for native SQL)

Optimized statements

optimized SELECT... INTO ITAB and SELECT SINGLE >= SP5*

optimized FOR ALL ENTRIES-clause >= SP5*

* ‘>= SP5’ means SP5 or later


Basic idea: support more standard SQL features

(for SAP HANA as well as for traditional database platforms)

Extended JOIN-support

it is possible to nest JOIN-statements (i.e. to specify JOINs on the ‘right hand-side’ of a JOIN-condition)

Support for literals

ABAP literals and ABAP variables can be used in the projection list

Expression support

CASE-statements can be used in the projection list, nesting is possible

Arithmetic expressions for integral, decimal and floating point calculations (e.g. +, -, *, DIV, MOD)

String expressions (concatenation of columns with &&)

SQL functions (e.g. CAST (currently restricted to FLTP), COALESCE)

Advanced Open SQL Overview


Advanced Open SQL Example

SELECT product_id && ',' && @space && category AS product,

CASE tax_tarif_code

WHEN 1 THEN price * @lc_factor_1



END AS price_vat,

currency_code AS currency

FROM snwd_pd

INTO CORRESPONDING FIELDS of @ls_result

WHERE product_id IN @so_prod.

WRITE: / ls_result-product,

ls_result-price_vat CURRENCY ls_result-currency,

ls_result-currency.

ENDSELECT.

concatenation of columns

CASE-statement

ABAP constants / variables

escaped by @

projection list separated by

commas


Advanced View Building Overview

Basic idea: support code pushdown to the database layer and

simplify consumption of relational data models by means of ‘view entities’

View entities

are defined through the Eclipse-based ABAP development environment by means of so-called ‘DDL source’

can be directly used in Open SQL (but also have a corresponding DDIC view)

Support code pushdown

similar features as advanced Open SQL; in addition: UNION, UNION ALL, non-equi INNER JOIN, support of literals in ON-condition

Simplify consumption

nested views (a view consumes another view)

definition of associations between view entities (and consumption within WHERE-clause as well as within expressions)


Advanced View Building Example (1/2)

@AbapCatalog.sqlViewName: 'ZCD201_VRVNU'

DEFINE VIEW zcd201_revenues AS SELECT FROM snwd_so

{

snwd_so.buyer_guid,

SUM(snwd_so.gross_amount) AS gross_amount,

SUM(snwd_so.net_amount) AS net_amount,

SUM(snwd_so.tax_amount) AS tax_amount,

snwd_so.currency_code AS currency

} GROUP BY snwd_so.buyer_guid, snwd_so.currency_code

name of DDIC view

name of view entity (can be used

in Open SQL)


Advanced View Building Example (2/2)

@AbapCatalog.sqlViewName: 'ZCD201_VCUST'

DEFINE VIEW zcd201_customer AS SELECT FROM snwd_bpa

INNER JOIN snwd_ad ON

snwd_ad.node_key = snwd_bpa.address_guid

ASSOCIATION[*] TO zd201_revenues AS revenues ON

revenues.buyer_guid = snwd_bpa.node_key

{

snwd_bpa.node_key, snwd_bpa.bp_id,

snwd_bpa.company_name,

snwd_ad.country,

snwd_ad.postal_code,

snwd_ad.city,

revenues.gross_amount,

revenues.currency

} WHERE bp_role = '01'

definition of an association to

previously created view entity ZCD201_REVENUES

consumption of the association

within the projection list


Basic idea: manage database procedures and their lifecycles from

within the ABAP development infrastructure

ABAP method is used as container for database procedures

addition BY DATABASE PROCEDURE indicates that a method is executed in the database

addition FOR db_platform indicates the database platform (currently supported HDB = SAP HANA database)

Implementation of method body is done in SQLScript

editing environment is the ABAP class editor

embedded syntax check is available for SQLScript code (different from native SQL!)

other languages might be supported in the future (for example R)

ABAP server creates database procedure in database catalogue during generation

runtime artifact is generated into schema SAP<SID> (and not _SYS_BIC)

ABAP-managed Database Procedures Overview


ABAP-managed Database Procedures Example

CLASS zcd201_cl_snippet_amdp DEFINITION...

INTERFACES if_amdp_marker_hdb.

METHODS: determine_sales_volume

IMPORTING VALUE(iv_client) TYPE mandt

EXPORTING VALUE(et_sales_volume) TYPE tt_sales_volume.

ENDCLASS.

CLASS zcd201_cl_snippet_amdp IMPLEMENTATION.

METHOD determine_sales_volume BY DATABASE PROCEDURE

FOR HDB LANGUAGE SQLSCRIPT

USING snwd_so_i snwd_so_sl snwd_pd.

-- here you use SQLScript

ENDMETHOD.

ENDCLASS.

Marker interface

method additions

forward declaration of accessed

database objects


How can my ABAP code benefit from SAP HANA? Recommendations

Possibility to use HANA-specific features

Code running on all supported database systems

Prefer Advanced ABAP database programming features to optimize ABAP code advanced Open SQL

advanced ABAP view building use ABAP-managed procedures where needed,

but think of fallbacks in case the code also has to

support traditional database systems

Use SAP HANA content integration if requirements cannot be fulfilled by Advanced

ABAP database programming

to consume pre-packaged content (e.g. SAP HANA

Live)

to develop reusable artifacts across application

stacks

The SAP<SID> schema is managed by the

AS ABAP only create artifacts in SAP<SID> schema through

ABAP use APIs to write to tables in SAP<SID> schema

and avoid using native SQL / SQLScript to write to tables SAP<SID> schema

What are completely new

possibilities?


Example 1: Application Function Libraries Overview

Application Function Libraries

libraries of reusable functions linked dynamically to the SAP

HANA database kernel (implemented in C++)

can be used to perform data intensive and complex calculations

BFL and AFL generally released as of SAP HANA SPS5

Business Function Library (BFL)

contains pre-built, parameter-driven business functions

(examples: annual depreciation, days outstanding)

Predictive Analysis Library (PAL)

contains statistical algorithms from areas like clustering,

classification, social network analysis

SAP HANA

Index Server

Business Function

Library (BFL)

Predictive Analysis

Library (PAL)

…


Example 1: Application Function Libraries Exemplary functions

function: Linear Average (BFL)

calculates a linear average, in which larger weights are applied to more recent periods

weights decrease lineary as you look back over time

function: ABC Analysis (PAL)

classify objects (customers, employees, products…) based on a particular measure

group objects by their estimated importance (“A”, “B”, “C” objects)

function: K-means (PAL)

partition records into k clusters in which each record belongs to the cluster with the

nearest center

e.g. used to group customers into segments


Example: ABC analysis of products

Classification based on revenues attributed to the products

“A” products: 20% of the products accounts for 70% of the revenue

“B” products: 30% of the products accounts for 20% of the revenue

“C” products: 50% of the products accounts for 10% of the revenue

Example 1: Application Function Libraries Demo scenario

Product ID Product Name Revenues Currency Classification

HT-1000 Notebook Basic 15 1.500.000 EUR A

HT-1001 Notebook Basic 16 800.000 EUR B

HT-2000 7” Widescreen DVD

player

1.700.000 EUR A

HT-2001 10” portable DVD player 500.000 EUR C


Example 2: Business rules SAP NetWeaver Decision Service Management (DSM)

SAP NetWeaver Decision Service management helps to externalize

decision-making logic from ABAP-based application code

Decision Service

Repository

Decision Service

Versioning & Reporting

Decision Service

Modeling

Decision Service

Lifecycle Management

Model and change

decision services

Centrally store

decision services

Manage technical

aspects of decision

services

Gain full transparency for

analysis and legal

purposes

SAP NetWeaver Decision Service Management

IT Business

domain

expert

SAP

ERP 1

SAP

ERP 2

SAP

CRM …

Non-

SAP …

Decision Service Distribution Distribute decision

services for local

execution


Example 2: Business rules DSM and SAP HANA (1/2)

Dynamic Database View: deployment to SAP

HANA generates a result view based on

database (tables, views) and decision table

content

database table or view as data source

decision table for calculated columns

result view to combine data source with calculated

columns

Decision table processed on SAP HANA database by SELECT on result view

can be used directly in ABAP and in database lookup

expressions

Database table or view

Decision table

Result view

Column 1 Column 2

A 10

B 20

Condition

Column 1

Condition

Column 2

Result

Column 3

Result

Column 4

A 100 1.000

B 0.. 10 200 2.000

B > 10 300 3.000

Column 1 Column 2 Column 3 Column 4

A 10 100 1.000

B 20 300 3.000


Example 2: Business rules DSM and SAP HANA (2/2)

Usage of database procedures in DSM

functions on SAP HANA

either based on database procedure proxies …

… or based on ABAP-managed procedures

allows to push down expensive expressions to the

database

Analyze DSM expressions with SAP HANA

Example: decision table

Which row is mostly evaluated as true?

How do values develop over time?

Row 1 (100, 1.000)

Row 2 (200, 2.000)

Row 3 (300, 3.000)


Example 3: Text Search / Analysis Overview

SAP HANA

Text Analysis

result database

table

Fu

llte

xt

ind

ex

Database

table

Text

Analysi

s

Text

Searc

h

Text Search

search character-type columns of database tables in

column store

search operations: exact, fuzzy, linguistic…

functions: ranking, snippets, facets…

requires creation of so-called fulltext index

Text Analysis

structure, transform, enrich unstructured data for the

purpose of discovery or analysis

happens based on fulltext index (optional configuration

for a fulltext index)

results of text analysis are stored in a database table


Example 3: Text Search / Analysis Using text search capabilities in ABAP

Fulltext index in the SAP HANA database

is implicitly created only for column store data types TEXT and SHORTTEXT

has to be explicitly created for other character-

type columns

Creation of fulltext index through ABAP

create (extension) index in transaction SE11

choose exactly one field for the index and specify

that the index is a fulltext index

available as of ABAP 7.4 SP2

Access by means of native SQL SELECT pd.product_id, tx.text

FROM snwd_pd AS pd

INNER JOIN snwd_text_key AS tk

ON tk.node_key = pd.name_guid

INNER JOIN snwd_texts AS tx

ON tx.parent_key = tk.node_key

WHERE CONTAINS(tx.text, 'Notebuch',

FUZZY(0.7))

ORDER BY SCORE( );

Notebook Professional 15,

Notebook Basic 17,

Notebook Professional 17…

How can I get started with

ABAP on SAP HANA?


How can I get started with ABAP on SAP HANA? Overview / our offers

Book a training at SAP Education

for example: HA400 ABAP Programming for SAP HANA

Visit our web site to browse through documentation and get access to trial

systems (http://scn.sap.com/community/abap/hana)

ABAP for SAP HANA reference scenario

tutorials (e.g. how use database procedures) and best practices (e.g. performance

guidelines)

AS ABAP 7.4 trial as virtual appliance

Visit one of the many SAP TechEd lectures, hands-on workshops,

expert networking sessions

http://scn.sap.com/community/abap/hana


How can I get started with ABAP on SAP HANA? Trial systems

AS ABAP 7.4 trial as virtual appliance

application server and database are deployed at a cloud service

provider (currently supported Amazon EC2)

including integrated development environment in Eclipse…

… plus sample scenarios, tutorials and best practices

free of charge (except the charges of the cloud service provider)

Currently available systems

SAP NetWeaver Application Server ABAP 7.4 on SAP HANA

(AS ABAP 7.4 SP2, HANA Revision 1.00.56)

SAP NetWeaver Application Server ABAP 7.4 on SAP MaxDB

(AS ABAP 7.4 SP2, MaxDB 7.9.08)

An AS ABAP 7.4 trial on SAP MaxDB is planned to be available for download in

Q4/2013.

Cloud

Developer

Tune Your Custom ABAP Code

Get Ready for SAP HANA


SAP Business Suite powered by SAP HANA Migrate to SAP Business Suite powered by SAP HANA

SAP HANA Database

AS ABAP 7.4

SAP Business Suite 2013

Any Database

AS ABAP

SAP Business Suite

Migrate

based on ABAP 7.4 = the “to-go” release for

all SAP HANA based ABAP applications

SAP Business Suite on SAP HANA

Upgrade

Upgrade SAP Business

Suite

Custom Code

SAP Business Suite 2013

Custom Code

Custom code is a central part of the

business functionality

How can customers smoothly migrate and

optimize their code for SAP HANA?

Adapt


Migrate to SAP Business Suite powered by SAP HANA Impact on custom ABAP development

Which of my ABAP code must be changed to

avoid potential functional issues?

Which of my ABAP code shall be optimized to

ensure good performance?

Which of my main business processes have

potential for HANA so that I can speed up those

massively?

And how can I find all this ABAP code easily

and do the adaption efficiently…?

Prepare Custom Code for SAP HANA ABAP Test Cockpit and Static Checks


Prepare your custom code for SAP HANA Functional corrections for existing ABAP code

In general existing ABAP code runs on SAP HANA as before

Only if ABAP code relies on technical specifics of the old database, ABAP code changes

might be necessary

Which of my ABAP code must be changed to avoid potential functional issues


Prepare your custom code for SAP HANA Functional corrections for existing ABAP code

Technical changes with SAP HANA that may affect existing DB specific

ABAP code

Technical Change Details and Examples – Effect on DB specific ABAP code

DB migration Each DB has specific features and unique technical behavior.

DB specific code may rely on these features of the used database.

SAP HANA architecture Column based architecture - as a consequence e.g. secondary DB indexes are

less important.

DB specific code may rely on the existence/usage of certain DB indexes.

Depooling/Declustering During the migration to SAP HANA most pool and cluster DB tables are

transformed to transparent DB tables (depooling/declustering) so that the tables

can be used in analytic scenarios.

DB specific code may rely on the technical specifics of pool and cluster tables.

Details: See OSS note 1785057


Technical Change ABAP code which must be checked / corrected Recommended solution

DB migration Native SQL, Hints, ADBC which uses DB specific features Use Open SQL if possible, adapt native SQL and DB hints if necessary

Prepare your custom code for SAP HANA Functional corrections for existing ABAP code – Details and examples

Example for native SQL Example for a DB hint


Technical Change ABAP code which must be checked / corrected Recommended solution

SAP HANA

architecture

Transparent DB tables:

- Relying on not guaranteed / implicit sorting of DB query

results

- (Rare problem) Analysis of technical DB index information using

function modules 'DB_EXISTS_INDEX‘, 'DD_INDEX_NAME‘

Recommendation:

Never rely on any implicit sorting - no robust code!

Always use ORDER BY … or SORT … BY to sort query result sets

explicitly

Try to get rid of DB index analysis

Depooling/

Declustering

Pool/Cluster tables:

- Relying on not guaranteed / implicit sorting of DB query

results

- (Rare problem) Direct access to physical clusters / pools

Recommendation:

Never rely on any implicit sorting - no robust code!

Always use ORDER BY … or SORT … BY to sort query result sets

explicitly

Get rid of direct access to physical clusters / pools

Prepare your custom code for SAP HANA Functional corrections for existing ABAP code – Details and examples

Read Table … BINARY SEARCH works only for a

correctly sorted internal table.

But no explicit sorting of the query result is done

-> Error in code which may work by chance

because of DB Index usage

Correction

Add explicit sorting by adding ORDER BY

to the SELECT or by using an ABAP SORT

Example for ABAP code which

relies on sorted DB query

results


Prepare your custom code for SAP HANA Functional corrections for existing ABAP code – static checks I

For all main functional error categories we provide Code Inspector checks to find the issues easily


Prepare your custom code for SAP HANA Functional corrections for existing ABAP code – static checks II

“Search problematic statement…w/o ORDER BY” check

Preferred check to find code which relies on sorted DB content

searches for statements like READ BINARY SEARCH, DELETE

ADJACENT DUPLICATES, … accessing unsorted DB content

works for transparent and pool/cluster DB tables

low false positive ratio

“Depooling/Declustering: Search for…w/o ORDER BY” check

Searches for SELECT <pool/cluster table> without ORDER BY but no

further analysis of the data flow is done

works only for pool/cluster DB tables

(before and after depooling/declustering)

high false positive ratio

Recommendation

Run both checks for your custom code

Concentrate on findings (priority 1 and 2) of the “Search problematic

statement…w/o ORDER BY” check

Treat findings of “Depooling/Declustering: …” check with reduced priority


Prepare your custom code for SAP HANA ABAP Test Cockpit – Run your Code Inspector checks in a new tool

ABAP Test Cockpit (ATC) is a

framework for running static checks on

ABAP objects to find: functional,

usability, performance, security, … bugs

ABAP Test Cockpit is based on Code

Inspector and adds new quality gates,

quality processes and better usability


Prepare your custom code for SAP HANA ABAP Test Cockpit – Quality process and Q gates

Development System

Quality System

Q-experts run mass checks and distribute the results

Developer run static / unit / scenario tests on their objects

Periodic check runs to validate the code of a development team

Q-Gate:

Check during

transport release

Q-Gate:

Mass check run and

consolidation test


Summary - Prepare your custom code for SAP HANA ABAP Test Cockpit and static checks

In general existing ABAP code runs on SAP HANA as

before. Only DB specific ABAP code must be analyzed

New static Code Inspector checks are available which

help to find the code which must be analyzed

ABAP Test Cockpit (ATC) is the new standard tool for

static ABAP code checks

ATC can be used now to prepare the custom ABAP

code for HANA and to improve the general quality of the

code

ATC availability starts with NW 702 SP12 / NW 731 SP5.

In older releases the Code Inspector can be used

Optimize Custom Code for SAP HANA

SQL Monitor, Static Checks and SQL Performance Tuning Worklist


Optimize Custom Code for SAP HANA Tooling

How can I find the ABAP code which shall be optimized or which has potential

for massive acceleration using code push down to SAP HANA

In general no changes are necessary if your SQL code follows the golden Open SQL rules

Use static checks to find SQL patterns that violate the golden Open SQL rules

Add runtime performance data from production to rank the findings and to find potential for

massive acceleration


Optimize Custom Code for SAP HANA Classic golden SQL rules


Optimize Custom Code for SAP HANA Shift in priorities of classic golden SQL rules for SAP HANA

Shift in priorities for SAP HANA

Avoid selecting large numbers of not

required columns

Prefer array operations for INSERT,

UPDATE und DELETE when changing

many records

Avoid nested SELECT loops

In most cases SAP HANA does not require

secondary indices

Keep unnecessary load away from the

database

Code pushdown for data-intensive

calculations to benefit from SAP HANA


Optimize Custom Code for SAP HANA Classic Static Performance Checks

Most golden SQL rules are covered by static checks already

Gaps for “Minimize amount/number of data transfer(s)”


Optimize Custom Code for SAP HANA New Static Performance Checks I

Improved and new static performance checks:

SELECT * - Used columns vs. selected columns

Nested SELECT statements (across modularization units)

SELECT + SELECT FOR ALL ENTRIES

which can be transformed into JOIN or SUBQUERY

SELECT * check

No field used – pure existence check

Nested SQL check

SELECT in LOOP (across

modularization unit)

SELECT + SELECT FOR ALL ENTRIES

transform into JOIN or SUBQUERY


Optimize Custom Code for SAP HANA New Static Performance Checks II

New checks close gaps for rules “Minimize

amount/number of data transfer(s)”

Check appendix for availability of the new

checks.

NW 740 SP3


Optimize Custom Code for SAP HANA Static Performance Checks – Efficient analysis of the findings

NW 740 SP3

“Shall I analyze and correct each and every nested SELECT or

SELECT * …? This may destabilize code which is not even

performance critical…”

In general static performance checks:

Have a higher false positive ratio as functional checks

Provide mainly performance hints - not all findings are errors

Should be combined with performance/runtime data to reduce

the false positive ratio

The findings of the following static performance

checks shall be analyzed like functional errors:

SELECT, UPDATE, DELETE without any WHERE

Unsecure use of FOR ALL ENTRIES

Table attribute checks

Check DB table settings (especially buffer characteristic)


Custom Code Base

Optimize Custom Code for SAP HANA Static Performance Check – Efficient analysis I

Covered by Static Checks

Concentrate on the findings

in performance hotspots

We need runtime data from

the productive system to find

the relevant code

Unused

Code

Dynamic

Code

Relevant

for performance

optimizations

Not relevant

for performance

optimizations


Optimize Custom Code for SAP HANA Static Performance Check – Efficient analysis II

1. Use Usage Procedure

Logging (UPL) to separate

not used (dead) code

VISIT ITM114: Real Software Utilization with Usage and

Procedure Logging

Use monitoring and trace

tools to measure the usage

of ABAP and SQL in your

productive system

3. Find the performance

hotspots with the SQL

Monitor

Dynamic

Code

Unused

Code

Not relevant

for performance

optimizations

Relevant

for performance

optimizations

Covered by monitoring tools

2. Use the standard

performance trace tools

(ST05, SAT, …) to analyze

your core business

processes


Example: A productive ERP System with

6000 concurrent users

Two weeks in this system means • 50k different business processes

• 130k different ABAP SQL statements

(>15% in customer code)

• 13B SQL requests (~1B / day)

• 140B records (~10B / day) read or changed

Where to start an optimization …?


New SQL Monitor allows to get

performance data for each and every

OPEN SQL executed in production...

Answer questions like • What are the most expensive and most

frequently executed SQLs?

• Which SQL reads/writes millions of

records?

• What is the SQL profile of my business

process VA01?

Total transparency of the SQL

running in production!


Business process list ranked by total DB time

Summary - Optimize Custom Code for SAP HANA SQL Monitor – DB performance profile of the business processes

SQL Monitor

• Displays performance data on process level (report, transaction, …)

• Allows to drill down from process level to the single DB operations

• Can run in production with minimal performance overhead (<3%)

Drill down to the SQL profile of

report ZSQLM_TEST

(98 SQL statements)

Navigate to the code


Summary - Optimize Custom Code for SAP HANA SQL Monitor

SQL Monitor collects performance data for each and

every SQL executed in an ABAP application

SQL Monitor can run in a productive system without

disturbing business processes

SQL Monitor allows to link the monitored SQL to the

driving business process

SQL Monitor can be used in the “old” productive system

before doing the HANA migration

SQL Monitor availability is starting with NW 700 (see

appendix for details)


Optimize Custom Code for SAP HANA Static Performance Check – Efficient analysis III

Covered by Static Checks

Processes covered by SQL

Monitor after a week / month

running in production. Quarter-

end or year-end processes

may not be covered

Relevant for

performance

Dynamic

Code

Unused

Code

Not relevant

for performance

optimizations

Combine SQL Monitor data and

Code Inspector / ATC check

data to rank the findings and for

complete coverage

Eliminated by UPL


Optimize Custom Code for SAP HANA Guided performance optimization

New SQL monitor (SQLM)

Can run in production; also before migration

Captures aggregated runtime data (number of executions,

database times, tables, transactions)

Runtime

data

Correlate and optimize

Define context from business perspective (entry point,

transaction, involved data)

Find code with best cost/benefit ratio

Fix performance bugs and “unexpected” database

accesses

Find code which can be “pushed down” to SAP HANA

(e.g. via a SQL script procedure)

SQL Performance Tuning Worklist (SWLT)

Improved / new code inspector checks, e.g.

Nested SELECT statements across modularization units

Used columns vs. selected columns

Static code

analysis


Optimize Custom Code for SAP HANA Guided performance optimization – System view

Production System

Database

Development / Q- System

Database

SQL Monitor

ABAP Test Cockpit

SQL Performance Tuning Worklist

Download / RFC


Tune your custom code – Get ready for SAP HANA Summary – Process view – Tool usage

Use ABAP Test Cockpit for your general quality assurance (avoid

regressions)

Use SQL Monitor in the SAP HANA test system ( + SWLT in the

development system) to tune the top custom SQL statements of the

running test cases

Use ABAP Test Cockpit for your general quality assurance

Use SQL Monitor in the SAP HANA productive system ( + SWLT in

the development system) to iteratively tune the custom SQL

statements and to detect and realize HANA potential

Use ABAP Test Cockpit to correct potential functional issues in your

custom code (“old” system or sandbox system)

Use SQL Monitor in the „old“ productive system ( + SWLT in the

development system) to correct custom SQL issues / tune the top

custom SQL statements in your system

Eliminate unused custom code to reduce migration effort

Define your main business processes for testing and tuning

Evaluation in sandbox system

Plan SAP HANA

migration

Prepare SAP

HANA migration

SAP Business Suite

Any Database

Development / Test System

SAP HANA Database

Upgrade

Migrate

SAP HANA

migration

SAP Business Suite

Any Database

Production System

SAP HANA Database

Upgrade

Migrate

SAP Business Suite

Any Database

Sandbox System

SAP HANA Database

Upgrade

Migrate


Tune your custom code – Get ready for SAP HANA Short overview – Realize SAP HANA potential - Code pushdown

AS ABAP

SAP HANA Database

Calculation

Calculation

“Data to Code” “Code to Data”

Code pushdown means delegating data intense calculations to the

database layer e.g. by using advanced Open SQL, advanced ABAP views

and SQL Script


Further Information

SAP Education and Certification Opportunities

www.sap.com/education

Watch SAP TechEd Online

www.sapteched.com/online

SAP Public Web

ABAP Test Cockpit: http://scn.sap.com/docs/DOC-31773

ABAP Performance guidelines for SAP HANA:

http://scn.sap.com/community/abap/hana/blog/2013/03/24/performance-guidelines-for-abap-development-on-the-sap-hana-database

Community ABAP4HANA: http://scn.sap.com/community/abap/hana

Central SAP Notes

1885926, “ABAP SQL Monitor”

1794297: “Secondary Indexes for the business suite on HANA”

1785057: “Recommendations for migrating suite systems to SAP”

1622681: “DBSL hints for SAP HANA”

1847431: “SAP NetWeaver BW ABAP Routine Analyzer”

1051576: “Conversion of Single Code Page Systems to Unicode”

1812713: “Add-on compatibility of SAP NW AS ABAP 7.4 for Suite”

1760306: “SAP EhP6 for SAP ERP 6.0, version for SAP HANA 1.0”

1855666: “Suite on HANA : 3rd party Add-ons



http://www.sap.com/education

http://www.sapteched.com/online

http://scn.sap.com/docs/DOC-31773

























Appendix Availability "Guided performance optimization"


Tools in the context of guided performance optimization Availability (1/4)

ABAP Test Cockpit

The ABAP Test Cockpit (ATC) is a tool for doing static and dynamic quality checking of ABAP code and associated repository objects.

SAP_BASIS 7.00 SAP_BASIS 7.01 SAP_BASIS 7.02 SAP_BASIS

7.10, 7.11, 7.20,

7.30

SAP_BASIS 7.31 SAP_BASIS 7.40

SAP_BASIS not planned not planned SP12 not planned SP5 SP0


7.10, 7.11, 7.20,

7.30


SAP_BASIS SP0 SP0 SP0 SP0 SP0

SP0

Code Inspector

The Code Inspector is a tool to check repository objects in terms of performance, security, syntax and compliance with naming conventions.


Tools in the context of guided performance optimization Availability of new static Code Inspector checks (2/4)

Area Check title Availability (SAP_BASIS Version)

Functional / Robust

Code

Search problematic statements for result of SELECT/OPEN CURSOR without ORDER BY

740 SP3

740 SP2 via special transport(note 1875529)

731 SP9

702 SP14

Functional / Robust Depooling/Declustering: Search SELECT for Pool/Cluster-Tables w/o ORDER BY

(Priority/Significance of this check is reduced when the more sophistcated check “Search problematic

statements for result of SELECT/OPEN CURSOR without ORDER BY” is used)

740 SP2

731 SP9

702 SP14

Functional / Robust

Code

Unsecure use of FOR ALL ENTRIES 740 SP2

731 SP5

702 SP12

701 SP14

700 SP29

Performance Search problematic SELECT * statements 740 SP2

731 SP9

702 SP14

Performance Search SELECT .. FOR ALL ENTRIES-clauses to be transformed 740 SP2

731 SP9

702 SP14

Performance Search DB Operations in loops across modularization units (reading and changing SQLs)

740 SP2:

Search SELECTs in loops across modularization units

Changing Database Accesses in Loops (no cross stack analysis !)

Old releases:

SELECTs in Loops (no cross stack analysis !)

Changing Database Accesses in Loops (no cross stack analysis !)

740 SP3

731 SP9

702 SP14

Performance Search SELECT statement with DELETE statement 740 SP2

731 SP9

702 SP14



SQL Monitor

The SQL Monitor is a monitoring tool for ABAP based SAP systems which allows to collect performance data of all executed SQL statements in a

productive environment. It exists in two flavours: one is delivered via SAP_BASIS, one via ST-PI plug-in. In release 7.02, 7.31 and 7.40 we recommend to

use the SQL monitor delivered via SAP_BASIS.

SQL Performance Tuning Worklist

The SQL Performance Tuning Worklist allows to combine static performance check results with relevant runtime data coming e.g. from a productive

system.

SAP_BASIS

7.00, Kernel 7.21

(PL>118)

SAP_BASIS

7.01, Kernel 7.21

(PL>118)

SAP_BASIS

7.02, Kernel 7.21

(PL>118)

SAP_BASIS

7.10, 7.11, 7.20,

7.30, Kernel 7.21

(PL>118)

SAP_BASIS

7.31, Kernel 7.21

(PL>118)

SAP_BASIS

7.40, Kernel 7.40

SAP_BASIS not planned not planned SAP_BASIS

SP14

not planned SAP_BASIS SP9 SAP_BASIS SP2

ST-PI ST-PI 2008_1_*

SP8

ST-PI 2008_1_*

SP8

ST-PI 2008_1_*

SP8

ST-PI 2008_1_*

SP8

ST-PI 2008_1_*

SP8

ST-PI 2008_1_*

SP8


7.10, 7.11, 7.20,

7.30


SAP_BASIS not planned not planned SP14 not planned SP9 SP2



Further remarks Background information for the ABAP Test Cockpit can be found here:


The current SAP Support Package Schedule can be found here:

https://service.sap.com/~sapidb/011000358700000294692004E

When using the SQL Monitor delivered via ST-PI, in all releases < SAP_BASIS 7.40 OSS

notes 1806015 and 1831460 must be applied

In release SAP_BASIS 7.40, SP2 OSS note 1834930 must be applied before using the SQL

Monitor





















Summary

ABAP is and will continue to be the basis for applications

like SAP Business Suite, SAP NetWeaver BW, new applications

on-premise as well as in virtual (private) cloud environments

AS ABAP 7.4 facilities leveraging SAP HANA features

by pushing down calculations to the database layer

by means of SAP HANA content integration as well as advanced ABAP database

programming

SAP HANA offers many new possibilities for ABAP-based applications

Examples: application function libraries, business rules, text search

Get started with ABAP on SAP HANA today! Use our offers and trial systems!


Further Information

SAP Education and Certification Opportunities

http://training.sap.com/v2/course/ha400-abap-programming-for-sap-hana-classroom-099-g-en/



SAP Public Web

Developer Center: http://scn.sap.com/community/developer-center/abap

ABAP Development Tools for SAP NetWeaver: http://scn.sap.com/community/abap/eclipse

ABAP for HANA: http://scn.sap.com/community/abap/hana

ABAP 7.4 Trial Editions: http://scn.sap.com/docs/DOC-41566





















http://scn.sap.com/community/developer-center/abap



http://scn.sap.com/community/abap/eclipse





Documents

Chicago Inside Track Application Development Based on ABAP and SAP HANA (1)