Oracle Database 12c High Availability · Cloud and High Availability Strategy •Oracle provides both a complete suite of cloud services, and the infrastructure that are used to build

Oracle Database 12c High Availability

Wei Hu Vice President of Development High Availability Technologies, Oracle Corporation

Vitor Pacheco Associate Director, Data Services Amadeus Data Processing GmbH September 29, 2014

Copyright © 2014, Oracle and/or its affiliates. All rights reserved. |

For Consolidation and Cloud Deployments


Consolidation and Cloud Deployments

• Consolidating databases into a private cloud can lower cost and increase agility – Challenge: Consolidation into private cloud requires placing many databases into

a shared infrastructure to simplify management and control cost. If that platform goes down, so do all of the databases that run on it

– Opportunity: It is much easier to make a smaller number of platforms highly available... if they are built using the right infrastructure

• Oracle Database 12c can help you build a highly available platform required for consolidation and private cloud

• 12.1.0.2 is the ruggedized version of 12c

Challenge and Opportunity


Cloud and High Availability Strategy

• Oracle provides both a complete suite of cloud services, and the infrastructure that are used to build these cloud services –Oracle’s cloud offerings are robust because of our best-in-class high

availability technologies –Oracle’s high availability technologies are battle-tested by being

massively deployed in our cloud offerings

• You, our customers, benefit when you use our cloud services or the underlying HA technologies

• Oracle MAA – a Cloud-tested architecture

Achieve Synergy through the Cloud


Edition-based Redefinition, Online Redefinition, Data Guard, GoldenGate – Minimal downtime maintenance, upgrades, migrations

Active Data Guard – Data Protection, DR – Query Offload

GoldenGate – Active-active replication – Heterogeneous

Active Replica

RMAN, Oracle Secure Backup, Recovery Appliance – Backup to disk, tape or cloud

Enterprise Manager Cloud Control – Site Guard, Coordinated Site Failover Application Continuity – Application HA Global Data Services – Service Failover / Load Balancing

RAC – Scalability – Server HA

ASM – Local storage

protection

Production Site

Flashback – Human error

correction

Cloud-tested Architecture Oracle Maximum Availability Architecture


Program Agenda

Active Data Guard

Global Data Services

Application Continuity

Backup and Recovery

Other HA Enhancements

MAA at Amadeus

1

2

3

4

5

6

6


Replication Required for Best Availability

• A consolidated database (or private cloud) must be physically replicated across multiple data centers

• Good news: as you consolidate, it becomes even more cost effective to deploy disaster recovery – Fewer databases, fewer data centers, lower

operational costs

• Best fault tolerance is achieved by replicating across long distances

• But long-distance replication can affect performance

7


Speed-of-light and Zero Data Loss

• Everyone wants to run Data Guard with zero data loss (Maximum Availability Mode), but are worried about performance impact of synchronous communication

• The longer the distance, the larger the performance impact

Performance Impact of Long Distance Replication

Primary Standby

Commit

Commit Ack

Network Send

Network Ack


Primary

Standby

Minimal, But Some Data Loss Exposure on Disaster Compromise by running Data Guard Customers in Async mode

ASYNC


Active Data Guard Far Sync

• Far Sync is a light-weight Oracle instance with a standby control file, standby redo logs, archived redo logs, but no data files

• Receives redo synchronously from primary, forwards redo asynchronously in real-time to standby

• Upon failover the asynchronous standby transparently obtains the last committed redo from the Far Sync instance for zero data loss failover

• A second Far Sync instance can be pre-configured to transmit in reverse direction after failover/switchover

Zero Data Loss For Async Deployments


Primary

Standby

Operational Flow Active Data Guard Far Sync

Far Sync instance

SYNC (no data loss)

ASYNC


Primary

Standby

Operational Flow Active Data Guard Far Sync

ASYNC

Zero Data Loss

No Compromise Between Data Protection and Performance!

Remaining Redo Blocks

Far Sync instance

New Primary


Active Data Guard Far Sync

• Best data protection, least performance impact

• Low cost and complexity

• Best way to implement a near DR + Far DR model

• Every Data Guard ASYNC configuration can benefit from using this

• Ability to do a no data loss failover improves your availability – since you will be able to failover immediately on any issues

• Benefits both types of DR cloud deployments – Centralized on-premise DR Service and DR in the public cloud

– In both cases, the DR center will likely be far from all the primary data centers

Benefits


Active Data Guard Fast Sync

Fast Sync • Remote site acknowledges received redo

before writing it to standby redo logs • Reduces latency of commit on primary • Better DR – increased SYNC distance • If network round-trip latency is less than

time for local online redo log write, synchronous transport will not impact primary database performance

• Fast Sync is supported with Fast-Start Failover and/or Far Sync

Reduced Primary Database Impact for Maximum Availability

Primary Standby

Redo Logs

Standby Redo Logs

Commit

Commit Acknowledge

NSS

Commit

Commit Acknowledge

Acknowledge returned on receipt

LGWR

Primary Standby

Redo Logs

Standby Redo Logs

NSS RFS LGWR

RFS SYNC

FAST SYNC


Other New Features in Oracle Database 12c Data Guard

Rolling Upgrade With Active Data Guard Automate complexity through simple PL/SQL Package: DBMS_ROLLING (12.1.0.2 onwards), with simple Init, Build, Start, Switchover, Finish procedures

Additional Data Type Support: XML OR, Binary XML, Spatial, Image, Oracle Text, DICOM, ADTs (simple types, varrays), …

Validate Role Change Readiness Ensure Data Guard configuration ready for switchover with

automated health checks – verify no log gaps, perform log switch, detects any inconsistencies, ensures online log files cleared on standby, …

DML on Global Temporary Tables Temporary undo is not logged in redo logs

Enables DML on global temporary tables on Active Data Guard: more reporting support

Set by default on Active Data Guard standby

Unique Sequences Primary allocates a unique range of sequence numbers to

each Standby

Enables more flexible reporting choices for Active Data Guard


EM Database Rolling Upgrade Assistant

16

Automation for Active Data Guard Rolling Upgrades

Copyright © 2014, Oracle and/or its affiliates. All rights reserved. | 17

It Just Works Better – Seconds to Failover Data Guard Failover and Switchover Times

Failover Single Instance Mount

Single Instance with Active Data Guard RAC Mount RAC with

Active Data Guard

18 n/a 25 n/a

11.2 7 13 18 44

12.1.0.2 7 8 9 15

Switchover

10.2 52 n/a 55 n/a

11.2 20 31 38 54

12.1.0.2 11 13 15 21


Oracle Database Cloud Service RMAN, Data Guard, and Active Data Guard

• Available today: RMAN backup and recovery

• Coming soon: local Data Guard and Active Data Guard standby databases

• Planned in calendar year 2015: – Active Data Guard for remote DR – Off-load RMAN backup to standby using fast

incremental backups – Database rolling upgrades and standby-first

patching using physical standby – 12c Global Data Services to load balance and manage

services across replicated databases – Active Data Guard Far Sync -- zero data loss

protection at ANY Distance Compute Storage

Compute Cluster

/u04 /u03 /u02 /u01

Compute

OVM GlassFish

REST Listener RDBMS

Cloud Utilities

Compute Cluster

/u04 /u03 /u02 /u01

Compute

OVM GlassFish

REST Listener RDBMS

Cloud Utilities

Redo Log Shipping

Primary Standby

Customer Tenant On Premise

R/W R/O


A A

Disaster Recovery Automation with Oracle Site Guard

• Automates entire stack switchover or failover between sites – Web, application, database and storage

– Data Guard and storage replication

– Manages start order dependencies

• Validated with: – Oracle Databases

– Oracle Fusion Middleware

– Oracle Fusion App deployments

– Extensible for 3rd party components (IBM WebSphere, SAP etc.)

• Part of the effort to industrialize HA/DR for managing a fleet of application deployments

Being Deployed with Oracle SaaS

Site A Site B

Data Guard/ Active Data Guard Transport

Storage Replication

Oracle Site Guard Enterprise Manager Cloud Control

ZFS Storage

Appliance

ZFS Storage

Appliance

DB Host1

DB Host2

Web Tier

App Tier

DB Tier

Web Host1

Web Host2

App Host1

App Host2

DB Host1

DB Host2

Web Tier

App Tier

DB Tier

App Host1

App Host2

Web Host1

Web Host2


Program Agenda

Active Data Guard



Backup and Recovery


MAA at Amadeus

1

2

3

4

5

20

6


Database Service in Cloud and Consolidation

• Fundamental requirement of consolidation and cloud deployment is to isolate applications from where the database service is physically offered

• Applications connect to a logical database service – Service can be offered by instances in a RAC database, replicated databases, local or remote data

centers, in a PDB or dedicated database – The set of physical database that offers a service can change due to failures, load, and configuration

changes – Service can be moved around (e.g., migrate a database from a dedicated server onto a consolidated

platform in another data center)

• When a service fails, the application should be transparently directed to a new instance of this service, whether it is running locally or in another data center

• Oracle Database 12c provides this using Global Data Services

Location Transparency and Handling Database Outages

21



• Before Oracle 12c, a Database Service was designed to run within a single RAC database

• Oracle 12c introduces Global Database Services that can run across multiple RAC and non-RAC databases – Provides RAC-style service failover and load balancing across

multiple databases within and across data centers

– Supports Data Guard, GoldenGate, and non-RAC databases

– Takes into account network latency, replication lag, and service placement policies

• Achieves higher availability, improved manageability and maximize performance

Service Failover and Load Balancing for Replicated Databases



Global Data Services • Global Services running across primary and standby

– Reporting Service: read-only service available on primary and standby – Call Center Service: read-write service running on primary database

• Clients connecting to the Reporting Service are routed to either primary or standby database – Based on location, response time, data, acceptable data lag

– Reports will automatically run on least loaded server – If preferred database not available, will route connections to another

database

• Global service migration – Automatically migrates services based on failover/switchover. Will

automatically start services on the most appropriate replica according to policies configured by the administrator

• Works similarly with GoldenGate – including support for GoldenGate configurations using active-active replication

Active Data Guard Example

Active Data Guard

Call Center Services

Reporting Services



• Key Benefits – Provides the location transparency that allows

you to consolidate and/or move databases into a cloud

– Allows your applications to more easily use Active Data Guard and GoldenGate for processing

• Improves resource utilization, agility, and manageability for Active Data Guard and GoldenGate deployments

Summary of Benefits



Program Agenda

Active Data Guard



Backup and Recovery


MAA at Amadeus

1

2

3

4

5

25

6

Copyright © 2014, Oracle and/or its affiliates. All rights reserved.

In-flight works continues • Replays in-flight work on recoverable errors • Masks hardware, software, network, and

storage failures and planned maintenance • Supports JDBC-Thin, Universal Connection

Pool (UCP), WebLogic Server • Oracle 12.1.0.2 adds support for IBM

Websphere, Apache Tomcat (using native PooledConnection interface), and other mid-tiers when using UCP

• Included with RAC, RAC One, & Active Data Guard

26


Application Servers

Database Servers

End User

Database Request Replayed

Mask Unplanned and Planned Failures

Copyright © 2014, Oracle and/or its affiliates. All rights reserved.

Known outcome after failure • Tracks outcome of the last transaction • Without Transaction Guard, retry can

cause logical corruption • Transaction Guard allows applications to

deal correctly with failures • Application Continuity uses Transaction

Guard • Also available for applications that do not

use Application Continuity

27

Transaction Guard

Application Servers

Database Servers

End User

Preserve and Retrieves Known Commit Outcome


Improvements to Services and Planned Maintenance

• Continuous Connections. Oracle clients delay retries while a service is failing over to eliminate connect failures for RAC and Data Guard. In 12.1.0.2 client, backported to 11.2.0.4

• Graceful relocation of sessions away from node during maintenance. Supports Oracle connection pools in conjunction with RAC (10g and later) and with Oracle 12c Global Data Services (Data Guard, GoldenGate, RAC and non-RAC databases)

• FAN (Fast Application Notification) notifies clients when a database service fails or moves (network, maintenance, hangs, joins). This allows clients to rapidly react. – Auto-configuration for FAN out-of-the-box – Global Data Services supports FAN on Active Data Guard, GoldenGate, RAC, single instance

• Cloud-tested: Oracle Database 12.1.0.2 just works better

28


UCP 12.1.0.2: Transaction rate unchanged during service migration OLTP Workload During Service Migration

0 500

1000 1500 2000 2500 3000 3500 4000 4500 5000

18:4

2:14

18

:47:

02

18:5

1:49

18

:56:

37

19:0

1:24

19

:06:

16

19:1

1:04

19

:15:

53

19:2

0:41

19

:25:

31

19:3

0:18

19

:35:

07

19:3

9:54

19

:44:

41

19:4

9:32

19

:54:

20

19:5

9:07

20

:03:

55

20:0

8:42

20

:13:

30

20:1

8:18

20

:23:

06

20:2

7:53

20

:32:

42

20:3

7:46

20

:42:

34

20:4

7:21

20

:52:

08

20:5

6:56

21

:01:

44

21:0

6:31

21

:11:

18

21:1

6:07

21

:20:

54

21:2

5:42

21

:30:

29

21:3

5:18

21

:40:

06

21:4

4:53

21

:49:

40

21:5

4:28

21

:59:

17

22:0

4:04

22

:08:

52

22:1

3:39

22

:18:

28

22:2

3:15

22

:28:

08

22:3

2:55

22

:37:

46

TPS

0

1000

2000

3000

4000

5000

6000

21:4

0:56

21

:45:

39

21:5

0:22

21

:55:

05

21:5

9:49

22

:04:

32

22:0

9:15

22

:13:

58

22:1

8:41

22

:23:

24

22:2

8:07

22

:32:

50

22:3

7:33

22

:42:

17

22:4

7:00

22

:51:

43

22:5

6:26

23

:01:

09

23:0

5:52

23

:10:

35

23:1

5:18

23

:20:

01

23:2

4:45

23

:29:

28

23:3

4:11

23

:38:

54

23:4

3:37

23

:48:

20

23:5

3:03

23

:57:

46

0:02

:29

0:07

:13

0:11

:56

0:16

:39

0:21

:22

0:26

:05

0:30

:48

0:35

:31

0:40

:14

0:44

:57

0:49

:40

0:54

:24

0:59

:07

1:03

:50

1:08

:33

1:13

:16

1:17

:59

1:22

:42

1:27

:25

1:32

:08

1:36

:52

UCP 12.1.0.2

UCP prior to 12.1.0.2

TPS


DBA Operation Maintenance Result Time to Drain all Sessions

RAC rolling PSU apply using opatch No errors to application 5s

RAC rolling Instance parameter change No errors to application 7s

Data Guard switchover Site maintenance No errors to application 29s

Data Guard switchover Site maintenance fallback No errors to application 25s

Planned Failover with FAN WebLogic Server Active GridLink and Real Application Clusters


Program Agenda

Active Data Guard



Backup and Recovery


MAA at Amadeus

1

2

3

4

5

31

6


Best Way to Backup your Consolidated Database and Private Cloud Zero Data Loss Recovery Appliance (ZDLRA)

Zero Impact Backups

Production only sends deltas. All backup and tape processing offloaded

Zero Data Loss Real-time redo shipping ensures data protection in sub-second time

DB Protection as a Service

Easily protect all databases using scalable centralized service

End-to-End Robustness End-to-end visibility, control, and data validation


Zero Data Loss Recovery Appliance Extreme Protection for All Oracle Databases

33

Delta Store • Stores validated, compressed DB changes on disk • Fast restores to any point-in-time using deltas • Built on Exadata scaling and resilience • Single console for end-to-end visibility and control

Zero Data Loss Recovery Appliance (ZDLRA)

Delta Push • DBs access and send only changes • Minimal impact on production • Real-time redo transport for

protection till sub-seconds

Protected Databases

Protects all DBs in Data Center • Scales to 1000s of databases • Petabytes of data, any DB release • No expensive backup agents

Tape Library

Autonomous Tape Archive • Offloads tape backup • Tapes utilized all day

Replicate to Remote ZDLRA or Oracle Cloud


Data Protection As A Service Facilitates Deployment of Private Cloud-Based IT Shared Services

ZDLRA Protection Policies Easy-to-deploy standardized data protection templates, e.g. recovery window, disk retention, etc.

Gold Policy, Mission Critical Disk: 45 days Tape: 90 days Tape

Silver Policy, Business Critical Disk: 30 days Tape: 45 days

Bronze Policy, Test/Dev Disk: 15 days Tape: 30 days

Replica

Possible to set up Protection Policies

also at Replica ZDLRA

34


RMAN Fine-grained Table Recovery From Backup

• Simple RECOVER TABLE command to recover one or more tables (most recent or older version) from an RMAN backup

• Eliminates time and complexity associated with manual restore, recover & export – Enables fine-grained point-in-time recovery

of individual tables instead of the contents of the entire tablespace

RMAN Backups


Simplified Platform Migration RMAN Cross-Platform Backup & Restore

• Simplifies platform migration. – Includes from Big-Endian systems (AIX, HP-UX) to Little-Endian systems (Exadata and Intel X86)

• Minimizes read-only impact with multiple incremental backups • Also useful for instantiating standby databases and GoldenGate replicas

Source Database (AIX)

Backup to Disk/Tape (data files, optional endian conversion, metadata export)

Restore Backup (optional endian conversion, metadata import)

Destination Database (Solaris)


Offsite Backups with Simple and Low Cost Cloud Backup Oracle Database Backup-as-a-Service

Data Center

Backup on-premise or Cloud Databases to Oracle Database Backup Service hosted in Oracle Public Cloud

Cost effective, scalable cloud storage

End-to-end enterprise-grade data encryption, compression and protection

– Clients: Data is encrypted with keys kept locally at client, optionally compressed, and securely transmitted

– Cloud: Encrypted data is protected with 3-way mirroring on every write

DBaaS Oracle Public Cloud

Oracle Public Cloud


Program Agenda

Active Data Guard



Backup and Recovery


MAA at Amadeus

1

2

3

4

5

38

6


Online Redefinition Enhancements Improved sync_interim_table performance

Ability to redefine table with VPD policies

Improved resilience of finish_redef_table

Better handling of multi-partition redefinition


Online Datafile Move Relocate a datafile while users are actively accessing

data: ALTER DATABASE MOVE DATAFILE … Maintains data availability during storage migration

Separation of Duties SYSDG / SYSBACKUP: Data Guard & RMAN specific

administrative privileges

No access to user data: enforce security standards throughout the enterprise

Additional Online Operations Drop index online / Alter index unusable online / Alter index

visible / invisible online

Drop constraint online / Set unused column online

Online move partition: ALTER TABLE … MOVE PARTITION … ONLINE


ASM Overview Oracle Database 11.2 or earlier

ASM Cluster Pool of Storage

Node2

Disk Group B Disk Group A

Node3 Node5 Node4

Shared Disk Groups Wide File Striping

One to One Mapping of ASM Instances to Servers

Node1

ASM Instance

Database Instance

ASM Disk

RAC Cluster

Public 40


Introducing Flex ASM Oracle 12.1: Removal of One to One Mapping

ASM Cluster Pool of Storage

Disk Group B Disk Group A Shared Disk Groups Wide File Striping

Databases share ASM instances

ASM Instance

Database Instance

ASM Disk

RAC Cluster

Node5 Node4 Node3 Node2 Node1 Node5

Node5 runs as ASM Client to Node4




Public 41


Oracle GoldenGate 12c Enhancements Focus Area Enhancements

Oracle Integration Support for Oracle Multitenant, Integrated Replicat for automatic dependency-aware apply, integration with Data Guard for auto-restart of GG components after failover

Improved Ease of Use Simpler configuration of DDL replication, Coordinated Replicat for all databases to allow for better, simpler scalability

Enhanced Security Integration with Oracle Credential Store for database authentication

Enhanced Heterogeneity New capabilities for replication for MySQL, SQL Server, Sybase, Teradata, and DB2.

Cloud Ready Capture from and delivery to the Cloud, both Cloud to Cloud and on-premises to cloud


Active Data Guard and GoldenGate – Complementary • Active Data Guard best for unplanned downtime

– DR capabilities like: Zero data loss, Far Sync, Fast Start Failover – Primary and standbys are physically the same (down to the SCNs) – Good for error detection: corruptions, lost writes – Good for repair: Automatic Block Repair, using a backup from the standby to recover

the primary, and vice versa

• GoldenGate best for planned downtime – GoldenGate replicas are never physically the same – Great when you want to perform planned operations like upgrades – Critical when you want to replicate among different platforms

43


Program Agenda

Active Data Guard



Backup and Recovery


MAA at Amadeus

1

2

3

4

5

44

6

Oracle High Availability in Amadeus Mission Critical Applications

© 2

014

Am

adeu

s IT

Gro

up S

A

Vitor Pacheco Moreira Amadeus Data Processing

Data Services

Page 46

Amadeus is a technology company dedicated to the

global travel industry. We are present in 195 countries with a worldwide team of more than 11,000 people. Our solutions help improve the

business performance of travel agencies, corporations, airlines, airports, hotels, railways and more.

© 2

014

Am

adeu

s IT

Gro

up S

A

Page 46

Amadeus in a few words

Page 47

Our customers

© 2

014

Am

adeu

s IT

Gro

up S

A

AIRLINES

HOTELS C

AR

S

RAIL

FER

RY

CR

UIS

E

INSURANCE

TOUR OPERATORS CONTENT PROVIDERS

TRAVEL INTERMEDIARIES

AIR

PO

RTS

CORPORATIONS

GROUND HANDLERS

Page 48

Amadeus Global Operations

Page 49

Service oriented • Layered • Modular • Composite services • Service self-registration

Highly distributed design • Hundreds of interconnected servers

Scalability • Scale out: Additional servers bring capacity on demand • Scale up: Processing power benefits from evolutions in the semiconductor industry

Open Back End – including Departure Control Systems

Architecture overview

© 2014 Amadeus IT Group SA

Travel Agent Sales

Kiosks DCS Provider Sales

Airport e-Commerce

SERVICES INTEGRATOR Protocol adapter

AUTHENTICATION Access control

APPLICATIONS AND SERVICES

DATABASES BIG DATA FILE SYSTEMS

app1 app2 app3 app4

Hotel Corporations

-3 + 8 0

Miami

London

Sydney

Bangkok

Erding

Bad Homburg

50

Mission Critical Applications Current DB topology _ CRS services in Active/Passive

• Role-switch: internal process to free computing resources

• Rolling PSU and one off-patches _ ASM in normal redundancy with failure

groups in different fire-cells • Volume mirroring leveraged for storage

evolution _ Amadeus MW instrumented for recovery

• Leveraging similar concepts as of TAF and FAN _ Data Guard in Maximum Performance

Mode • Planned switch over using Data Guard Broker: 2

minutes • Exadata for isolation and consolidation • Flashback DB enabled in DG:36 hours/900GB

flashback in less than 30 minutes

Second Data Center (MUC)

Instance 1

Mirror 2

Mirror 1

Instance 2

Data Guard

Data Guard

-3 + 8 0

Miami

London

Sydney

Bangkok

Erding

Bad Homburg

51

Mission Critical Applications DB futures – Isolation and distribution _ Infrastructure enhancements

• Minimize the shared components in order to mitigate impact

_ Oracle features • RAC in a box for rolling patches

• Retaining the manageability for patch management

• ASM and other HA enhancements • Improved manageability of our platforms

• 12c Application Continuity / Transaction Guard • To be studied

Second Data Center (MUC)

Data Guard

Data Guard

PK 1 EMEA

PK 2 ASIA PAC

PK 3 AMER

-3 + 8 0

Miami

London

Sydney

Bangkok

Erding

Bad Homburg

52

Data Guard

PK 1 EMEA

PK 2 ASIA PAC

PK 3 AMER

Data Guard

Data Guard

Data Guard

Mission Critical Applications DB futures – DR in the Cloud _ Oracle features

• 12c Active Data Guard Far Sync Replication • Improved resiliency in cross continent replications

• 12c Global Data Services • Simplification of the App-to-DB connectivity settings • Improved resilience against failures • Workload balancing

• Monitoring purpose • RO workloads

_ … and beyond

• Distribution of Amadeus services in the Cloud • DR services to start • Own and third party data centers

• Licensing model under negotiation

-3 + 8 0

Miami

London

Sydney

Bangkok

Erding

Bad Homburg

53

Mission Critical Applications Transform while perform

Data Guard

PK 1 EMEA

PK 2 ASIA PAC

PK 3 AMER

Data Guard

Data Guard

Data Guard

_ Oracle GoldenGate • Zero (minimum) down-time for upgrades • Bi-directional cross version/endian fundamental

feature to allow the fallback • Reorganization enabler Data dictionary and data

objects

• 12c • Improved manageability

• Tuned upgrade use case • Integration with Global Data Services

• Complements Active Data Guard in MAA configuration

You can follow us on: AmadeusITGroup amadeus.com/blog amadeus.com

Thank you

© 2

014

Am

adeu

s IT

Gro

up S

A


Amadeus – a Cloud MAA Example

• Amadeus is a Travel SaaS provider that relies on MAA and Exadata for mission-critical availability

• Oracle Database 12c allows Amadeus to further improve availability using – Active Data Guard Far Sync to globally distribute their databases – GoldenGate to perform zero-downtime upgrades – Global Data Services to load balance and failover requests across their global data

centers – Lower operational costs through manageability improvements in ASM and RAC, and

more effectively use Active Data Guard for applications

55


Database 12c for Consolidation and Cloud Deployments

• Oracle Database 12c offers high availability (HA) capabilities needed for consolidation and private clouds

• This allows you to benefit from consolidation and cloud, while improving application availability

• Oracle Database 12.1.0.2 is the ruggedized version of 12c

• Cloud-ready and cloud-tested. You get to use the same proven HA technologies that are used in Oracle’s public and private clouds, and mission-critical customers like Amadeus


Resources • OTN HA Portal:

http://www.oracle.com/goto/availability

• Maximum Availability Architecture (MAA): http://www.oracle.com/goto/maa

• MAA Blogs: http://blogs.oracle.com/maa

• Exadata on OTN: http://www.oracle.com/technetwork/database/exadata/index.html

• Oracle HA Customer Success Stories on OTN: http://www.oracle.com/technetwork/database/features/ha-casestudies-098033.html


Documents

Oracle Database 12c High Availability · Cloud and High Availability Strategy •Oracle provides both a complete suite of cloud services, and the infrastructure that are used to build