2. Getting the Most out of Your Disaster
Recovery Infrastructure
Using Active Data Guard
Manoj Machiwal
Consulting Director
Jade Global
3. AGENDA
• Oracle Database High Availability
• Traditional Disaster Recovery Solution
• Active Data Guard Overview
• Enabling Active Data Guard
• Active Data Guard Cool Features
4. AGENDA (CONTINUED)
• Data Guard Broker Setup
• Maintenance Tips
• Customer Case Study
• Question & Answer Session
6. DATA GUARD
• Both HA/DR Solution
• Protection from Data Corruption
• Addresses Planned and unplanned
Downtime
• Built-in Oracle Integration
• Still Failover site is Redundant, and used
only in case of disaster
7. ACTIVE DATA GUARD
• Read only operation can be
offloaded to standby site
• Backups can be done from standby
site, giving more I/O and processing
power at Primary site
8. ENABLING ACTIVE DATA GUARD
• Using SQLPLUS
If physical standby database is shutdown
Open database read-only and start redo apply
10. ENABLING ACTIVE DATA GUARD
• Using Data guard Broker
• Broker automatically stops the redo apply and resumes after
database open in 11GR2,
• In Release 1, have to manually stop redo apply and restart after
database open
11. ACTIVE DATA GUARD – COOL FEATURES
• Write Operations to Primary
• Automatic Repair of Block Corruption
14. CUSTOMER CASE STUDY
• Industry
– Solar Manufacturing
• Technical foot Print
– 11gR2
– 2 Node RAC
– Exadata Machine
• Application
– Factory Operations
– MES Reporting
15. CUSTOMER CHALLENGES
• Manufacturing Operations and Reporting Operations
running on Same database
• No High Availability Architecture
• Exadata Patching
16. EXISTING ARCHITECTURE
• 24X7 Factory Operations
• MES Reporting
• Backups
• Downtime means, Halt Factory Operations
and have significant impact on business
• Sluggish Performance of Factory Operations
during reporting period
• Can not afford another exadata machine for
HA
• Exadata machine could not be patched
17. PROPOSED ARCHITECTURE
Primary Site Standby Site
• Read Write Factory Operations • MES Reporting
• Backups ( IO performance better on Exadata • Failover Site for Exadata Patching
Machine)
18. LESSONS LEARNED
• Failover Site was not tuned for Factory
Operations
(Exadata/ Non- Exadata Environment)
• Rework on Reporting Queries to tune to
non-exadata environment
• RAT Capture and RAT Replay
19. SUMMARY
• Utilization of Disaster Recovery Infrastructure
• Reporting Load offloaded to standby site
• RAT Capture and RAT Replay should be used
extensively to migrate to Active Data Guard
• Added Benefits – Automatic fixing on Block
Corruption, Write Operations on standby
(using DB links)