Thursday, December 17, 2015

RDBMS/Exadata/RAC -- What do we recommend for recoverability?

Here is a list of techonologies and configurations which we recommend for having an optimal recoverability according to the general RPO and RTOs defined in Customer environments.
Ofcourse, these are just titles but we have lots of things to say about these subjects as well. 
On the other hand, we are going into the details when we are doing recoverability assessments.. So this is not the place for giving the details:)
  • Implement Oracle RAC
  • Run services on multiple instance Oracle RAC 
  • Use ASM for placing the Oracle Database files
  • Contact third party vendors for storage based replication technologies and ensure best practices are implemented
  • Place the redolog files in multiple mount points
  • Place the control files in multiple mount points 
  • Implement Dataguard and utilize Standby Database (one local , one remote)
  • Configure archive_lag_target (for standby)
  • Use Enterprise Manager
  • Use synchronized Dataguard in the local site
  • Use DataGuard Broker:
  • Document the restore, recovery and standby failover,switch operations
  • Enable Flashback Database option
  • Use Guaranteed recovery points
  • Use Retention guarantee for Flashback Queries: 
  • Use automated Backups
  • Increase Backup & Recovery speed 
  • Transfer Database Backups to multiple Locations
  • Configure a recovery window based backup retention:
  • Build the restore, recovery, failover task flow , define the communication methods and document everything
  • Test restore, recovery and disaster scenarios
  • Periodically run orachk or exachk(if it is an Exadata) , and run the reported recoverability related problems.
  • For Exadata : Ensure  Exadata best practices documented in the whitepaper named “Backup and Recovery Performance and Best Practices for Exadata Cell and Oracle Exadata Database Machine” implemented or at least make plans for implementing them
  • Example to Recoverability related problems (exachk):
o   One or more Ethernet network cables are not connected.
o   Database parameters log_archive_dest_n with Location attribute are NOT all set to recommended value
o   Storage Server alerts are not configured to be sent via email
o   Active system values should match those defined in configuration file "cell.conf"
o   System is exposed to Exadata Critical Issue DB18
o   System is exposed to Exadata Critical Issue DB23
o   System is exposed to Exadata Critical Issue DB24
o   System is exposed to Exadata Critical Issue EX19
o   Grid Version is older than the recommended
o   Storage Server Version is older than the recommended
o   Exadata bundle patch  Version is older than the recommended
o   Database parameter LOG_BUFFER is NOT set to recommended value
o   fast_start_mttr_target should be greater than or equal to 300
o   The data files should be recoverable
o   Flashback on PRIMARY is not configured
o   Primary database is NOT protected with Data Guard (standby database) for real-time data protection and availability
o   RMAN controlfile autobackup should be set to ON
o   Oracle database does not meet certified platinum configuration

No comments :

Post a Comment