Tuning ORACLE to Minimize Recovery Time: For Solaris Operating System on SPARC
Like this article? We recommend
Like this article? We recommend
Like this article? We recommend
This article provides recommendations for tuning ORACLE on SPARC® processor-based systems running the Solaris™ Operating System (Solaris OS) to minimize recovery in the event of a system or database failure.
The article contains the following topics:
"Background"
"Availability"
"Recovery"
"ORACLE Cache Recovery Tuning"
"Recovery and Performance Measurements"
"Best Practices"
"About the Author"
"Acknowledgments"
"Related Resources"
"Ordering Sun Documents"
"Accessing Sun Documentation Online"
Background
Production systems are evaluated and measured by different criteria, but performance and availability are consistently the critical items evaluated when assessing whether a system is meeting or can meet business requirements. The relationship between performance and availability exists both at the measurement and the configuration and tuning levels.
At the measurement level, we know a that when a system is down, it is providing a performance level of zero. Also, it is possible for a system's performance level to degrade to the point where the system is down from an end-user or business point of view. For example, a system that normally provides a response time of five seconds or less is likely to generate a flurry of user calls for support when response time degrades to two or three minutes.
At a configuration and tuning level, what we do to improve performance can adversely affect availability. Caching large quantities of data in memory and disabling checkpointing can boost performance, but have a devastating affect on recovery time (thus, availability). Such a scenario could increase the probability of data loss in the event of a system crash. Alternatively, aggressive synchronous writes to disk, frequent checkpointing, and sophisticated fault and error handling in application code can make for a very robust platform, but create challenges in meeting performance goals.
Thus, we must seek a balance in tuning production systems to meet both performance and availability requirements. Like any configuration and tuning exercise, it takes careful planning and an understanding of the trade-offs, with repeated measurement and analysis of the target platform and application.