G
George Potemkin
Guest
> and the rest is probably related to buffers needing to be reloaded from disk. IMHO, it's unlikely. Did you get the promon statistics for both periods of time? Like: Activity: Summary Activity: Performance Indicators Activity: Latch Counts I'm used to ask our customers to gather the statistics for 30-sec period using my dbmon script and dbstatmon.p program. This information is enough to identify the bottleneck and to find out an application's code that caused the problems.
Continue reading...
Continue reading...