20 September 2007

HALP

Getting slammed this week. My production instance hanging for 10-15 minutes 2-3 times a day during business hours due to latching on the shared pool and library cache. Oracle "support" seems to be reading from a "fix ur app codez kthxbye" script.

This post from Jonathan Lewis describes the symptoms to a tee.

I'll try to follow-up when we get past this.

2 comments:

  1. Are you using automatic memory management?

    We had similar symptoms on two of our systems, and they were resolved after reverting to manual memory management. We did suspect the application at the time (Many of the queries didn't use bind variables), but we did run Confio, so maybe it was actually related to that.

    ReplyDelete