Re: intermittent library cache pin/lock problems

NetComrade wrote:

>I also don't fully understand why the pin would even be needed for sql
>statements that appear to be frequently executed and use bind
>variables (i've seen a whole bunch of those for the sessions waiting
>for library cache pin/load lock)

Because that's what the CURSOR_SPACE_FOR_TIME tells it to do. Shared SQL areas need to be pinned as long as anyone is accessing them.

>How does one determine if the pool (library cache) is fragmented, or
>look for other degradation pointers? I feel there is too much 'magic'
>surrounding the shared pool. There isn't much description of
>'internals' exactly what latch (children) are needed for what
>operation in the documentation.

You got that right.
https://metalink.oracle.com/metalink/plsql/ml2_documents.showFrameDocument?p_database_id=NOT&p_id=61623.1 maybe you are invalidating things somehow?

jg

--
@home.com is bogus.
"If love is blind, why do people like lingerie so much?" - Nick Brooke
"What's love got to do with it?" - Tina Turner