Re: about "cursor: pin S wait on X" wait event.
Date: Wed, 27 Aug 2008 13:18:11 -0700 (PDT)
On Aug 27, 2:09 pm, "Shakespeare" <what..._at_xs4all.nl> wrote:
> "Helma" <helma.vi..._at_hotmail.com> schreef in berichtnews:557bdb59-e225-4844-9b5c-96e897673ecb_at_r15g2000prh.googlegroups.com...
> > Subject: Bug 5907779 - Self deadlock hang on "cursor: pin S wait on
> > X" (typically from DBMS_STATS)
> > Note:5907779.8
> > a bug on 10.2.0.3.
> In spite of the docs, we see a lot of "cursor: pin S wait on X" wait events
> occurring in a 10.2.0.4 db as well, causing (according to RDA/AWR reports)
> the main portion of out wait events.
> We are looking at it right now, will report back!
I always see the notes on Mtealink for DBMS_STATS, but on 10.2.0.3 we see it periodically for no *apparent* reason on a new three node x86-64 cluster. We have seen up to five wall clock days of wait time for this even in a half hour AWR period in the last month.
We haven't yet set the "super spy decoder ring" hidden parameter (_kks_use_mutex_pin), but do have a sev 3 SR open to see if Oracle recommends it. Received on Wed Aug 27 2008 - 15:18:11 CDT