'enq: TX - row lock contention' and 10046 trace
- From: Rich <richa03@xxxxxxxxx>
- To: Oracle-L Freelists <oracle-l@xxxxxxxxxxxxx>
- Date: Thu, 29 Nov 2012 08:27:08 -0800
Hi List,
This is 10.2.0.4 on Solaris 10.
We are seeing large amounts of the blocking enqueue 'enq: TX - row lock
contention' on insert of a normal/heap table as well as smaller amounts on
a couple of updates of this table.
We have repro'd the issue in test and have 10046 (level 12) traces of all
of the sessions involved.
From the traces, we see these locks are mostly on one of the
(normal/B-tree) indexes (USERSTATUS), however, some of them are on another
(normal/B-tree) index (ROUTINGSTATUS).
We are trying to consolidate these traces into a timeline. We tried to use
trcsess, however, it requires at least one of the parameters be set. This
application (COTS) is not instrumented such that any of these fields are
populated, so that isn't working very well.
IIRC, there were tools (Perl script? maybe Hotsos/Cary Millsap?) allowing
many tracefiles to be consolidated in relation to time, however, I cannot
seem to find them.
Would someone on this list please direct me to such tools or give me
another avenue to try to determine what the application is doing?
Thanks,
Rich
--
//www.freelists.org/webpage/oracle-l
Other related posts:
