An interesting listener issue

From: Mike Haddon <m.haddon_at_tx.rr.com>
Date: Mon, 17 Oct 2011 17:45:35 -0500
Message-ID: <4E9CB00F.1000903_at_tx.rr.com>

Today I was brought into a performance issue on one of our databases. There were a couple of resource issues and during the analysis I noticed that the listener was attempting to be started when it is already running. The database is 9.2.0.8 (yeah, I know, it supports a legacy app), and the listener is on default port.

The problem is that I continue to see the system try to start the same listener. This of course fails because the port is already open but the attempt is made again in a minute or two.

*Hostname->SID:> ps -ef | grep lsnr

   oracle 3768386 1 120 May 27 - 13233:59
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit

   oracle 4231390 3805240 1 17:23:10 pts/0 0:00 grep lsnr Hostname->SID:> ps -ef | grep lsnr

   oracle 3768386 1 120 May 27 - 13234:00
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit

   oracle 5882040 1 102 17:23:12 - 0:00
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit

*Wait a few minutes
*
ps -ef | grep lsnr

   oracle 3768386 1 60 May 27 - 13235:58
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit
**Hostname->SID**:> r ps
ps -ef | grep lsnr

   oracle 3768386 1 120 May 27 - 13235:58
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit
**Hostname->SID**:> r ps
ps -ef | grep lsnr

   oracle 3768386 1 120 May 27 - 13236:00
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit

   oracle 5042184 1 17 17:30:14 - 0:00
/u00/app/oracle/product/9.2.0.8/bin/tnslsnr LISTENER -inherit

*On top of this the original instance of the listener is taking up a complete CPU on the system.

*Name            PID  CPU%  PgSp Owner
tnslsnr     3768386  27.2 782.2 oracle

*
I have checked everywhere for the reason the listener is being started again. There are no cron jobs, at jobs, running processes, dbms jobs that would be trying to restart. The only thing I can think of is maybe the listener is bound and trying to restart itself. We do not have Oracle Restart installed and configured so I am at a lack.

I will be shutting down and restarting the listener later this evening but this is such a vexing issue I don't want to walk away from this issue without truly understanding what is happening. I think it is related to the size of the listener log (> 2GB) and will be resetting the log this evening.

Has anyone seen this before? The OS is AIX 5.3 (LPAR) with 9.2.0.8 Oracle binaries.

Any ideas would be very welcome.

Thanx in advance,

Mike

--
http://www.freelists.org/webpage/oracle-l

Received on Mon Oct 17 2011 - 17:45:35 CDT