[Spacewalk-list] Connections Idle in Transaction

[Wojtak, Greg (Superfly)]

There was an SQL select statement you could run inside of postgres to find the queries that were causing those, but I can't seem to find it any more.  Something to do with select * from pg_activity something or other.

Greg Wojtak
Sr. Unix Systems Engineer
Office: (313) 373-4306
Cell: (734) 718-8472


From: Jonathan Scott <lists at xistenz.org<mailto:lists at xistenz.org>>
Reply-To: "lists at xistenz.org<mailto:lists at xistenz.org>" <lists at xistenz.org<mailto:lists at xistenz.org>>, "spacewalk-list at redhat.com<mailto:spacewalk-list at redhat.com>" <spacewalk-list at redhat.com<mailto:spacewalk-list at redhat.com>>
Date: Thursday, September 27, 2012 2:09 PM
To: "spacewalk-list at redhat.com<mailto:spacewalk-list at redhat.com>" <spacewalk-list at redhat.com<mailto:spacewalk-list at redhat.com>>
Subject: Re: [Spacewalk-list] Connections Idle in Transaction

There have been a few recent threads relating to this, but nothing has come out of them yet. It's interesting that more folks are starting to report this problem. I myself am still trying to figure out how I can capture it properly to file a bug report or something more substantial than an email.

- Jonathan

On Thu, Sep 27, 2012 at 1:54 PM, Wojtak, Greg (Superfly) <GregWojtak at quickenloans.com<mailto:GregWojtak at quickenloans.com>> wrote:
I asked this a while back and I don't think I ever saw an answer.  I've got a spacewalk 1.6 install on 64-bit CentOS 6 with Postgresql backend.  Every day I see about 50 postgres processes queue up as "postgres: swuser swdb [local] idle in transaction."

I do a restart every night in order to keep my OSA implementation working (it stops responding if I don't do a restart of the services periodically), and the 'idle in transaction' processes start immediately after that.  Sometimes I get more throughout the day, but it seems the mostly pile up within the first 30 seconds or so of restarting.  I see a lot of slowness while doing certain tasks in spacewalk (especially as it relates to kickstart profiles, which, when I make a change to one, seem to time out and give me a 500 error, at which point the changes I made seem to be committed).

Anyone have any ideas how to troubleshoot fix these?

I saw in another thread about monitoring.  I do not have monitoring enabled in my spacewalk implementation.  The processes do not seem to abate when I stop OSA dispatcher or jabberd, so it doesn't seem OSA related either.

Greg Wojtak
Sr. Unix Systems Engineer
Office: (313) 373-4306<tel:%28313%29%20373-4306>
Cell: (734) 718-8472<tel:%28734%29%20718-8472>


_______________________________________________
Spacewalk-list mailing list
Spacewalk-list at redhat.com<mailto:Spacewalk-list at redhat.com>
https://www.redhat.com/mailman/listinfo/spacewalk-list