[Spacewalk-list] Spacewalk/Postgresql database issues [RESOLVED]

[Paul Robert Marino]

Good
Glad to hear it you should plan to run the vacuum atleast once a month but
you may find running it more often may be better. It always takes a little
trial and error to get the frequency right witout over doing it.
On Sep 2, 2012 3:58 PM, "Jonathan Scott" <lists at xistenz.org> wrote:

> Looks like it was something in the database. As you suggested, I ran a
> vacuum analyze and followed up with a re-index for good measure; Spacewalk
> app has been up and running without any further issue for a solid week now.
>
> Thank you very much for your assistance, Paul!
>
> - Jonathan
>
> On Fri, Aug 17, 2012 at 10:53 AM, Jonathan Scott <lists at xistenz.org>wrote:
>
>> I am not running monitoring.
>>
>> This has started happening rather recently, and it all began with
>> connection issues a few weeks ago. I have been steadily increasing the
>> connection count, which has replaced the nightly flood of traceback emails
>> with complete lack of response from the databases.
>>
>> I have started an analyze and vaccum run, let's see if this helps!
>>
>> Thanks!
>> - Jonathan
>>
>>
>> On Fri, Aug 17, 2012 at 9:53 AM, Paul Robert Marino <prmarino1 at gmail.com>wrote:
>>
>>> Are you running monitoring in spacewalk. If so that may be the culprit
>>> its a known issue in spacewalk 1.7 with postgresql.
>>> You can use pgpool (connectoin pooling proxy) to consolidate and reuse
>>> the connections, but due to an other know bug in 1.7 you would have to
>>> change the port postgresql listens on and put pgpool on that postgresqls
>>> port instead.
>>> You also may want to try doing an analyze if there have been a lot of
>>> changes in the database so its planner statistics become more acurate.
>>> And as always with postgresql its best to plan to do a full vaccum
>>> analyze on a regular basis. While the auto vacuum introduced in the 8
>>> series reduces the need for it, it doesn't completely eliminate it.
>>> On Aug 17, 2012 9:12 AM, "Jonathan Scott" <lists at xistenz.org> wrote:
>>>
>>>> Good Morning All,
>>>>
>>>> I've started having some rather odd issues with my Spacewalk install
>>>> using postgres. Every night, we do an errata load which typically consists
>>>> of a handful of RPMs if anything at all, but recently (and for no apparent
>>>> reason) the database has just started slowing down. This is causing
>>>> everything on the system to back up;  when I checked this morning, there
>>>> were 542 postgres processes in various states (idle, idle in transmission,
>>>> UPDATE waiting [which concerns me the most]), and 258 httpd processes.
>>>> Restarting the database seems to clear the issue up, but I cannot find
>>>> anything specific in the logs to tell me what is going on here. There are
>>>> no traceback emails or other forms of alerting coming from the application.
>>>>
>>>> Any insight?
>>>>
>>>> Thanks!
>>>> Jonathan
>>>>
>>>> _______________________________________________
>>>> Spacewalk-list mailing list
>>>> Spacewalk-list at redhat.com
>>>> https://www.redhat.com/mailman/listinfo/spacewalk-list
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/spacewalk-list/attachments/20120902/98035b39/attachment.htm>