[Spacewalk-list] Alternative ways without OSAD

[Matt Moldvan]

Depends on the volume of your systems... I tried this as well, but
rhn_check initially makes HTTPS calls and with 6,000+ systems it
overwhelmed the Apache instances and caused deadlocks in the Postgres
databases (even with Apache and pgsql variables increased).

I ended up switching the jabberd database to SQLite and it made a lot of
difference over the default Berkeley DB.

For systems going offline, we have a few lines in our Puppet configs to
restart OSAD on stale connections on a daily basis.  To debug your own
connection issues, I recommend increasing the debug value for OSAD in
/etc/sysconfig/rhn/osad.conf (debug = 9) then restart OSAD, to see if you
can narrow down what might be causing the systems to go offline.  If it's
not at the clients you might be able to find something at the server,
somewhere in /var/log/rhn or so.

        $date = strftime("%F")
        $datetime = strftime("%F @ %T")
        exec { "restart_osad_on_conn_error":
                command => "/sbin/service osad restart",
                onlyif => "/bin/grep -q '^${date}.*Unable to connect to
jabber servers' /var/log/osad && sed -i 's/Unable to connect to jabber
servers/fixed on ${datetime} by spacewalk::c_osad.pp/g' /var/log/osad",
        }
        exec { "restart_osad_on_stale_conn":
                command => "/sbin/service osad restart",
                onlyif => "/bin/netstat -an | /bin/grep -q
':5222.*CLOSE_WAIT'",
        }


On Tue, Jan 17, 2017 at 5:33 AM Zeal Vora <zeal at freecharge.com> wrote:

> Hi
>
> Thanks for the inputs everyone.
>
> Basically we need the rpm's to be installed in real time ( around 2-3
> minutes delay is fine ) from SpaceWalk. Since OSAD was giving bit of
> trouble, we thought rhn_check was ideal solution.
>
> As we cannot configure rhn_check in the configuration to check every few
> minutes, I thought to put a cron for 2 minutes which does rhn_check but it
> didn't seem to work properly. Can this be an ideal way ?
>
>
>
> On Thu, Jan 12, 2017 at 5:17 PM, Lichtinger, Bernhard <
> Bernhard.Lichtinger at lrz.de> wrote:
>
>
> >
> >> Due to some reason, many of the servers have issues related to OSAD
> going offline time and again so sometimes we just go for a way of running
> rhn_check manually.
> >>
> >> So I was wondering, can we have rhn_check in all the servers in CRON ?
> Can it work in this situation ? I had bit trouble with putting it in cron
> so wanted to verify.
> >
>
> You don't need to run rhn_check via cron: rhnsd is already checking in to
> spacewalk server (per default) every 4 hours.
> So if you don't need to trigger actions in "real time" then just disable
> osad on the clients.
> _______________________________________________
> Spacewalk-list mailing list
> Spacewalk-list at redhat.com
> https://www.redhat.com/mailman/listinfo/spacewalk-list
>
>
> _______________________________________________
> 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/20170117/938dd796/attachment.htm>