[Freeswitch-users] Sofia freezing for 5 minutes then starting again

[Steven Ayre]

>
> Freeswitch is 1.0.head (git-7531fed 2011-08-17 11-27-20 -0500)


As you've already acknowledged it's a very old version.

It's possible that your issue has already been found and fixed, but if it
hasn't then the code will have changed significantly since then and you'd
really need to reproduce it on the latest code for it to be investigated.


As some general thoughts though, are you able to spot it happening while
it's happening or only afterwards?

If you're able to get on the system during one of those times look at what
else is happening. Is the load average/cpu usage/io high? Perhaps
something's running that's blocking all access or causing very high IO.

What DB backend are you using for Sofia? Is it possible that that's hanging
for a moment? For example if you're running a backup on the DB that blocks
all writes to the DB while Sofia is trying to update the DB that perhaps
would cause this.

Try running a SIP OPTIONS ping your your sofia profile from the localhost
during that time, which should exclude it being any issue on the ethernet.

-Steve




On 22 November 2012 19:31, Tim St. Pierre <fs-list at communicatefreely.net>wrote:

> Hello,
>
> I'm having a bit of an odd problem.
>
> Intermittently, often every 2-3 days or so, Freeswitch stops replying to
> SIP for about 5
> minutes.  I can't verify if it's EXACTLY 5 minutes, but it seems to be
> pretty close.
>
> During this time, no new registrations or invites can happen, but existing
> calls stay
> connected for at least a minute or two.  In the logs, you can see calls
> slowly hanging up
> with "NORMAL_CLEARING".  In 5 minutes, everything starts up again with no
> word about it at
> all in the logs.
>
> When calls resume, I notice that the number of sessions returned by the
> status command is
> one higher than the actual number sessions returned by show channels, or
> by looking in the
> database.  Every time this happens, the discrepancy increases by one.
>
> The interruption happens on all SIP profiles, but calls originated from
> the socket API
> still work, insofar as they return with PROGRESS_TIMEOUT since the
> profiles are still
> running, but stuck.
>
> We are using ODBC/MySQL for the core database, and the database server
> only runs this
> database and some basic PHP/xml-curl stuff.
>
> We have 416 endpoints registered, and usually sit at about 30 sessions
> during the day.
>
> This never happens at night, only during busier times, but not necessarily
> busy hour.
>
> I'm running on FreeBSD 8.2-RELEASE AMD 64(2 XEON cores, 4G ram)
>
> Freeswitch is 1.0.head (git-7531fed 2011-08-17 11-27-20 -0500)
>
> Yes, I know it's old and I'm trying to upgrade, but I'm still having some
> problems getting
> all my phones to work properly with 1.2 stable.  This is a production
> system, so I can't
> just blindly put out the newest release.  Mostly, I need to buy myself
> some time so that I
> can get the kinks worked out of the latest version and then upgrade the
> production box.
>
> I'm grateful for any insights as to what could be happening, even if a
> solution is just a
> temporary workaround.
>
> Thanks!
>
> -Tim
>
> _________________________________________________________________________
> Professional FreeSWITCH Consulting Services:
> consulting at freeswitch.org
> http://www.freeswitchsolutions.com
>
> 
> 
>
> Official FreeSWITCH Sites
> http://www.freeswitch.org
> http://wiki.freeswitch.org
> http://www.cluecon.com
>
> FreeSWITCH-users mailing list
> FreeSWITCH-users at lists.freeswitch.org
> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
> http://www.freeswitch.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freeswitch.org/pipermail/freeswitch-users/attachments/20121123/23b2e582/attachment-0001.html