[Freeswitch-users] Help with debugging stuck threads/channels

[Patrick Lists]

On 27-02-12 02:43, [Redhosting] Egbert Groot wrote:
> Hi All,
>
> I'm running a freeswitch server with moderate load (at least, in my
> opinion..). Everything runs fine and stable.
> There is however one issue I have: occasionally a channel/thread gots
> stuck (and keeps that way until a full restart). Doing a uuid_kill
> doesn't help, and a uuid_dump got locked itself as well. I managed to
> get a core dump from the running process, and I can see a couple of
> threads in a locked state: "#0 0x00007f3878764be4 in __lll_lock_wait ()
> from /lib/libpthread.so.0" etc.
> This is where my experience ends. How should I proceed in finding the
> cause of these locks? I've read these kind of locks can happen when
> doing 'fancy' things with cdr processing etc. In my setup this doesn't
> seem te be the case, as this is a 'simple' gateway only using the
> default modules.
>
> (for the stack trace see http://pastebin.freeswitch.org/18526 version:
> commit 224f4c0f8aefcb969ecab7ff9d3a66dce92bd943 / Tue Jan 24 11:38:41
> 2012 -0600).

Bugs should always be submitted to Jira: http://jira.freeswitch.org
The developers will probably want to see that backtrace so please 
include it in the bugreport.

Please also checkout the following link so you provide the developers 
with the information they need: 
http://wiki.freeswitch.org/wiki/Reporting_Bugs#Reporting_A_Bug_With_JIRA

Since your version is more than a month old you might get the request to 
first update to today's git master and verify if the bug still happens. 
So I guess you could try that first.

Regards,
Patrick