Definitely something going on with the sipxchange user on this box (as 
mentioned below), but I had no problem running other development builds (for my 
companies product) after I noticed the issue with sipxchange (the inability to 
fork).

One concrete question: what can cause a "503 Maximum Calls In Progress" on a 
test system with no user traffic and 6 Polycom phones? None of the sip logs 
were growing when this was happening, there wasn't a lot of processes running 
(parented by the manager or otherwise). Maybe there's an fd leak?

-Eric

On Dec 18, 2009, at 5:41 PM, Scott Lawrence wrote:

> On Fri, 2009-12-18 at 16:34 -0600, Eric Varsanyi wrote:
>> Sorry for incomplete information.
>> 
>> Yes, the call comes in to the SPA3102 'fxo1' at 192.172.252.21 and it
>> is programmed to dial 0...@... to get the auto attendant. I enter '200#'
>> to transfer to extension 200 "Kitchen", this is 192.172.252.42. The
>> sipX code is running on 192.172.252.163.
>> 
>> Here's another extract I did with "d3 ~/sip1/INSTALL/bin./sipx-trace
>> -a -o /tmp/poly2.xml 5434d8c5-c89e2db7".
>> 
>> One other data point, for some time over the last day something (that
>> I couldn't track) ran wild and all my shells logged in as sipxchange
>> started reporting 'out of resources' when they tried to fork any
>> command. I could log into the box (which was set up by the EDE stuff)
>> on another account w/o any issue and run things, there wasn't a huge
>> number of processes active per ps auxww (nor a horde of zombies). This
>> is quite a mystery and after an 'sstop' / 'sstart' the problem went
>> away (for now). I don't now if this is related but its worth
>> mentioning. The box has 2T of disk on a SAS array in the root FS
>> (which is as expected mostly empty) and 32G of physical RAM (of which
>> very little is in use according to top and /proc/meminfo).
>> 
>> I notice that freeswitch is dropping core files now and then in the
>> log directory, but they don't seem related to the 'REJECT' button
>> issue (I haven't dug into that issue yet).
> 
> You've got some wider system issue - the call trace looks perfectly
> normal as far as the phone is concerned.  The reject returns a normal
> 486 Busy response, and then the call is forked to voicemail, but the
> voicemail server returns:
> 
> SIP/2.0 503 Maximum Calls In Progress
> 
> Since Busy is a 'better' status than the 503, the proxy ends up
> returning the busy to the caller (SIP only allows a proxy to return one
> error response to a request).
> 
> 

_______________________________________________
sipx-dev mailing list [email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
sipXecs IP PBX -- http://www.sipfoundry.org/

Reply via email to