[Asterisk-Dev] Re: [Asterisk-Users] Warning -- chan_iax2.c: ast_sched_runq tasks

[Mark Hulber]

I can get it to repeat.  It happens in a condition that has had no problem 
normally in the past.  I have an incoming call, SIP or ZAP it doesn't really 
matter and then I'm dialing out on IAX to try and connect the call to one or 
more numbers on an IAX channel.  For each IAX channel I create I seem to be 
getting at least one of these messages:

     -- SIP/sipura2_1-085a is ringing
     -- SIP/sipura2_2-0f28 is ringing
     -- Call accepted by 64.34.45.100 (format ulaw)
     -- Format for call is ulaw
     -- IAX2/voipjet-out-5 is making progress passing it to 
SIP/147.135.20.128-08d36c60
Nov  2 18:01:06 WARNING[2781]: chan_iax2.c:7948 network_thread: chan_iax2: 
ast_sched_runq ran 3271 scheduled tasks all at once
     -- Hungup 'Zap/3-1'

In this second case above I was only dialing out to a single number.  In the 
previous case below I was dialing out to two numbers.

Here's what I am running.  I only started seeing it recently:

Asterisk CVS HEAD built by root at asterisk.hulber.com on a i686 running Linux on 
2005-11-02 22:22:26 UTC

MARK.


Rich Adamson wrote:
>> I've started to see this lately on outgoing IAX calls using CVS Head:
>>
>>     -- IAX2/voipjet-out-5 is making progress passing it to 
>> SIP/64.26.157.252-094e4058
>>     -- IAX2/voipjet-out-6 is making progress passing it to 
>> SIP/64.26.157.252-094e4058
>> Nov  2 16:56:15 WARNING[22566]: chan_iax2.c:7948 network_thread: 
>> chan_iax2: ast_sched_runq ran 2502 scheduled tasks all at once
>> Nov  2 16:56:15 WARNING[22566]: chan_iax2.c:7948 network_thread: 
>> chan_iax2: ast_sched_runq ran 3131 scheduled tasks all at once
>> Nov  2 16:56:15 WARNING[22566]: chan_iax2.c:7948 network_thread: 
>> chan_iax2: ast_sched_runq ran 3074 scheduled tasks all at once
>>     -- IAX2/voipjet-out-6 answered SIP/64.26.157.252-094e4058
>>
>> Any interpretation?
> 
> The above would suggest a _very_ busy box. Any idea how many calls
> were in progress during this?
> 
> Some of the folks on the -dev list might be interested in seeing
> the above numbers, so I'll cross-post my response over there as
> well.
> 
> Rich
> 
>