hello,
i am currently working on tracking down a bug in asterisk which is causing my system to crash.
What I am wondering is: what is the properly protocol for configuring asterisk to help track down bugs so they can properly be submitted to the bug list.
I currently only have WARNING and ERROR messages going to my log and they don’t really provide any useful information. The last thing I see in my log is:
[Sep 3 18:56:34] WARNING[24221] channel.c: SIP/bw-077d3210 is already going to masquerade as Local/OFFHOOK@acd_dial_offhook-62e9,1
thanks, looks to me like the only real preparation was to disable compiler optimization and recompile to ensure good info in the core dump. Guess it’s just a waiting game now.
my previous core dumps are totally useless, but i guess that’s to be expected with the compiler optimizing things.
SIP/bw-077d3210 is already going to masquerade as Local/OFFHOOK@acd_dial_offhook-62e9,1
Well that does tell you a fair bit realy, I would do a show dialplan OFFHOOK@acd_dial_offhook and see what the dialplan is as it seems to be a problem area.
The version of * which crashed was 1.4.20.1. I have since upgraded to 1.4.21.2, i can’t say if the problem has gone away as I only upgraded last night. I don’t believe the dialplan is actually the problem as it is a sporadic crash. This context is run successfully hundreds or thousands of times before a crash actually occurs.
I have a fairly complex application which uses AMI redirects, FastAGI, and app_conference. This is just a small piece of the puzzle. Basically we request a manager originate for a Local channel which lands in this context.