[asterisk-dev] Asterisk 1.6.1.15-rc2 Crash

[Leif Madsen]

Mark Michelson wrote:
> Bryant Zimmerman wrote:
>> I made the mistake of rushing the rc2 to a production server to try and 
>> get T.38 to work correctly and it crashed and asterisk stopped running 
>> after about 4 hours in production.  I fell back to 1.6.1.14.  How can I 
>> figure out what caused the crash.
>> The system had taken only about 150 calls when it crashed.
>>
>> It deffinatly does not appare to be stable.  
> 
> If Asterisk was run with the -g option, then it would have dumped a core file 
> upon crashing. This core file can then be inspected with gdb to get a backtrace 
> where the crash occurred. However, unless the DONT_OPTIMIZE option was turned on 
> in menuselect, it is likely that the backtrace will be unhelpful since a lot of 
> the data has been optimized away.
> 
> All information about getting backtraces from a crash can be found in 
> doc/backtrace.txt in the Asterisk source directory.

If you can provide this information per Mark, then that would greatly help us to 
resolve it before we release 1.6.1.15.

Once you've gathered the information, including console traces with debug level 
logging (per logger.conf), core set verbose 10, core set debug 10, along with 
the backtrace output, please open an issue on the issue tracker 
(https://issues.asterisk.org) so we can track this appropriately.

Be sure to attach the files as text to the issue.

Thanks!
Leif.