| Summary: | ASTERISK-07949: Frequent asterisk lockups | ||
| Reporter: | Nikolay (skvt) | Labels: | |
| Date Opened: | 2006-10-18 06:10:20 | Date Closed: | 2006-12-07 10:33:18.000-0600 |
| Priority: | Critical | Regression? | No |
| Status: | Closed/Complete | Components: | Applications/app_queue |
| Versions: | Frequency of Occurrence | ||
| Related Issues: | |||
| Environment: | Attachments: | ( 0) 01:19:34.10.18_01 ( 1) 01:19:35.10.18_01 ( 2) bt.10.18_01 | |
| Description: | Hello, We have asterisk based call center. We are using realtime configuration with MySQL. We have frequent asterisk lockups. Usually couple per week on a production system. I can reproduce the situation in hour or two using call emulation system. After lockup has occurred sometimes we can connect to asterisk and only app_queue is locked, but sometimes all Asterisk PBX locks up. Nikolay | ||
| Comments: | By: Nikolay (skvt) 2006-10-18 08:43:58 I tried to upload core dumps , logs backtraces, but for some reason system does not accept them. How to upload those files ? By: Nikolay (skvt) 2006-10-18 08:54:20 Finali I uploaded all traces. Before uploading, I compressed and split directory with commands: tar cvf a.tar ./!/ bzip2 -zv ./a.tar split -a 4 -b 1000000 ./a.tar.bz2 I uploaded files from xaaaa to xaaai, all previous uploads can be deleted (unfortunately system do not give me ability to delet them). By: Serge Vecher (serge-v) 2006-10-23 13:47:17 please read the bug-guidelines, what's needed is a backtrace of a core, not the core file itself. Please review the backtrace section here http://www.voip-info.org/tiki-index.php?page=Asterisk%20debugging Also, please try the latest Asterisk release, which is 1.2.13. By: BJ Weschke (bweschke) 2006-10-25 13:07:50 we're likely going to need a build with DEBUG_THREADS enabled to get a better idea of what's going wrong where... By: Steve Murphy (murf) 2006-11-16 14:06:45.000-0600 skvt-- how goes the battle? are you still having lock-ups? Have you moved on to newer 1.2 releases? it's been a little while; what's going on? I'll give you a few days to update the status on this bug, and close it if no status. I'll study the code from the bt you did provide, and see if I can spot the problem via review. If not, compiling a late-model 1.2 with NO_OPTIMIZATION, and DEBUG_THREADS , as well as attaching to a locked asterisk with gdb, and getting a bt and a copy of the logs will help a lot to find and kill the bug. By: Steve Murphy (murf) 2006-11-16 14:09:39.000-0600 Oh, and I deleted the split up corefile, it is of no use. Unless we have the exact same source, same compiler, same machine architecture, it's probably going to have problems. We'd just extract the stack trace from it anyway. By: Matt O'Gorman (mogorman) 2006-12-07 10:14:07.000-0600 skvt still with us? By: Steve Murphy (murf) 2006-12-07 10:33:17.000-0600 skvt hasn't responded; I'll close this bug for the time being. If skvt wishes to reopen this (or anyone else for that matter) with some more info, we'll tackle it then. | ||