| Summary: | ASTERISK-17617: Crash when posting a CDR record | ||
| Reporter: | Kirill Katsnelson (kkm) | Labels: | |
| Date Opened: | 2011-03-29 16:00:25 | Date Closed: | 2011-04-05 16:07:29 |
| Priority: | Critical | Regression? | No |
| Status: | Closed/Complete | Components: | Resources/res_odbc |
| Versions: | Frequency of Occurrence | ||
| Related Issues: | |||
| Environment: | Attachments: | ( 0) 19037-backtrace.txt | |
| Description: | Both CEL and CDR are enabled for a transition. Using ODBC driver for Microsoft SQL. This is a random crash that happened only once. ****** ADDITIONAL INFORMATION ****** Looking at the backtrace, the crash seem to have happened well inside unixodbc library. Do not know if asterisk could be held at any fault at all. | ||
| Comments: | By: Tilghman Lesher (tilghman) 2011-03-31 02:35:37 The crash is actually within the pthreads library, and the crash suggests a memory corruption problem. You're right that this isn't necessarily an Asterisk problem, though, especially as the mutex in question isn't in our code; however, it's still possible that the application code is where the memory corruption occurred. The only way for us to be able to track this down would be a reproducible case with valgrind. By: Kirill Katsnelson (kkm) 2011-03-31 16:42:55 I do not have a repro for this. Logged a one-time event, just in case it might click together with other known problems. Might be a one-time accident, too. By: Leif Madsen (lmadsen) 2011-04-05 16:07:29 Closing this issue as tilghman suggests the data required to look further into this. If it happens again or you can reproduce and provide the requested data please feel free to reopen this issue. | ||