Asterisk crash

Asterisk Asterisk Support
1

Dear all,

i have installed Asterisk 11.2.1.
2/3 times every day, Asterisk crashes. In the log message see:

Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] WARNING[22477][C-00000023] channel.c: Unable to write to alert pipe on (null) (qlen = 0): Bad file descriptor!
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:11] ERROR[22477][C-00000023] astobj2.c: bad magic number for 0x217c1f8. Object is likely destroyed.
[Mar 25 09:07:12] WARNING[22477][C-00000024] chan_sip.c: Got an attempt to replace own Call-ID on 625e930563b64b05b05dd0fb0df845d6@10.10.1.5
[Mar 25 09:07:12] NOTICE[22477][C-00000024] chan_sip.c: Got OK on REFER Notify message
[Mar 25 09:07:18] WARNING[22477] chan_sip.c: Autodestruct on dialog ‘78ff892d4dcab8194c51d8ae7e85605e@10.10.1.10:5060’ with owner (null) in place (Method: BYE). Rescheduling destruction for 10000 ms

I reinstalled Asterisk in another server but the problem is the same. What can be problem or the error in the configuration?

Thanks and regards,
Davide

2

You have a memory corruption. That needs reporting on issues.aterisk.org/jira/, although you will need a backtrace from an unoptimised build. You should probably also build with thread debugging. Note that this can degrade throughput.

Ideally, you should run it under valgrind, but it is unlikely that it will perform well enough for production use.

3

Dear David,

what do you mean exactly with memory corruption?
I reinstalled Asterisk in another server and the problem is the same, i tried to open issues.aterisk.org/jira/ but the website is not available.

Thank you for your support.

Davide

4

A part of the RAM that is still in use has been overwritten by some other part of the code. This is a software bug (or less likely a hardware failure). It is one of the most difficult ones to diagnose, because the original fault can happen some time before the symptom presents.