[OpenSIPS-Users] Opensips presence server. How to check for Out Of Memory errors?

[Suryaveer Chauhan]

Hi Bogdan,

I didn't do the SIGUSER1 test, instead I checked at time of shutdown. At
the end huge log is generated and I don't know if all of that is unfreed
memory.
You may wish to look at the log http://pastebin.com/M9tmXk5F. This is not
complete log as complete one was very huge and I couldn't post it therefore
removed some from middle.

Below are few lines from the log:

   1. Oct  9 13:40:21 [25420]:   423. N  address=0x7febb9c15a18
   frag=0x7febb9c159e8 size=128 used=1
   2. Oct  9 13:40:21 [25420]:     428. N  address=0x7febb9c20870
   frag=0x7febb9c20840 size=25216 used=1
   3. Oct  9 13:40:21 [25428]:             alloc'd from pvar.c:
   pv_parse_format(3782)
   4. Oct  9 13:40:21 [25420]:             alloc'd from io_wait.c:
   init_io_wait(609)
   5. Oct  9 13:40:21 [25428]:         start check=f0f0f0f0, end check=
   c0c0c0c0, abcdefed
   6. Oct  9 13:40:21 [25420]:         start check=f0f0f0f0, end check=
   c0c0c0c0, abcdefed
   7. Oct  9 13:40:21 [25428]:     424. N  address=0x7febb9c15af8
   frag=0x7febb9c15ac8 size=128 used=1
   8. Oct  9 13:40:21 [25420]: dumping free list stats :
   9. Oct  9 13:40:21 [25428]:             alloc'd from pvar.c:
   pv_parse_format(3782)
   10. Oct  9 13:40:21 [25420]: hash= 261. fragments no.:     1, unused:
     0#012#011#011 bucket size:      2088 -      2088 (first      2088)
   11. Oct  9 13:40:21 [25428]:         start check=f0f0f0f0, end check=
   c0c0c0c0, abcdefed
   12. Oct  9 13:40:21 [25428]:     426. N  address=0x7febb9c16460
   frag=0x7febb9c16430 size=8192 used=1
   13. Oct  9 13:40:21 [25420]: hash= 2056. fragments no.:     1, unused:
     0#012#011#011 bucket size:   2097152 -   4194304 (first   3814544)
   14. Oct  9 13:40:21 [25428]:             alloc'd from mi_writer.c:
   mi_writer_init(51)
   15. Oct  9 13:40:21 [25428]:         start check=f0f0f0f0, end check=
   c0c0c0c0, abcdefed
   16. Oct  9 13:40:21 [25428]:     427. N  address=0x7febb9c184c0
   frag=0x7febb9c18490 size=33616 used=1
   17. Oct  9 13:40:21 [25428]:             alloc'd from io_wait.c:
   init_io_wait(560)
   18. Oct  9 13:40:21 [25428]:         start check=f0f0f0f0, end check=
   c0c0c0c0, abcdefed


Does all those line numbers represent the unfreed memory?

Thanks,
Surya

On 11 January 2016 at 04:36, Bogdan-Andrei Iancu <bogdan at opensips.org>
wrote:

> Hi Surya,
>
> Even for checking if all the memory was properly release, you still have
> to use the Memory debugger - have you managed to get the memory dump on
> SIGUSR1 (as per example in the page I pointed)? if so, such a dump simply
> contains a list of all allocated memory chunks - size and where from code
> they were allocated from.
>
> Best regards,
>
> Bogdan-Andrei Iancu
> OpenSIPS Founder and Developer
> http://www.opensips-solutions.com
>
> On 09.01.2016 10:24, surya wrote:
>
>> Hi Bogdan,
>>
>> I checked the method mentioned on the link but really, I am unable to
>> understand anything generated from the log.
>>
>> Now, my problem is not out of memory. I just want to check if there are
>> some
>> memory which is not *freed*.
>>
>> Is there a way I can run opensips with Valgrind, because I am bit
>> comfortable with it and help is also easily available.
>>
>> I am not expert in Unix or OS therefore finding it difficult to understand
>> the logs.
>> Hope you understand.
>>
>> Thanks,
>> Surya
>>
>>
>>
>> --
>> View this message in context:
>> http://opensips-open-sip-server.1449251.n2.nabble.com/Opensips-presence-server-How-to-check-for-Out-Of-Memory-errors-tp7599129p7600541.html
>> Sent from the OpenSIPS - Users mailing list archive at Nabble.com.
>>
>> _______________________________________________
>> Users mailing list
>> Users at lists.opensips.org
>> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20160111/3b610e57/attachment-0001.htm>