[OpenSIPS-Users] Opensips 1.11.3 crash

Răzvan Crainea razvan at opensips.org
Wed May 20 09:29:48 CEST 2015


Hi, Federico!

Is there any chance you could ping pe on IRC, freenode, #opensips so we 
can debug this further?

Best regards,

Răzvan Crainea
OpenSIPS Solutions
www.opensips-solutions.com

On 05/15/2015 04:12 PM, Federico Edorna wrote:
> Sure, you have 3 different cores pasted on the following urls:
>
> http://pastebin.com/xZ2zqJ0F
> http://pastebin.com/8DWhsMfK
> http://pastebin.com/9ERCD3mZ
>
>
> The opensips logs for those cores are pasted below in my email (Apr 6, 
> 2015 at 4:46 PM)
>
> Thanks!
>
>
> On Fri, May 15, 2015 at 7:14 AM, Răzvan Crainea <razvan at opensips.org 
> <mailto:razvan at opensips.org>> wrote:
>
>     Hi, Federico!
>
>     Can you attach the output of the core file on pastebin?
>
>     Best regards,
>
>     Răzvan Crainea
>     OpenSIPS Solutions
>     www.opensips-solutions.com  <http://www.opensips-solutions.com>
>
>     On 05/08/2015 04:51 PM, Federico Edorna wrote:
>>     Hi Răzvan, It was happening at least once a day. It started to
>>     happen when we reach ~80 registered terminals, not a big load.
>>     Now I'm using the event_rabbitmq module with 400 terminals and
>>     it's working fine.
>>
>>     The event is a custom one, I called "E_REGISTERED", it's just to
>>     notify an external process that a particular terminal has
>>     registered, this is part of the configuration that crashes:
>>
>>     /startup_route {/
>>     /subscribe_event("E_REGISTERED", "xmlrpc:10.10.11.2:10080:OSEvent");/
>>     /}/
>>     /.../
>>     /.../
>>     /route {/
>>     /.../
>>     /.../
>>     /        if (is_method("REGISTER")) {
>>     /
>>     /.../
>>     /.../
>>     /$avp(attr-name) = "username";
>>     /
>>     /                $avp(attr-val) = $tU;/
>>     /$avp(attr-name) = "domain";/
>>     /                $avp(attr-val) = $td;/
>>     /raise_event("E_REGISTERED", $avp(attr-name), $avp(attr-val));/
>>     /.../
>>     /.../
>>     /}/
>>
>>     Another thing: when I compiled with DBG_QM_MALLOC instead of
>>     F_MALLOC to debug, I didn't have any crashes for about 5 days.
>>     Maybe I should have waited more time to confirm, but it seems
>>     that the first memory manager solved the issue.
>>
>>     Regarding to the core files, it seems than some module
>>     (even_xmlrpc for me..) it's freeing memory that it should not.
>>     After this issue I realized that the module was in beta, so I
>>     moved to the rabbitmq
>>
>>     Thanks for your reply
>>     Federico
>>
>>
>>     On Fri, May 8, 2015 at 7:47 AM, Răzvan Crainea
>>     <razvan at opensips.org <mailto:razvan at opensips.org>> wrote:
>>
>>         Hi, Federico!
>>
>>         Is this easily replicating, or it happens once in a while?
>>         Also, what events are you raising?
>>
>>         Best regards,
>>
>>         Răzvan Crainea
>>         OpenSIPS Solutions
>>         www.opensips-solutions.com  <http://www.opensips-solutions.com>
>>
>>         On 04/24/2015 05:44 PM, Federico Edorna wrote:
>>>         Just in case somebody deal with the same issue, the problem
>>>         seems to be event_xmlrpc module. I tried with the
>>>         event_datagram to notify the external process and I got no
>>>         more crashes for a couple of weeks.
>>>         Now I'm using event_rabbit module instead of datagram
>>>         without problems for a couple of days.
>>>
>>>
>>>         On Mon, Apr 6, 2015 at 4:46 PM, Federico Edorna
>>>         <fedorna at anura.com.ar <mailto:fedorna at anura.com.ar>> wrote:
>>>
>>>             Hello, I'm getting core dumps in version 1.11.3.
>>>             Unlike other opensips we are running without problems,
>>>             we're using some extra modules in this config because
>>>             opensips needs to notify an external process (via
>>>             event_xmlrpc) when a terminal registers, and that
>>>             external process afterwards sends opensips (via
>>>             mi_datagram/t_uac_dlg) a MWI NOTIFY for the terminal.
>>>
>>>             I'm pasting 3 backtraces (commit cbaf569, but it
>>>             happened for previous commits too)
>>>
>>>             http://pastebin.com/xZ2zqJ0F
>>>             http://pastebin.com/8DWhsMfK
>>>             http://pastebin.com/9ERCD3mZ
>>>
>>>             This is what syslog shows:
>>>
>>>             2015-04-03T13:45:16.228227-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24272]:
>>>             CRITICAL:core:recv_all: 1st recv on 36 failed:
>>>             Connection reset by peer
>>>             2015-04-03T13:45:16.228249-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24272]:
>>>             CRITICAL:core:handle_tcp_child: read from tcp child 0
>>>             (pid 24240, no 0) Connection reset by
>>>              peer [104]
>>>             2015-04-03T13:45:16.228260-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24272]:
>>>             CRITICAL:core:receive_fd: EOF on 38
>>>             2015-04-03T13:45:16.250712-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24214]:
>>>             INFO:core:handle_sigs: child process 24240 exited by a
>>>             signal 11
>>>             2015-04-03T13:45:16.250727-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24214]:
>>>             INFO:core:handle_sigs: core was generated
>>>             2015-04-03T13:45:16.250735-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24214]:
>>>             INFO:core:handle_sigs: terminating due to SIGCHLD
>>>             2015-04-03T13:45:16.250800-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[24270]:
>>>             INFO:core:sig_usr: signal 15 received
>>>
>>>             ----
>>>
>>>             2015-04-03T13:54:48.179260-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21747]:
>>>             CRITICAL:core:recv_all: 1st recv on 36 failed:
>>>             Connection reset by peer
>>>             2015-04-03T13:54:48.179289-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21747]:
>>>             CRITICAL:core:handle_tcp_child: read from tcp child 0
>>>             (pid 21715, no 0) Connection reset by
>>>              peer [104]
>>>             2015-04-03T13:54:48.179307-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21747]:
>>>             CRITICAL:core:receive_fd: EOF on 38
>>>             2015-04-03T13:54:48.179373-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21688]:
>>>             INFO:core:handle_sigs: child process 21715 exited by a
>>>             signal 11
>>>             2015-04-03T13:54:48.179388-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21688]:
>>>             INFO:core:handle_sigs: core was generated
>>>             2015-04-03T13:54:48.179402-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21688]:
>>>             INFO:core:handle_sigs: terminating due to SIGCHLD
>>>             2015-04-03T13:54:48.179417-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21746]:
>>>             INFO:core:sig_usr: signal 15 received
>>>             2015-04-03T13:54:48.179426-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21745]:
>>>             INFO:core:sig_usr: signal 15 received
>>>             2015-04-03T13:54:48.179435-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[21743]:
>>>             INFO:core:sig_usr: signal 15 received
>>>
>>>             ----
>>>
>>>             2015-04-03T14:44:01.064875-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31736]:
>>>             CRITICAL:core:recv_all: 1st recv on 36 failed:
>>>             Connection reset by peer
>>>             2015-04-03T14:44:01.064898-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31736]:
>>>             CRITICAL:core:handle_tcp_child: read from tcp child 0
>>>             (pid 31704, no 0) Connection reset by
>>>              peer [104]
>>>             2015-04-03T14:44:01.064922-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31736]:
>>>             CRITICAL:core:receive_fd: EOF on 38
>>>             2015-04-03T14:44:01.064943-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31678]:
>>>             INFO:core:handle_sigs: child process 31704 exited by a
>>>             signal 11
>>>             2015-04-03T14:44:01.064954-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31678]:
>>>             INFO:core:handle_sigs: core was generated
>>>             2015-04-03T14:44:01.064963-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31678]:
>>>             INFO:core:handle_sigs: terminating due to SIGCHLD
>>>             2015-04-03T14:44:01.066539-03:00 bermeja
>>>             /home/gc/local/opensips/sbin/opensips[31736]:
>>>             INFO:core:sig_usr: signal 15 received
>>>
>>>             Thanks in advance
>>>             Federico
>>>
>>>
>>>
>>>
>>>         _______________________________________________
>>>         Users mailing list
>>>         Users at lists.opensips.org  <mailto:Users at lists.opensips.org>
>>>         http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>>
>>
>>         _______________________________________________
>>         Users mailing list
>>         Users at lists.opensips.org <mailto:Users at lists.opensips.org>
>>         http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>>
>>
>>
>>
>>     _______________________________________________
>>     Users mailing list
>>     Users at lists.opensips.org  <mailto:Users at lists.opensips.org>
>>     http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
>
>     _______________________________________________
>     Users mailing list
>     Users at lists.opensips.org <mailto:Users at lists.opensips.org>
>     http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
>
>
>
> _______________________________________________
> 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/20150520/588e780d/attachment-0001.htm>


More information about the Users mailing list