Hello,
I have a frequent problem
See bellow.
Queue show, before hangup… 1. SIP/TRUNKVOIP-0000372d (wait: 1626:05, prio: 0)
Command hangup request…
[root@np-pbx ~]# asterisk -rx "hangup request SIP/SONAVOIP-0000372d"
Queue show, after hangup… 1. SIP/TRUNKVOIP-0000372d (wait: 1626:21, prio: 0)
Hangup has no effect…
Also,
Anyone had a similar case?
Thanks…
(Sorry my english).
Upgrade from: version unspecied, to: latest version. If the problem is still there, build with thread debugging enabled and run core show locks when the queue entry stalls.
Hi David, thanks for reply.
I’m using the latest version, but it happens with all others.
I enabled thread debug and I could follow the problem occurring, in one of my servers.
See bellow…
[root@vm_dev callcenter]# asterisk -rx "queue show 10003"1. Local/002125211749@from-internal-00000067;1 (wait: 26:35, prio: 0)
One agent is iddle on queue… but not delivery the call…
Queue command used…
When I try hangup command on Cli it locks…
The result of “core show locks”…
[root@vm_dev callcenter]# asterisk -rx “core show locks ”
=======================================================================
===
=== Thread ID: 0x7f1f9ba2c700 (do_devstate_changes started at [ 750] devicestate.c ast_device_state_engine_init())#0 (astobj2.c): MUTEX 1087 internal_ao2_callback c 0x2705690 (1)#1 (channel.c): MUTEX 1475 ast_channel_by_name_cb chan 0x7f1e901336d0 (1)
=== Thread ID: 0x7f1f3629d700 (iax2_process_thread started at [12609] chan_iax2.c start_network_thread())#0 (chan_iax2.c): MUTEX 3037 __find_callno &iaxsl[x] 0x7f1f3e9674b8 (1)
=== Thread ID: 0x7f1f36221700 (iax2_process_thread started at [12609] chan_iax2.c start_network_thread())#0 (chan_iax2.c): MUTEX 1297 iax2_lock_owner iaxs[callno]->owner 0x7f1e901336d0 (0)
=== Thread ID: 0x7f1f34cd1700 (handle_tcptls_connection started at [ 734] tcptls.c ast_tcptls_server_root())#0 (astobj2.c): MUTEX 1293 internal_ao2_iterator_next iter->c 0x2705690 (1)
=== Thread ID: 0x7f1f34fb9700 (fast_originate started at [ 4545] manager.c action_originate())#0 (astobj2.c): MUTEX 1085 internal_ao2_callback c 0x2705690 (1)
=== Thread ID: 0x7f1f35491700 (pbx_thread started at [ 6868] pbx.c ast_pbx_start())#0 (channel.c): MUTEX 6903 ast_do_masquerade channels 0x2705690 (1)
=== Thread ID: 0x7f1f99193700 (pbx_thread started at [ 6868] pbx.c ast_pbx_start())#0 (channel.c): MUTEX 3889 __ast_read chan 0x7f1e901336d0 (1)
=== Thread ID: 0x7f1f34c55700 (pbx_thread started at [ 6868] pbx.c ast_pbx_start())#0 (channel.c): MUTEX 6903 ast_do_masquerade channels 0x2705690 (1)
=== Thread ID: 0x7f1f35589700 (pbx_thread started at [ 6868] pbx.c ast_pbx_start())#0 (channel.c): MUTEX 6903 ast_do_masquerade channels 0x2705690 (1)
=== Thread ID: 0x7f1f35779700 (pbx_thread started at [ 6868] pbx.c ast_pbx_start())#0 (channel.c): MUTEX 6903 ast_do_masquerade channels 0x2705690 (1)
=== Thread ID: 0x7f1f3550d700 (fast_originate started at [ 4545] manager.c action_originate())#0 (astobj2.c): MUTEX 899 internal_ao2_link c 0x2705690 (1)
=== Thread ID: 0x7f1f34b5d700 (netconsole started at [ 1486] asterisk.c listener())#0 (cli.c): RDLOCK 2589 __ast_cli_generator &(&helpers)->lock 0x813d70 (2)#1 (astobj2.c): MUTEX 1087 internal_ao2_callback c 0x2705690 (1)
=== Thread ID: 0x7f1f349e9700 (netconsole started at [ 1486] asterisk.c listener())#0 (cli.c): RDLOCK 2589 __ast_cli_generator &(&helpers)->lock 0x813d70 (2)#1 (astobj2.c): MUTEX 1293 internal_ao2_iterator_next iter->c 0x2705690 (1)
=======================================================================
I have more than 30 servers with Asterisk/Callcenter (1.8 and 11)…
Thanks…
david55
September 9, 2014, 3:23pm
4
You have a deadlock between thread 0x7f1f99193700 and thread 0x7f1f9ba2c700. Unfortunately, you seem to have stripped the debug symbols from your Asterisk modules, so you don’t have enough to raise a bug report.
Hm…
david55
September 9, 2014, 4:42pm
6
It looks like you did enable them in the build and then ran strip on the modules directory. I’m fairly sure that the thread debug code can find the symbols, if they are there, but it is a while since I last used it in anger.
If you have got symbols, but it failed to find them, you may be able to translate the file offsets into subroutines using gdb, but you’ll have to work out the details for yourself.
The most you can expect, in the short term is a an understanding of what is going wrong, and therefore possibly how to avoid it. Locking problems are difficult to fix safely.
yateya
January 20, 2015, 7:29pm
7
This might be similar to the bug:
IAX hits maxcallnumber limit of 2048 with only one concurrent channelissues.asterisk.org/jira/browse … ent-224584
In the bug attachments (dead_show_locks.txt), there is similar deadlocks.
yateya
January 20, 2015, 7:30pm
8
To enable debug symbols, I usually use development build. Don’t know if there is better solution or not.
I use ./configre --enable-dev-mode
Then I select required compiler flags from the menu using command make menuselect then select Compiler Flags.
Hi guys,
Thanks for the replies
So, I installed Ast 12 or Ast 13. Using, the same way as before
I think that the changes in the new releases, like Bridge API and the end of chan_sip, contributed to this doesn’t happen more.
Thanks all.
