Summary: | ASTERISK-22889: Segmentation fault when RTP going via ICE | ||||||||||||
Reporter: | Alexandr Byvaltsev (alfox) | Labels: | |||||||||||
Date Opened: | 2013-11-22 04:05:25.000-0600 | Date Closed: | 2017-12-18 11:10:48.000-0600 | ||||||||||
Priority: | Critical | Regression? | |||||||||||
Status: | Closed/Complete | Components: | Channels/chan_sip/General Resources/res_rtp_asterisk | ||||||||||
Versions: | 11.2.0 11.5.0 11.6.0 | Frequency of Occurrence | |||||||||||
Related Issues: |
| ||||||||||||
Environment: | Debian 7.2 on Intel(R) Core(TM) i5-3330 CPU @ 3.00GHz, Debian 7.0 on Intel(R) Xeon(R) CPU E5-2620 0 @ 2.00GHz | Attachments: | ( 0) backtrace_27.11.2013.txt ( 1) backtrace.txt ( 2) backtrace-13_09.txt ( 3) rtp.conf ( 4) sip.conf | ||||||||||
Description: | I have 2 servers with Asterisk 11.5, both have config like this:
{noformat} [trunk_74] type=friend host=192.168.66.99 port=5060 disallow=all allow=alaw context=trunk_74_in t38pt_udptl=no call-limit=0 qualify=no canreinvite=no dtmfmode=inband nat=force_rport,comedia avpf=yes icesupport=yes videosupport=no directmedia=no encryption=yes {noformat} when i make 200 calls at one time from one asterisk to another, in dialplan {noformat} 007777 => { Answer(); Wait(10); Hangup(); }; {noformat} i got Segmentation fault in 1-5 minutes, only when i used icesupport=yes. Same problem i have when sip users connect to Asterisk from browser Chrome over WEBRTC. | ||||||||||||
Comments: | By: Matt Jordan (mjordan) 2013-11-26 12:41:11.347-0600 The fact that this is crashing during a free in the PJSIP memory pool is troubling, as that's indicative of a memory corruption somewhere. Can you attach your full {{sip.conf}} and {{rtp.conf}}? By: Alexandr Byvaltsev (alfox) 2013-11-27 00:11:14.442-0600 yes, sip.conf and rtp.conf in attachments By: Joshua C. Colp (jcolp) 2017-12-18 11:10:49.044-0600 I'm suspending this issue since we now use PJSIP upstream and work has gone in there to fix these kind of problems. We also haven't seen any other reports in recent versions. |