[asterisk-bugs] [JIRA] (ASTERISK-24633) Asterisk Replies : Call leg/transaction does not exist

[Rusty Newton (JIRA)]

    [ https://issues.asterisk.org/jira/browse/ASTERISK-24633?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=224214#comment-224214 ] 

Rusty Newton commented on ASTERISK-24633:
-----------------------------------------

I don't see an example of the "transaction does not exist" incident in the log.

Also, it looks like Asterisk is crashing?

{quote}
Dec 23 03:29:24 callback01my kernel: asterisk[2652]: segfault at 9d0 ip 000000000048c120 sp 00007f9463ffc858 error 4 in asterisk[400000+208000]
{quote}

Are you running Asterisk with -g and does it dump a core? If so, can you provide a backtrace? Please follow these instructions closely: https://wiki.asterisk.org/wiki/display/AST/Getting+a+Backtrace

{quote}
If I set debug mode and recompile, using gdb and debug, shall asterisk be able to take heavy load to collect the logging information to resolve this issue?
{quote}
There may be a little more load, but you should be fine. It is hard to know as it really depends on whether we are talking about tens of calls or thousands of calls.

If Asterisk is crashing.. that would explain why you are losing all your calls. 

{quote}
Note: At above point asterisk uptime 'core show uptime' is reset to zero, but strangely asterisk recover or continue to operate.
{quote}
You probably have a script restarting Asterisk, like safe_asterisk.

We can't sign an NDA for your logs. We can lock the issue down to only bug marshals but that doesn't guarantee you anything legally.

{quote}
I have to ensure our client's phone numbers and destination calls are safe and kept private.
{quote}
You would need to manually scrub all private information from your files before uploading. This is what most users do.

> Asterisk Replies : Call leg/transaction does not exist 
> -------------------------------------------------------
>
>                 Key: ASTERISK-24633
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-24633
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Channels/chan_sip/General
>    Affects Versions: 11.15.0, 13.1.0
>         Environment: CentOS 6.5 64bit
>            Reporter: Sid Mason
>            Assignee: Sid Mason
>            Severity: Critical
>         Attachments: DEBUG_LOG Asterisk 11.15.txt, DEBUG_LOG.txt
>
>
> It seems asterisk erases the information about the ongoing call through Chan_SIP 
> This happens when :
> A) This is right before the time Asterisk looses all SIP Dialog status related second leg on Bridges
> Dec 23 03:29:24 callback01my kernel: asterisk[2652]: segfault at 9d0 ip 000000000048c120 sp 00007f9463ffc858 error 4 in asterisk[400000+208000]
> Note:  At above point asterisk uptime 'core show uptime' is reset to zero, but strangely asterisk recover or continue to operate.
> B) 20 or AMI listener is attached to asterisk on port 5038 and they all have filter as below:
> C) AMI has filter of the UNIQUE ID of the channels and others:
> '!Event: AGIExec'
> '!Event: VarSet'
> '!Event: RTCPSent'
> '!Event: RTCPReceived'
> '!Event: VarSet'
> D) the call durations are usually above 10 minutes
> E) the bridge is doing unlink before the hang-up and somehow asterisk must hold the information for SIP Dialog right before the hang-up
> I tried asterisk 11.15 and the same result was there.
> At this point ANSWEREDTIME is reported as NULL as well
> F) At incident asterisk is loosing track of multiple calls at once.



--
This message was sent by Atlassian JIRA
(v6.2#6252)