Asterisk crashing on vicibox 9

[dspaan]

Asterisk is simply crashing without warning after about 20 mins with just 4 agents logged in on a RATIO campaign.

Version: 2.14b0.5
SVN Version: 3199
DB Schema Version: 1588
DB Schema Update Date: 2020-03-23 16:05:02

In the CLI i see this:

ERROR[3402]: chan_sip.c:4271 __sip_reliable_xmit: Serious Network Trouble; __sip_xmit returns error for pkt data


But i also see that on other servers which do not have a crashing asterisk, so not sure if it's related.

Anyone have an idea how to approach this?

When asterisk crashes the agents get the famous time synchronization error. But i think that's just a symptom.
I can manually start asterisk again with service asterisk start and then it works for a while but after 20 min it simply disconnects.

[williamconley]

Serious Network Trouble

Have you any reason to believe that this is NOT merely a networking problem?

are there any system logs that indicate a problem right around the time this problem starts?

[mubeen]

I'm having same issue, did you configured WEBRTC? In my case I'm having same issue with INB_MANUAL campaign as well

[williamconley]

if using webrtc, try using a soft phone instead to see if the problem is limited to webrtc.

[dspaan]

I contacted vicidial support about this and here is the diagnosis:

There is a known bug in MeetMeAdmin in Asterisk 13 that can cause crashes when users use the volume and mute controls on the agent screen and not on their phones. Check the screenlog for calls to MeetMeAdmin with the 'T', 't', 'M', or 'm' option specified. If you are seeing a bunch of them, this is the most likely cause for the crash. Please note that there will also be calls with the 'K' option. Those are harmless.

If you are seeing those the simplest solution is to comment out the calls to MeetMeAdmin in extensions.conf. So change this:

; MeetMe Adjustment Extensions
exten => _X48600XXX,1,MeetMeAdmin(${EXTEN:2},T,${EXTEN:0:1})
exten => _X48600XXX,2,Hangup()

exten => _X38600XXX,1,MeetMeAdmin(${EXTEN:2},t,${EXTEN:0:1})
exten => _X38600XXX,2,Hangup()

exten => _X28600XXX,1,MeetMeAdmin(${EXTEN:2},m,${EXTEN:0:1})
exten => _X28600XXX,2,Hangup()

exten => _X18600XXX,1,MeetMeAdmin(${EXTEN:2},M,${EXTEN:0:1})
exten => _X18600XXX,2,Hangup()

to this:

; MeetMe Adjustment Extensions
;exten => _X48600XXX,1,MeetMeAdmin(${EXTEN:2},T,${EXTEN:0:1})
exten => _X48600XXX,1,Hangup()

;exten => _X38600XXX,1,MeetMeAdmin(${EXTEN:2},t,${EXTEN:0:1})
exten => _X38600XXX,1,Hangup()

;exten => _X28600XXX,1,MeetMeAdmin(${EXTEN:2},m,${EXTEN:0:1})
exten => _X28600XXX,1,Hangup()

;exten => _X18600XXX,1,MeetMeAdmin(${EXTEN:2},M,${EXTEN:0:1})
exten => _X18600XXX,1,Hangup()

Be sure to leave the ones with the 'K' option though. You will also want to notify the users that these buttons have been disabled and not to use them.

I checked our messages log and and indeed this was the case.


About these messages

ERROR[3402]: chan_sip.c:4271 __sip_reliable_xmit: Serious Network Trouble; __sip_xmit returns error for pkt data

VERBOSE[31316][C-00000002] res_srtp.c: SRTCP unprotect failed because of authentication failure

they said:

As for those two messages, I have not fully figured out what cause those. They are related to using a webrtc phone, but they do not cause issues so they are safe to ignore.

[williamconley]

Excellent Postback!

[carpenox]

They're is a fix for this, a patch you can apply:

http://download.vicidial.com/asterisk-p ... trol.patch
http://download.vicidial.com/asterisk-p ... .7.1.patch
http://download.vicidial.com/asterisk-p ... 2.16.patch
http://download.vicidial.com/asterisk-p ... 2.23.patch