vicidial.org

Team,

I am having an issue with a cluster installation. This are all Physical Servers and handle about 150 agents.

We are running the Lastest Vicidial Version = v9.x86_64-9.0.1.iso
Asterisk version: 13.29.2-vici
SVN: 3182
Cluster Installation Mode = Vicibox-install/expert mode
Database = 1
Vicidials = 4

The issue is that the screens (8 screens on each cluster vicidial/asterisk) are dying and the servers stop dialing.

I looked at several logs and found this error below on the update.2020-01-XX. log

Error:
"2020-01-03 16:31:40|WAITFOR ERRMSG: |read error: Connection reset by peer|2020-01-03 16:31:40|167760|1|996|Command: Action: CoreShowChannels
ActionID: 1578087100.979372||
2020-01-03 16:31:40|WAITFOR ERRMSG: |write error: filehandle isn't open|2020-01-03 16:31:40||1|997|Command: Action: CoreShowChannels
ActionID: 1578087101.7963||
2020-01-03 16:31:41|Asterisk server shutting down, PROCESS KILLED... EXITING|ONE DAY INTERVAL:0|write error: filehandle isn't open||
2020-01-03 16:31:41|------ UPDATER BAD GRAB!!!
0|0:350 0|0:0 0|0:0 0|0:196 0|0:25"

This issue happens randomly and at different times in different vicidial/asterisk cluster servers.

Can someone bring some light to this issue?

All appreciate it!

This message is fired by the AST_manager_listen.pl script, have you verified system load when this happen ?

ambiorixg12 wrote:This message is fired by the AST_manager_listen.pl script, have you verified system load when this happen ?

Good morning!

Yes, I have and the system load is OK. We only have about 25 agents simultaneously on each server. That's why is so odd that this is happening when we have enough system resources.

You've shown the error log from a screen that fails to contact another screen that's not running. So ... show the error log from the screen that dies. If there's no log, bring that screen up and leave it up until it dies and see what happens at the end. If that screen has a debug mode or option, turn it on.