Page 1 of 1

odd system freeze up

PostPosted: Mon Aug 20, 2018 3:03 pm
by TwistedFister
Vicibox 7.0.3 from .iso | Vicidial 2.14-579a Build 161128-1746 | Asterisk 11.22.0-vici | CLuster setup: 1 web 1 DB 1 Slave DB 8 telephony | No Digium/Sangoma Hardware | No Extra Software After Installation

Outbound ratio dialing and all of the agent screens and the realtime report completely freeze for 3-4 minutes however all 110 of the agents remain on the calls they are on but cant hang up or dispo the calls. After the 3-4 minutes the realtime report refreshes and everyone is either paused (which the agents didnt do on their own) or waiting for a call and it resumes dialing just as it has been for the last 2 years... Anyone seen this before? First time for me

Re: odd system freeze up

PostPosted: Mon Aug 20, 2018 3:25 pm
by williamconley
Given the odds: Your server has matured. You now have so many records in one or more tables that running a large report or any fresh configuration changes such as a new sql filter on a campaign can hammer the DB server and lock tables for quite a while and regularly.

Test which mysql query has been running the longest (ie: the problem) with:
Code: Select all
watch -n1  "mysql asterisk -pROOTMYSQLPASSHERE -e \"select * from information_schema.processlist where command not in ('Sleep') and info not like '%processlist%' and info is not null \""


If you have multiple millions of Vicidial Leads in the system, consider deleting as many as possible. Try to get below 5M if you can. Note that this should happen before the nightly DB optimization routine runs (shows in the crontab, probably 1AM-ish).

You can also activate the log archiving system (also in crontab) which can dramatically improve performance on mature systems that have been collecting logs forever.

If this is a reporting triggered event, consider a reporting mysql/replication server.

Happy Hunting 8-)