HELP!!! 3 to 4 agent VD install, delay/slow to hang up calls

All installation and configuration problems and questions

Moderators: gerski, enjay, williamconley, Op3r, Staydog, gardo, mflorell, MJCoate, mcargile, Kumba, Michael_N

HELP!!! 3 to 4 agent VD install, delay/slow to hang up calls

Postby dfwonsite » Wed Mar 21, 2007 5:02 pm

I am fairly versed in *, but am brand new to ViciDial, I am in a position to land myself a full time position maintaining and running this dialer with a friend's company, all contingent on getting this setup working :/

The problem is this -

Auto Dial works
Manual Dial Works (Adapt or Ratio)

but, the calls are EXTREMELY SLOW to hang up if there is more than 1 agent logged in to the system. The system specs are as follows -

CPU - AMD 64 3500+
RAM - 512 MB
HDD - 120GB Western Digital UDMA-100
MOBO ECS

OS - Centos 4.4
Asterisk - Tried both 1.4 and 1.2.16
ViciDial - 2.0.3b1 - problem was also present in 2.0.2 hence the U/G.
Agents - SIP (Snom 300 Hard phones)
Trunks - IAX2 over to the TrixBox phone system (this works fine, VD is on a seperate server :) )

Zaptel 1.2.15 (used 1.4.0 w/ * 1.4)

The load page says -

VICIDIAL: Server Performance 2007-03-21 16:59:53
Time range: 2007-03-20 08:45:00 to 2007-03-20 15:30:00

---------- TOTALS, PEAKS and AVERAGES
Total Calls in/out on this server: 44
Total Off-Hook time on this server (min): 4.03
Average channels in use for server: 8.0031
Average load for server: 7.9253
Peak load for server: 92
Average USER process cpu percentage: 3.8142 %
Average SYSTEM process cpu percentage: 1.1278 %
Average IDLE process cpu percentage: 95.0970 %

---------- LINE GRAPH:

and what else you may need?


PLEASE tell me somebody has had a similar problem and has some fixes i haven't thought of :) If i get this fixed, i have a very nice full time job w/ good pay & benefit's waiting for me!! :)

-Andy
dfwonsite
 
Posts: 9
Joined: Sun Mar 18, 2007 12:08 pm

Postby mflorell » Wed Mar 21, 2007 6:35 pm

This might be that phantom CentOS problem. It only happens to some people on CentOS. The problem is that perl process prioritization "nice" is somehow turned down by default and the perl scripts execute after they are supposed to. Using any distro other than CentOS has solved this problem from everyone I have heard from.

It is possible that it could be another issue, but that is the most prevelent one.

Can you post some Real Asterisk CLI output with AGI STDERR turned on for the server when this happens?
mflorell
Site Admin
 
Posts: 18334
Joined: Wed Jun 07, 2006 2:45 pm
Location: Florida

Postby dfwonsite » Thu Mar 22, 2007 11:37 am

that had crossed my mind, so i pulled the drive the current install is on and am in the process of reinstalling using slackware 11, will advise on the outcome, i would like to put the centos installed drive into another server and further trace this down, but for right now i need to get a workable system in place :)

-Andy
dfwonsite
 
Posts: 9
Joined: Sun Mar 18, 2007 12:08 pm


Return to Support

Who is online

Users browsing this forum: No registered users and 44 guests