Call not received by asterisk when dial level go high

[v.car94]

Hello everybody,
my problem occurs when I raise the dial level of a campaign and the calls grow.
In the realtime report the calls are still sent regularly (calls being placed) but nothing arrives in the asterisk (asterisk console). Even in carrier stats I don't detect any activity (all zero). In the asterisk console I don't detect any errors and neither do the calls that should come. The problem is solved simply by restarting the Web-Telephony server and reducing dial level or also by core stop now and restarting asterisk after a while. If I leave the dial level at low values the problem does not arise.
I cannot understand what happens and how to understand what is limiting my system.
I've enabled recording calls, Max Trunks:600,Max Calls per Second:20
(I never reached 600 calls, I stay under 150 call being placed)

My configuration is as follows:
1 Server Database:
Cpu: Xeon 8c Silver 4110 (8x core 2.10 GHz)
Ram: 32 GB
Disk: 2 x 4TB SATA 7.2k rpm

1 Web-Dialer Server:
CPU: Intel Quad Core
Ram: 8GB
240 GB SSD Hard Disk

Installed with vicibox 7.X | Vicidial Version: 2.14b0.5 | BUILD: 170609-1529 | SVN Version: 3010 |Asterisk Version 11.25.1-vici

If do you need other info I'm here
Thank's to all for any support

[williamconley]

"high" is not a number. "I stay under 150 calls" ... is that when you set it "high" or when you are normal and low?

Total calls shown in Real Time and total calls shown in "core show channels" in asterisk would be useful. Plus of course: System Load. Use either htop or uptime, this is also usually shown at the bottom of the real time screen in a single server system.

Then we start with Tracing of the issue:

1) You set the values for the campaign
2) The decision to dial or not dial numbers for each campaign is made by the script running in the autodial screen (screen -ls).
3) That script stores call requests in the Vicidial Manager table.
4) The manager scripts (send and listen, running in send and listen screens of course) then manage the actual calls.

Somewhere in that mess your system is dropping the ball! Happy Hunting!

[v.car94]

Thank's for the answer.
I also reached 400 calls placed with success but it crashed in the same situation to reach 50 calls placed.
I don't know how to find where the problem is. If someone can help me

This is what screen -ls output when system is crashed

Code: Select all

There are screens on:
        6665.ASTVDremote        (Detached)
        6662.ASTVDauto  (Detached)
        6659.ASTlisten  (Detached)
        6656.ASTsend    (Detached)
        6653.ASTupdate  (Detached)
        6420.asterisk   (Detached)
        6402.astshell20201026092203     (Detached)
        1321.ASTfastlog (Detached)
        1318.ASTVDadapt (Detached)
9 Sockets in /var/run/screens/S-root.


The same result (screen -ls) is obtained after restarting asterisk from shell ( "core stop now" in the console and then "asterisk"). After the restart the system is correctly working.

I noticed that the system crash if I increase or decrease too fast the number of call placed. Example:
if i change number of calls placed from 50 to 250 the system probably will crashy.
If i change number of calls placed from 50 to 60 to 80 .... until 250 the system probably will work fine

Thank's

[carpenox]

That seems high for the hardware specs you specified for the dialer. only quad core with 8gb? WHa does the CPU% go up to during peak dialing times? Why not upgrade to vicibox 9.0.3?

[v.car94]

My question is:
what is my bottleneck? Dialer RAM, Dialer CPU and why I can solve problem restarting Asterisk?
System Load Average: 1.18 0.45 0.36 M (with 120 calls)

I don't think that problem is hardware because sometimes it crashes in a call placed value that other time it's not crashed. (sorry for english)
Is there some idea?

Why can help me the upgrade to vicibox 9?

[carpenox]

Backup your system, the db, install vicibox 9.0.3 by ISO, restore your backup db into it, I can help you do it if you need, however it is not hard.
https://itgala.xyz/vicidial-backup-and-restore-guide/

Contact me through my business website if you need help : www.cyburityllc.com

[v.car94]

Thank's for answer,
I've found the problem!

I noticed that in my CLI log there was some error, before the crash, about the limit of file opened reached. This error appear only one time then the system crashes and in the log I could see only the calls placed before the crash. I solve editing the file:

Code: Select all

/etc/security/limits.conf

And adding at the end these row:

Code: Select all

* soft nofile 65536
* hard nofile 65536


Then reboot and the problem is solved.
I find this solution on google searching for the error found in the CLI log and I don't know if this solve definitively the problem.

I hope that this can help someone with my same problem.
If someone think that the value added in that file are wrong can tell me why and how to correctly solve the problem.

Thank's to all

[carpenox]

great postback

[williamconley]

Code: Select all

cat /proc/sys/fs/file-max
ulimit -Hn
ulimit -Sn