Thank you for the response Mike,

I did run into a CDR bottleneck as well and have already disabled it,

> module show like cdr
Module                         Description                              Use 
Count  Status      Support Level
0 modules loaded

# grep enable= /etc/asterisk/cdr.conf
enable=no

At this point I'm really just not sure what the current bottleneck is and how 
to prevent the tasks for pooling.  I expected that the CPU would cap out before 
this occurred.  I do feel like there must be something I'm missing but just 
can't to it.

Any further suggestions are very welcome.

Thanks
Joseph
________________________________
From: [email protected] 
<[email protected]> on behalf of Mike 
<[email protected]>
Sent: Friday, September 1, 2017 3:46 PM
To: Asterisk Users Mailing List - Non-Commercial Discussion
Subject: Re: [asterisk-users] ERROR during high volume MoH dialplan

I had that problem before – I believe “task processor queue reached 500 
scheduled tasks” crashing means your CDR records (queue) are being written as 
the call ends, and if you had many thousands of entries being written to disk 
it crashes asterisk (each ring to one phone is an entry, so it goes up fast – 
for example 10 busy phones, with a between-ring delay of 1 second means every 
second there are 10 entries being put in memory)

I was using a MySQL CDR, but I had left the “CSV” type of CDR on. I 
removed/disabled the CSV CDR module, kept on the SQL CDR only and things have 
been working fine ever since.

Mike

From: [email protected] 
[mailto:[email protected]] On Behalf Of Joseph Smith
Sent: September 1, 2017 16:41
To: [email protected]
Subject: Re: [asterisk-users] ERROR during high volume MoH dialplan


Thanks for the suggestion Tony,


I installed each codec for MoH, core sounds, and extra sound packages.  
Unfortunately the tests produce the same results.

[Sep  1 20:36:45] ERROR[10081][C-00007fe5]: frame.c:343 ast_frdup: FRACK!, 
Failed assertion Excessive refcount 100000 reached on ao2 object 0x20380b0 (

continuously for a while followed by a

[Sep  1 20:36:46] WARNING[7761][C-0000770d]: taskprocessor.c:888 
taskprocessor_push: The 'subp:PJSIP/sipp-00000020' task processor queue reached 
500 scheduled tasks.

Then this time Asterisk actually crashed. :(

________________________________
From: [email protected] 
<[email protected]> on behalf of Tony Mountifield 
<[email protected]>
Sent: Friday, September 1, 2017 11:01 AM
To: [email protected]
Subject: Re: [asterisk-users] ERROR during high volume MoH dialplan

In article 
<cy4pr2201mb14643c2177c953fa27ac9e2ba8...@cy4pr2201mb1464.namprd22.prod.outlook.com>,
Joseph Smith <[email protected]> wrote:
>
> Thanks for the feedback.
>
> I do agree with having multiple smaller servers.  When I was first approached 
> with this task I mentioned as much.
> However, the current desire is to work with already existing hardware.  That 
> is out of my hands at the moment unless it
> just can't be done.  I will explore Freeswitch a bit soon to compare it as 
> well.
>
>
> I am struggling to find what the bottle neck is in this scenario.  Does 
> anyone have any advice on what that could be or
> on steps to discover it?   Do you think that tasks are pooling up because of 
> transcoding?  If so would it help to change
> the codec that is being used?  I am not sure about the MoH but the audio 
> files I am using are gsm.

You will find it less taxing on the server if you have MoH files and sounds 
files
available in all the possible native formats. Then Asterisk can use the 
appropriate
one for the channel without transcoding.

On my systems I have MoH and sounds installed in wav, ulaw, alaw, gsm and g729.

They will also sound better than transcoding from the gsm versions.

Cheers
Tony
--
Tony Mountifield
Work: [email protected] - http://www.softins.co.uk
Software Insight - Welcome<http://www.softins.co.uk/>
www.softins.co.uk
Welcome. Software Insight Ltd is a small but expert company specialising in 
software and systems development and systems administration. We pride ourselves 
in ...



Play: [email protected] - http://tony.mountifield.org
[Image removed by sender.]<http://tony.mountifield.org/>

Tony Mountifield's Home Page<http://tony.mountifield.org/>
tony.mountifield.org
Tony Mountifield's Home Page. This page is still under construction (despite 
having been started a long time ago!) It will grow as I think of more things to 
put in ...




--
_____________________________________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --
api digital - problem solved.<http://www.api-digital.com/>
www.api-digital.com
API Digital Website




Check out the new Asterisk community forum at: https://community.asterisk.org/
[Image removed by sender.]<https://community.asterisk.org/>

Asterisk Community<https://community.asterisk.org/>
community.asterisk.org
The Asterisk Community's home for Discussion ... Is there any way to share same 
queue with same agents between multiple servers in a multiple server Asterisk 
...




New to Asterisk? Start here:
      https://wiki.asterisk.org/wiki/display/AST/Getting+Started
Getting Started - Asterisk Project - Asterisk Project 
Wiki<https://wiki.asterisk.org/wiki/display/AST/Getting+Started>
wiki.asterisk.org
When learning Asterisk it is important to start off on the right foot, so this 
section of the wiki covers orientation for learning Asterisk as well as 
installation ...




asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users
-- 
_____________________________________________________________________
-- Bandwidth and Colocation Provided by http://www.api-digital.com --

Check out the new Asterisk community forum at: https://community.asterisk.org/

New to Asterisk? Start here:
      https://wiki.asterisk.org/wiki/display/AST/Getting+Started

asterisk-users mailing list
To UNSUBSCRIBE or update options visit:
   http://lists.digium.com/mailman/listinfo/asterisk-users

Reply via email to