[Freeswitch-users] Does freeswitch forks his processes?

Ken Rice krice at freeswitch.org
Tue Feb 2 17:53:10 MSK 2016


FreeSWITCH uses threading…  in modern linux kernels threading and forking are very similar… if you look in top you’ll only see 1 FS process that’s because top by default rolls up all the threads… htop on the other hand by default shows you the individual threads.

 

In FreeSWITCH there are several threads running on just a base idle FreeSWITCH process. Each addition call leg is atleast 1 more thread. Depending on which applications are actually active on a call, there could be more then 1 thread per call leg.

 

 

On the load testing and load handling capabilities this is something the FS team typically does not support on the open source side. I would recommend contacting consulting at freeswitch.org <mailto:consulting at freeswitch.org>  to get the pros involved. There are a lot of factors to take into consideration and just 1 param can make orders of magnitude difference from test to test

 

K

 

From: freeswitch-users-bounces at lists.freeswitch.org [mailto:freeswitch-users-bounces at lists.freeswitch.org] On Behalf Of Vladislav Ivanov
Sent: Tuesday, February 2, 2016 6:09 AM
To: FreeSWITCH Users Help <freeswitch-users at lists.freeswitch.org>
Subject: [Freeswitch-users] Does freeswitch forks his processes?

 

Hey guys,

I have a question about freeswitch process/threading usage.
So far that I haven't noticed freeswitch to fork himself, I have only 1 freeswitch instance.
http://i.imgur.com/bdbYOwp.png

But then I found screenshot of htop with freeswitch and noticed that there is multiple freeswitch processes being run:
http://i.imgur.com/VNpl55z.jpg

I'm having issues with "loading" the freeswitch after 50 cps in any cpu/ram configuration.
Be it physical or virtual environment I cant pass the 50 cps mark.
I have strange issue with CPU usage on same CPS:

http://i.imgur.com/8BdQWVL.png
http://i.imgur.com/mWRnoGr.png

I timeload test freeswitch with 50cps for 5+ hours, and seems like there is some kind of leak somewhere.
I have tested configuration on:
Debian 8
2 core/8 gb ram
4 core/8 gb ram (graphs are from here)
8 core/32 gb ram

and in all the tests I were not able to send more than 50 cps without CPU dropping to 0 with all system starting to respond really laggy.

Test is:
sipp -> freeswitch -> sipp

Just 1 dialpeer with bridge action. No gateways. Just simple dialplan and 1 profile...
Any advice?

Thank you all





-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freeswitch.org/pipermail/freeswitch-users/attachments/20160202/bff9bc33/attachment.html 


Join us at ClueCon 2016 Aug 8-12, 2016
More information about the FreeSWITCH-users mailing list