[Freeswitch-users] CPU sys load issue

[Antonio Silva]

Hi Juan,

Thanks for the tips!!

Unfortunately i cannot test with -nosql option, i need to have the 
sip_registration fill to be able to send options to natted endpoints. 
But i don't think the issue is in db, i use postgres and to get better 
performance from it i run it on tmpfs.


I read about tcmalloc, but from jira FS-7315, i get the idea that it 
didn't change anything.. but i guess i can try it and post the results, 
from the bt i do see some calls been done to malloc, but not sure if 
they are locking or not.

Regards,


On 02/23/2017 06:52 PM, Juan Pablo Godoy Macari wrote:
> Hello Antonio,
>
> I had the same issue, to fix it i tried many things written in this 
> mail list. Try running FS with -nosql option, if that doesn't work 
> recompile FS with tcmalloc (from google's perfomance tools).
>
> That did it for me.
>
> Regards,
> Juan Pablo.
>
> 2017-02-23 13:26 GMT-03:00 Antonio Silva <asilva at wirelessmundi.com 
> <mailto:asilva at wirelessmundi.com>>:
>
>     Hi,
>
>     i'm having an issue with FS that gets lock every morning with cpu sys
>     hitting 80%, because is limited in fs with min-idle, it could
>     reach the
>     100%, so the calls are rejected by load, to solve the problem i
>     kill fs
>     and start it again.
>
>     I discard a load issue, because after restarting freeswitch the system
>     increases in number of sessions/cps and no problems at all until next
>     day.... it's given me crazy... i've enough memory, lots of cores...
>     after restart fs the system stays at less than 30% of cpu usage.
>
>
>     Right now i have hpet as clocksource, transcoding is minimal (use
>     inheric_codec, only happens because of different ptimes with some
>     endpoints) .
>
>
>     anyone having similar problem? or any tip i can do to debug this
>     issue?
>
>
>     For info the Jira ticket can be located at
>     https://freeswitch.org/jira/browse/FS-9957
>     <https://freeswitch.org/jira/browse/FS-9957>,
>
>     i've done a gcore, top, htop when the system is with high cpu sys
>     but i
>     can't figure the thread that is causing the issue... initial i thought
>     it was because of module lua (most of my logic is done there), but
>     today
>     only 2 threads where running mod_lua when it happen.
>
>
>     Appreciate the help.
>
>
>     --
>
>     Saludos / Regards / Cumprimentos,
>     António silva
>
>
>     _________________________________________________________________________
>     Professional FreeSWITCH Consulting Services:
>     consulting at freeswitch.org <mailto:consulting at freeswitch.org>
>     http://www.freeswitchsolutions.com
>     <http://www.freeswitchsolutions.com>
>
>     Official FreeSWITCH Sites
>     http://www.freeswitch.org
>     http://confluence.freeswitch.org <http://confluence.freeswitch.org>
>     http://www.cluecon.com
>
>     FreeSWITCH-users mailing list
>     FreeSWITCH-users at lists.freeswitch.org
>     <mailto:FreeSWITCH-users at lists.freeswitch.org>
>     http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
>     <http://lists.freeswitch.org/mailman/listinfo/freeswitch-users>
>     UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
>     <http://lists.freeswitch.org/mailman/options/freeswitch-users>
>     http://www.freeswitch.org
>
>
>
>
> _________________________________________________________________________
> Professional FreeSWITCH Consulting Services:
> consulting at freeswitch.org
> http://www.freeswitchsolutions.com
>
> Official FreeSWITCH Sites
> http://www.freeswitch.org
> http://confluence.freeswitch.org
> http://www.cluecon.com
>
> FreeSWITCH-users mailing list
> FreeSWITCH-users at lists.freeswitch.org
> http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
> UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
> http://www.freeswitch.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freeswitch.org/pipermail/freeswitch-users/attachments/20170223/0d812d52/attachment.html