[OpenSIPS-Users] Major performance issues question
Liviu Chircu
liviu at opensips.org
Wed Dec 21 12:25:56 EST 2016
Okay, so it seems lots of CPU cycles are spent in the kernel!
Are you sure there isn't any intensive I/O going on? For example, maybe
you are pushing lots of gateway pings. You could run "iostat 1" to
assess this.
Regards,
Liviu Chircu
OpenSIPS Developer
http://www.opensips-solutions.com
On 21.12.2016 18:58, Schneur Rosenberg wrote:
> procs -----------memory---------- ---swap-- -----io---- -system--
> ------cpu-----
> r b swpd free buff cache si so bi bo in cs us sy
> id wa st
> 7 0 0 9966412 296728 5239152 0 0 3 60 0 0 0
> 0 99 0 0
> 6 0 0 9966684 296728 5239152 0 0 0 20 2075 1569 12
> 64 24 0 0
> 7 0 0 9966588 296728 5239152 0 0 0 20 2023 1408 12
> 64 24 0 0
> 4 0 0 9966492 296728 5239152 0 0 0 108 2027 1434 11
> 63 26 0 0
> 5 0 0 9966588 296728 5239152 0 0 0 44 1687 1232 11
> 52 38 0 0
> 5 0 0 9966620 296728 5239156 0 0 0 72 1704 1136 9
> 53 37 0 0
> 6 0 0 9966492 296728 5239156 0 0 0 40 1874 1133 11
> 64 25 0 0
> 6 0 0 9966524 296728 5239156 0 0 0 20 1794 853 11
> 65 25 0 0
> 6 0 0 9966556 296728 5239156 0 0 0 56 1810 951 11
> 64 25 0 0
> 6 0 0 9966620 296728 5239156 0 0 0 0 1802 934 10
> 65 25 0 0
> 7 0 0 9966556 296728 5239160 0 0 0 2328 2356 1489 12
> 76 12 0 0
> 7 0 0 9966460 296728 5239160 0 0 0 0 2074 929 12
> 76 12 0 0
> 7 0 0 9966588 296728 5239164 0 0 0 0 2022 728 11
> 76 12 0 0
> 7 0 0 9966620 296728 5239164 0 0 0 0 2015 802 14
> 74 12 0 0
> 5 0 0 9966652 296728 5239164 0 0 0 72 2275 1049 14
> 81 5 0 0
>
>
> On Wed, Dec 21, 2016 at 6:42 PM, Liviu Chircu <liviu at opensips.org
> <mailto:liviu at opensips.org>> wrote:
>
> Hi Schneur,
>
> Those warnings are due to a rather harmless bug, which was fixed
> and backported ~ two weeks ago.
>
> Regarding 100% CPU usage though, could you please run "vmstat 1"
> for 10 seconds and paste the full output? Thanks!
>
> Regards,
>
> Liviu Chircu
> OpenSIPS Developer
> http://www.opensips-solutions.com <http://www.opensips-solutions.com>
>
> On 21.12.2016 18:17, Schneur Rosenberg wrote:
>> Hi, I'm running a instance of OpenSIPS (just signalling no RTP on
>> this machine) on a DigitalOcean VM, it was running fine for a
>> while and it does not process lots of CPS (4-5 max) and today I
>> was getting complaints of dropped calls, I checked the logs and I
>> saw lots of WARNING:core:utimer_ticker: utimer task <tm-utimer>
>> already scheduled for 3956760 ms (now 3956860 ms), it may overlap
>> and WARNING:core:handle_timer_job: utimer job <tm-utimer> has a
>> 30000 us delay in execution
>> I had 4 children and I upped it to 15, I think it helped but not
>> completely solved the problem. When running top I see all of the
>> sudden a few opensips processes jumping to 100% and the idle
>> resources slowly going down sometimes even to 0% then it
>> recovers, I haven't changed anything in my cfg file lately and
>> the traffic has not increased etc. Is there anyway to see in the
>> logs whats taking so much resources? I'm also wondering if its
>> can be a DigitalOcean issue where other users on the same
>> Hyperviser are using very high resources and its stealing my
>> resources.
>> thank you
>>
>> _______________________________________________
>> Users mailing list
>> Users at lists.opensips.org <mailto:Users at lists.opensips.org>
>> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
>> <http://lists.opensips.org/cgi-bin/mailman/listinfo/users>
> _______________________________________________ Users mailing list
> Users at lists.opensips.org <mailto:Users at lists.opensips.org>
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
> <http://lists.opensips.org/cgi-bin/mailman/listinfo/users>
>
> _______________________________________________
> Users mailing list
> Users at lists.opensips.org
> http://lists.opensips.org/cgi-bin/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20161221/f566e2fa/attachment.html>
More information about the Users
mailing list