[OpenSIPS-Users] WARNING:core:utimer_ticker: utimer task <tm-utimer>

xaled xaled at web.de
Fri Nov 9 10:08:51 EST 2018


Hi Bogdan,

 

I found the source of the problem. It is the use of localhost to connect to local postgresql database.

I changed it to 127.0.0.1 and opensips starts fast and without warnings. 

 

before:

modparam("db_virtual", "db_urls", "postgres://user:pass@localhost/db")

 

after:

modparam("db_virtual", "db_urls", "postgres://user:pass@127.0.0.1/db")

 

 

There is this warning I don’t understand:

 

Nov  8 09:52:40 fra-appsrv01 /usr/local/sbin/opensips[12635]: WARNING:avpops:avpops_db_bind: async() calls for

DB URL [default] will work in normal mode due to driver limitations

 

And once or twice a day I get timer warnings, but it is way better than it was before. 

 

Nov  8 09:59:45 fra-appsrv01 /usr/local/sbin/opensips[12640]: WARNING:core:handle_timer_job: timer job <tm-time

r> has a 70000 us delay in execution

Nov  8 09:59:45 fra-appsrv01 /usr/local/sbin/opensips[12641]: WARNING:core:handle_timer_job: timer job <blcore-

expire> has a 70000 us delay in execution

 

Nov  8 18:10:10 fra-appsrv01 /usr/local/sbin/opensips[12640]: WARNING:core:handle_timer_job: timer job <tm-time

r> has a 60000 us delay in execution

Nov  8 18:10:10 fra-appsrv01 /usr/local/sbin/opensips[12640]: WARNING:core:handle_timer_job: timer job <db_virt

ual-reconnect> has a 60000 us delay in execution

Nov  8 18:10:10 fra-appsrv01 /usr/local/sbin/opensips[12640]: WARNING:core:handle_timer_job: timer job <blcore-

expire> has a 60000 us delay in execution

 

Thanks,

Xaled

From: Bogdan-Andrei Iancu [mailto:bogdan at opensips.org] 
Sent: Friday, November 09, 2018 3:00 PM
To: xaled <xaled at web.de>; users at lists.opensips.org
Subject: Re: [OpenSIPS-Users] WARNING:core:utimer_ticker: utimer task <tm-utimer>

 

Hi Xaled,

(switching back to the list)

Again, it seems to be related to your postgres server - the backtrace shows that all opensips procs are stuck while trying to connect to the postgres server. Why the connect phase takes so long? not sure, it is a postgres stuff.

Try setting this param to 1 or 2:
    http://www.opensips.org/html/docs/modules/2.4.x/db_postgres.html#param_timeout

And see how it works.

Regards,



Bogdan-Andrei Iancu
 
OpenSIPS Founder and Developer
  http://www.opensips-solutions.com
OpenSIPS Bootcamp 2018
  http://opensips.org/training/OpenSIPS_Bootcamp_2018/

On 11/08/2018 12:14 PM, xaled wrote:

Hi Bogdan,

 

here is the compiled version from git sources:

 

root at fra-appsrv01:/usr/local/etc/opensips# /usr/local/sbin/opensips -V

version: opensips 2.4.2 (x86_64/linux)

flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, F_MALLOC, FAST_LOCK-ADAPTIVE_WAIT

ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535

poll method support: poll, epoll, sigio_rt, select.

git revision: a9b9169

main.c compiled on 19:30:38 Nov  7 2018 with gcc 4.9.2

 

and here is the debian repo version:

 

root at fra-appsrv01:/usr/local/etc/opensips# /usr/sbin/opensips -V

version: opensips 2.4.2 (x86_64/linux)

flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, F_MALLOC, FAST_LOCK-ADAPTIVE_WAIT

ADAPTIVE_WAIT_LOOPS=1024, MAX_RECV_BUFFER_SIZE 262144, MAX_LISTEN 16, MAX_URI_SIZE 1024, BUF_SIZE 65535

poll method support: poll, epoll, sigio_rt, select.

main.c compiled on  with gcc 4.9.2

 

Greetings,
Xaled

 

 

From: Bogdan-Andrei Iancu [mailto:bogdan at opensips.org] 
Sent: Thursday, November 08, 2018 10:21 AM
To: xaled  <mailto:xaled at web.de> <xaled at web.de>
Subject: Re: [OpenSIPS-Users] WARNING:core:utimer_ticker: utimer task <tm-utimer>

 

Hi Xaled,

What's the exact version / revision of OpenSIPS you are using ?

Best regards,




Bogdan-Andrei Iancu
 
OpenSIPS Founder and Developer
  http://www.opensips-solutions.com
OpenSIPS Bootcamp 2018
  http://opensips.org/training/OpenSIPS_Bootcamp_2018/

On 11/08/2018 01:44 AM, xaled wrote:

Hi Bogdan,

 

Is there something you can recommend to debug and fix it? 

 

The database is in the same VM as opensips and there are never time out warnings during the actual database access. Warnings come either at startup, or randomly if opensips gets through startup phase.

 

I will take a closer look at the database connection performance. It makes no sense to me though, that warning flood started after migrating the opensips VM to a newer hardware without changing anything else and then partially stopped after using compiled version of opensips vs installed from Debian repo on the same VM.

 

The older Debian VM with the same opensips/db  setup on a different physical server does not have the startup warning flood at all. Only random warning during run time.

 

Appreciate your time,

Xaled

 

 

On Nov 7, 2018 22:18, Bogdan-Andrei Iancu  <mailto:bogdan at opensips.org> <bogdan at opensips.org> wrote:

Hi Xaled, 

As I suspected, all the worker processes in OpenSIPS are blocked in the 
startup sequence (so called child_init routine) trying to connect to the 
postgres database. This is holding up without rejecting or accepting the 
connection from OpenSIPS.

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20181109/56f5747b/attachment-0001.html>


More information about the Users mailing list