Re: [OpenSIPS-Users] Clarification on Auto-scaling in Opensips

[Bogdan-Andrei Iancu]

Hi,

1. See the load (per proc) related stats [1] - they are used to check 
the combined load of the procs in the same auto-scaling group


2. I would say your syslog is very slow and the overhead with extra 
logging (from level 3 to 4) is triggering the up scaling.



[1] 
https://www.opensips.org/Documentation/Interface-CoreStatistics-3-2#toc13


Regards,

Bogdan-Andrei Iancu

OpenSIPS Founder and Developer
  https://www.opensips-solutions.com
OpenSIPS eBootcamp 2021
  https://opensips.org/training/OpenSIPS_eBootcamp_2021/

On 10/21/21 3:15 PM, Saurabh Chopra wrote:

Hi All,

We have successfully tested the auto-scaling in opensips after reading 
the blog.
_https://blog.opensips.org/2019/02/25/auto-process-scaling-a-cure-for-load-and-resources-concerns/ 
_ 



But, I am confused with two things,
1. How opensips calculates load to scale the processes UP/DOWN. Is 
there any function that we can use to check the load or any kind of 
xlog so that we can print in our config.
2. The config which i am running is scaling up the process when we are 
adding (*log_level=4*) not for *log_level=3*. Is it specific that we 
have to use log_level=*4*.


version: opensips 3.1.0 (x86_64/linux)
flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, 
Q_MALLOC, F_MALLOC, HP_MALLOC, DBG_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.
svn revision: 3748:3779M
main.c compiled on 06:18:30 Feb  1 2021 with gcc 5.4.0

Below is the basic of my configuration

*log_level=4
log_stderror=no
log_facility=LOG_LOCAL5
memlog=2
disable_dns_blacklist=no

auto_scaling_profile = PROFILE_SIP
 scale up to 6 on 40% for 4 cycles within 5
 scale down to 2 on 1% for 8 cycles

socket=udp:20.0.xx.xxx:5507 AS 54.227.xx.xx:5507 
use_auto_scaling_profile PROFILE_SIP*


Best Regards
Saurabh Chopra
+918861979979

___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users


___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

[OpenSIPS-Users] Clarification on Auto-scaling in Opensips

[Saurabh Chopra]

Hi All,

We have successfully tested the auto-scaling in opensips after reading the
blog.
*https://blog.opensips.org/2019/02/25/auto-process-scaling-a-cure-for-load-and-resources-concerns/
*


But, I am confused with two things,
1. How opensips calculates load to scale the processes UP/DOWN. Is there
any function that we can use to check the load or any kind of xlog so that
we can print in our config.
2. The config which i am running is scaling up the process when we are
adding (*log_level=4*) not for *log_level=3*. Is it specific that we have
to use log_level=*4*.

version: opensips 3.1.0 (x86_64/linux)
flags: STATS: On, DISABLE_NAGLE, USE_MCAST, SHM_MMAP, PKG_MALLOC, Q_MALLOC,
F_MALLOC, HP_MALLOC, DBG_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.
svn revision: 3748:3779M
main.c compiled on 06:18:30 Feb  1 2021 with gcc 5.4.0

Below is the basic of my configuration











*log_level=4log_stderror=nolog_facility=LOG_LOCAL5memlog=2disable_dns_blacklist=noauto_scaling_profile
= PROFILE_SIP scale up to 6 on 40% for 4 cycles within 5 scale down to 2 on
1% for 8 cyclessocket=udp:20.0.xx.xxx:5507 AS 54.227.xx.xx:5507
use_auto_scaling_profile PROFILE_SIP*

Best Regards
Saurabh Chopra
+918861979979
___
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users