Alexey,

I already read some of the articles from the links you provided, but answer
is still not straight forward. The relevant part states that "The new
dialog clustering support in OpenSIPS 2.4 is a complete one as it not only
takes care of dialog replication/sharing, but also of dialog handling in
terms of properly triggering dialog-specific actions."
But there is no information on how or "if" the module will  do call
representing / re-homing or restoring.

/Konrad

Hi Konrad!

Clustering OpenSIPS is some mix of:
- N OpenSIPS nodes
- N RTPEngine instances (I think you can use one, standalone,
on a different host than OpenSIPSes)
- at least 1 Redis, in case of >=2 RTPEngines, to share
RTP sessions
- some VRRP realisation, e.g. Keepalived or Pacemaker/Corosync.

OpenSIPSes share cluster-related critical data through BIN interface [1],
along with cluster-related options in some modules like usrloc [2],
dialog [3], tm [4]. Please find 'cluster' word on the documentation pages,
I hope this will give you answers on your questions.

[1] https://www.opensips.org/Documentation/Interface-Binary-2-4
[2] https://opensips.org/html/docs/modules/2.4.x/usrloc.html
[3] https://opensips.org/html/docs/modules/2.4.x/dialog.html
[4] https://opensips.org/html/docs/modules/2.4.x/tm.html

Another kind of clustering is for horizontal scaling,
not redundancy/high-availability.
A good example is registrar server
clustering:http://www.opensips.org/Documentation/Tutorials-Distributed-User-Location-Federation

And some more 
links:https://blog.opensips.org/2018/03/13/the-new-clustering-engine-of-opensips-2-4/https://blog.opensips.org/2018/03/23/clustering-ongoing-calls-with-opensips-2-4/https://blog.opensips.org/2018/03/27/clustering-presence-services-with-opensips-2-4/
https://github.com/sipwise/rtpenginehttps://github.com/sipwise/rtpengine/wiki/Redis-keyspace-notifications



On Tue, Mar 5, 2019 at 11:38 PM Konrad Malewski <
konrad.malew...@motorolasolutions.com> wrote:

> Hello,
>
> We ware trying out clusterer module with opensips. We are having problems
> understanding what is the purpose of clustering when it comes to calls, so
> if anyone has some insights we would appreciate:
>     - if in active-backup mode active fails how backup is going to recover
> the call (with one of our tries nothing happened) ? Does clusterer module
> handles it ?
>     - does the behavior differ when it comes to tcp/udp ?
>     - how does opensips support call representing / re-homing or restoring
> ?
>
> thanks for any hint,
>    Konrad
>
>
_______________________________________________
Users mailing list
Users@lists.opensips.org
http://lists.opensips.org/cgi-bin/mailman/listinfo/users

Reply via email to