[OpenSIPS-Users] Dispatcher within a K8s environment

[Bogdan-Andrei Iancu]

Hi Jonathan,

I guess this will be a good topic (DS and K8S) for the OpenSIPS Summit 
in Athens - I think this is the 3rd time in the last week coming across 
it :)

Unfortunately there is no way to skip at the moment that DNS failure 
when loading the destinations :(....even more, there some code that 
relies on the fact that there is an "IP" attached to any 
destination.....And I just checked, a local error in sending the ping 
(like the DNS err) does not results in marking the destination as failed 
or so..... so it is not so straight as ignoring the DNS error.

Best regards,

Bogdan-Andrei Iancu

OpenSIPS Founder and Developer
   https://www.opensips-solutions.com
OpenSIPS Summit 27-30 Sept 2022, Athens
   https://www.opensips.org/events/Summit-2022Athens/

On 8/24/22 12:24 AM, Jonathan Hunter wrote:
>
> Hi All,
>
> I have a query around dispatcher behaviour, I am running 3.2 in a k8s 
> environment.
>
> I have 2 freeswitch instances defined in a destination set, both of 
> which are pods.
>
> As people may be aware its fun implementing in k8s as pods can restart 
> and disappear at times so I ideally want this reflected in the cache 
> and output of opensips-cli -x mi ds_list where I was hoping the 
> freeswitch entries would be defined but with a state of probing or 
> inactive.
>
> With my current setup, when restarting opensips for example, I have 
> the dispatcher table populated in postgres db , and if opensips cant 
> resolve the URI it wont load it into cache, like wise if opensips is 
> running and freeswitch pod drops, I see this in the logs;
>
> Aug 23 21:22:01 [55] ERROR:dispatcher:add_dest2list: could not resolve 
> freeswitch-opensips-deployment-1.freeswitch-opensips, skipping it
>
> Aug 23 21:22:01 [55] WARNING:dispatcher:ds_load_data: failed to add 
> destination 
> <sip:freeswitch-opensips-deployment-1.freeswitch-opensips:5070;transport=tcp> 
> in group 10
>
> I therefore don’t see it listed in cache when I run ds_list.
>
> Does anyone know if its possible to tweak dispatcher to always load 
> the database entries into cache at startup, and also set their status 
> to probing/inactive if not reachable due to a resolving issue as above?
>
> My dispatcher settings are;
>
> #### Dynamic routing
>
> loadmodule "dispatcher.so"
>
> modparam("dispatcher", "db_url", "postgres://x.x.x.x/opensips")
>
> modparam("dispatcher", "ds_probing_mode", 1)
>
> modparam("dispatcher", "ds_probing_threshhold", 1)
>
> modparam("dispatcher", "persistent_state", 0)
>
> modparam("dispatcher", "ds_ping_interval", 5)
>
> modparam("dispatcher", "table_name", "dispatcher")
>
> modparam("dispatcher", "cluster_id", 1)
>
> Hope that makes sense!
>
> Many thanks
>
> Jon
>
>
> _______________________________________________
> 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/20220824/b4e54022/attachment-0001.html>