[OpenSIPS-Users] Dispatcher module disable gateway after X failures, auto re-enable after some time

[Adam Raszynski]

My first idea was to create failure counter for each gateway (variable in
local cache)

Than in onreply route to increment counter every time call is not
established, and reset counter if any call was successful. So we get
consecutive failures counter in cache and can make further decisions basing
on that information

Does it make sense? Is local cache fast enough (and causes no locking
problems) to be incremented in such intensive way?

Is it possible to change dispatcher gateway status from inside script? Or
it's possible only by using external interface?

2017-10-31 12:31 GMT+01:00 Bogdan-Andrei Iancu <bogdan at opensips.org>:

> Hi Adam,
>
> Maybe such a custom logic is outside OpenSIPS scope - still have you
> considered using the ds_set_state MI function to control from outside the
> status of the destinations ? Of course you will have to use the Event
> Interface too, to receive the notifications on the destination status (when
> OpenSIPS is doing the state transition) - see the E_DISPATCHER_STATUS event
> (http://www.opensips.org/html/docs/modules/2.3.x/dispatcher.
> html#idp5746368)
>
> So, when OpenSIPS switches a destination into probing (based on a failure
> detection), you can get the event and force the destination into Inactive ;
> after some custom amount of time, you can put it back into Probing mode, to
> allow OpenSIPS to re-enable it.
>
> Regards,
>
> Bogdan-Andrei Iancu
>   OpenSIPS Founder and Developer
>   http://www.opensips-solutions.com
>
> On 10/28/2017 04:13 PM, Adam Raszynski wrote:
>
> Hi
>
> Could you suggest some working configuration of the dispatcher module for
> the following setup:
>
> - Flag gateway as failed (even if it's still responding to OPTIONS) after
> some number of consecutive call failures. That's the easy part
>
> - Automatically un-flag gateway after some time, to allow re-checking
> (reset failure counter?)
>
> - Ideal solution would also include increasing re-check interval, ie first
> re-check after 1 minute, if next failure detected next re-check after 2
> minutes, next after 4 and so on
>
>
> Best Regards
>
>
>
> _______________________________________________
> Users mailing listUsers at lists.opensips.orghttp://lists.opensips.org/cgi-bin/mailman/listinfo/users
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensips.org/pipermail/users/attachments/20171031/4b558638/attachment.html>