[ejabberd] Clustering not working...

Gadi Srebnik gadi at rounds.com
Tue Aug 18 01:47:50 MSK 2015


When a node is going down, all clients connected to it should manage their
reconnection. Nothing to do with server or cluster configuration.

- If cluster is configured with the same DNS name, you should have
heath-checker that will remove bad client from list.
- In load balance based, node should be removed from LB and client should
reconnect to the same LB.
- Third option is send to client list of available nodes that he will
attempt to connect to.


On Tue, Aug 18, 2015 at 1:38 AM, Vikram Lalit <vikramlalit at gmail.com> wrote:

> Hi.. I am rather new to ejabberd so apologize if this is trivial.
>
> I've set up a ejabberd (v 15.04) cluster on AWS using 2 Ubuntu images.
> Whilst I am able to successfully cluster the two (using the command
> join_cluster from the 2nd node to the 1st node), I am not sure if the
> behavior is as expected... any thoughts would be much appreciated...
>
> To detail further, 2 different clients connected to the 2 nodes separately
> can communicate with each other. However, when I stop the server on the
> secondary node, I would still expect the two clients to be able to talk to
> each other. But instead, this 2nd client simply gets disconnected as the
> server is not running.
>
> Is there something possibly that am overlooking here?
>
> Many thanks!
>
>
>
> _______________________________________________
> ejabberd mailing list
> ejabberd at jabber.ru
> http://lists.jabber.ru/mailman/listinfo/ejabberd
>
>


-- 
Gadi Srebnik
VP Communications
 m: +972 54 5929261
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.jabber.ru/pipermail/ejabberd/attachments/20150818/84ceb3df/attachment.html>


More information about the ejabberd mailing list