[jira] [Commented] (ARTEMIS-285) Bridge static connectors must be identical to the server connectors for failover

["clebert suconic (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/ARTEMIS-285?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14978497#comment-14978497 ] 

clebert suconic commented on ARTEMIS-285:
-----------------------------------------

AFAIK you are supposed to use the same address you use on the cluster connection... I'm not sure this is fixable.


It's likely a won't fix IMHO.

> Bridge static connectors must be identical to the server connectors for failover
> --------------------------------------------------------------------------------
>
>                 Key: ARTEMIS-285
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-285
>             Project: ActiveMQ Artemis
>          Issue Type: Bug
>    Affects Versions: 1.1.0
>            Reporter: Jeff Mesnil
>
> * one live server with a remote connector
> * one backup server with a remote connector
> * a 3rd server with a core bridge using a static connector to the live server.
> During failover (when the live server is killed and the backup server fails over), the bridge server will be notified that the backup server is now UP.
> It will then figure out if it corresponds to its target by comparing its connector configuration to the one in the TopologyMember.
> I have tests that were failing because the connectors were not identical (one was using localhost, the other was using 127.0.0.1).
> This setup is brittle as it "forces" the bridge to use the same connector configuration than the one defined by the backup.
> Adding only one different parameter (such as use-nio=false) would make the comparison fail and the bridge would not fail over to the backup server.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)