[mod_cluster-issues] [JBoss JIRA] (MODCLUSTER-438) WebSocket support for mod_cluster

[Michal Karm Babacek (JIRA)]

    [ https://issues.jboss.org/browse/MODCLUSTER-438?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13290690#comment-13290690 ] 

Michal Karm Babacek commented on MODCLUSTER-438:
------------------------------------------------

[~gfleury] Thank you. The current version of httpd/mod_cluster supports websockets via proxy_wstunnel.  Does nginx/mod_cluster anything differently? e.g. what happens when worker dies? Does the balancer try to re-negotiate handshake with a different worker or it's up to the client to do so?

We arrived at the conclusion that there is no meaning in having websocket "ha", so when worker dies, the communication is broken and client has to initiate it again - balancer would give the client another, healthily worker then.

> WebSocket support for mod_cluster
> ---------------------------------
>
>                 Key: MODCLUSTER-438
>                 URL: https://issues.jboss.org/browse/MODCLUSTER-438
>             Project: mod_cluster
>          Issue Type: Feature Request
>          Components: Native (httpd modules)
>    Affects Versions: 1.3.2.Final
>            Reporter: Michal Karm Babacek
>            Assignee: Jean-Frederic Clere
>             Fix For: Awaiting Volunteers
>
>
>  * take a look at mod_proxy_wstunnel
>  * mod_cluster should be able to load balance WebSocket connections to worker nodes
>  * consider high-availability with respect to WebSockets
> Additional details TBD...



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)