[ 
https://issues.apache.org/jira/browse/ARTEMIS-803?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15585400#comment-15585400
 ] 

ASF subversion and git services commented on ARTEMIS-803:
---------------------------------------------------------

Commit c1ecddcb717be023d1d852c53325a9b08286fbd0 in activemq-artemis's branch 
refs/heads/master from [~jmesnil]
[ https://git-wip-us.apache.org/repos/asf?p=activemq-artemis.git;h=c1ecddc ]

[ARTEMIS-803] Fix colocated backups with http-upgrade acceptor

*  Do not offset ports for Netty connector/acceptor with http-upgrade
enabled.
* Pass the name of the ActiveMQ server to the HTTP request to initiate
the Upgrade so that the HTTP endpoint on the app server can find the
correct ActiveMQ broker that must handle the upgrade.

JIRA: https://issues.apache.org/jira/browse/ARTEMIS-803


> Do not offset port for http-upgrade acceptor for colocated backups
> ------------------------------------------------------------------
>
>                 Key: ARTEMIS-803
>                 URL: https://issues.apache.org/jira/browse/ARTEMIS-803
>             Project: ActiveMQ Artemis
>          Issue Type: Bug
>          Components: Broker
>    Affects Versions: 1.4.0
>            Reporter: Jeff Mesnil
>
> In our application server, we use an embedded Artemis server with netty 
> acceptors configured to enable HTTP Upgrade.
> This means that the app server is using the HTTP port to negotiate a HTTP 
> upgrade handshake with Artemis client. If the handshake is succesful, we 
> transfer the connection from our HTTP handler to Artemis netty channel.
> This causes issue with colocated backups as Artemis offsets all the ports for 
> the Netty acceptor. If HTTP Upgrade is enabled, the port should not be offset 
> (as it is "owned" by the app server).
> Additionally, with colocated backups we have a single entry point (the app 
> server's HTTP port) that can be used by multiple Artemis server (the main one 
> and any of its colocated backups). When Artemis sends the HTTP request to 
> initiate the upgrade, it should pass the name of the Artemis server that 
> should handle the upgrade so that the app server can delegate the actual 
> handshake to the correct Artemis server.



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

Reply via email to