[
https://issues.apache.org/jira/browse/ARTEMIS-3713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17507129#comment-17507129
]
Justin Bertram commented on ARTEMIS-3713:
-----------------------------------------
I can't speak directly to the behavior discrepancy you're seeing when the
brokers are on the same machine vs. different machines. However, the whole
point of using a network file-system is to access the same files from different
machines on the same network so it's not clear to me why the behavior of NFS 3
when you run both brokers are on the same machine is ultimately relevant here.
If you want to configure shared-storage for ActiveMQ Artemis you need to use a
networked file-system that supports the proper locking semantics. From all that
I can see NFS 3 does not and NFS 4 does. Therefore I recommend you use NFS 4.
> HA with shared store not working when configured master slave in two
> different servers with data folder which is accessible by both servers
> ---------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: ARTEMIS-3713
> URL: https://issues.apache.org/jira/browse/ARTEMIS-3713
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Affects Versions: 2.13.0, 2.20.0
> Reporter: gayatri beela
> Priority: Major
> Attachments: broker_master.xml, broker_slave.xml
>
>
> HA with shared store not working when configured master slave in two
> different servers with data folder which is accessible by both servers . when
> master is down slave is working, after that if master is again live records
> in the queue are not processing even though we have put allow-failback true
> in slave
> broker.xml for master and slave are attached.
>
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
