You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by Mark Greene <gr...@gmail.com> on 2011/05/19 16:12:30 UTC

failover transport recovery behavior

I have a primary and secondary node configured with the failover transport.
When the primary node dies, producers and consumers switch over to the
secondary node just fine. However, when the primary node comes back, only
the producer seems to recover and migrate back to the primary, however the
consumer stays connected to the secondary node. Is this expected behavior?
I'm using v5.3.2.

Thanks in advance.

-Mark

Re: failover transport recovery behavior

Posted by Jason Whaley <ja...@gmail.com>.
5.3.x of ActiveMQ is fairly old and lots of things have been fixed since then in the current 5.5 version.    You should consider upgrading.  

In addition, as of 5.4 there is a "rebalanceClusterClients" option you can set on your transportConnector (in conjunction with "updateClusterClients") that should proactively disconnect clients from their current broker when there is a change in your current cluster... such as a previously unavailable broker becoming available again.

See http://activemq.apache.org/failover-transport-reference.html for documentation on the failover transport and http://bsnyderblog.blogspot.com/2010/10/new-features-in-activemq-54-automatic.html or more detail on those two options.

--Whaley

On May 19, 2011, at 8:12 AM, Mark Greene wrote:

> I have a primary and secondary node configured with the failover transport.
> When the primary node dies, producers and consumers switch over to the
> secondary node just fine. However, when the primary node comes back, only
> the producer seems to recover and migrate back to the primary, however the
> consumer stays connected to the secondary node. Is this expected behavior?
> I'm using v5.3.2.
> 
> Thanks in advance.
> 
> -Mark