Fuse 6.1 + ActiveMQ Functionality

[schellinger]

We are trying to decide whether to use the ActiveMQ that is embedded in Fuse 6.1 or download, install, and use the standalone version of ActiveMQ.  We have heard there are differences between the embedded and standalone ActiveMQ versions.  Possibly including its scalability?  Is this true and/or are there other major differences between the 2 versions?  Is there somewhere I can see any differences between the versions?  Thanks.

Shawn

[davsclaus]

JBoss A-MQ comes in a standalone and embedded mode (the latter is embedded in karaf/fabric).

So the standalone mode is just like the Apache ActiveMQ (except it comes with hawtio as the web console out of the box), and it is not integrated with fabric, or cloud etc.

The embedded mode integrates with fabric, and allows you to use fabric to manage and provision the brokers, and make it easier to setup broker clusters and network groups and whatnot, as you do not need to manually configure ip addresses, port numbers, and so forth.

But the code in AMQ for the standalone and embedded mode is the same.

[schellinger]

Thanks Claus - exactly what I needed to see.  Thanks.

Shawn

[schellinger]

Claus,

As a follow up from my team, we wanted to know if there are any scalability or high-availability concerns if we use the embedded AM-Q inside Fuse vs AM-Q being standalone in a production environment.  Thanks.

Shawn

[davsclaus]

No there is no scalability and another other limitations between the two of them. Its the same broker code.

The embedded just runs inside a Karaf container which comes with a few additonal services such as SSH / OSGi runtime / container. So you can host other apps in the broker. But people tend to leave the broker as a broker. But some many want a little Camel app to ingest data into the broker.

Also as its the Karaf container, you can patch the broker without having to restart it.

But they should both handle as many messages as the other etc.