bootstrapping/deploying metrics db

John Sanda jsanda at redhat.com
Wed Sep 26 15:13:52 UTC 2012


I think we need to settle on some terminology. I'm going to need Jay's help :) For embedded I have been referring primarily to a cassandra node running in the same JVM as the rhq server. In some conversations, embedded has also referred to running in a separate process but on the same machine.

We can absolutely support the person who just wants to try out RHQ or who is running a small shop. In fact, we can even run a multi-node cluster for those use cases and make the installation and set up completely seamless. This is exactly what I am working on for our dev environment. I hope to have something to demo by the end of the week or early next week at the latest.

On Sep 26, 2012, at 11:02 AM, John Mazzitelli <mazz at redhat.com> wrote:

> We could provide documentation to folks on how to add new standalone cassandara nodes if they so choose. I'm more worried about the guy who just wants to install RHQ and try it out - or the small shop that just wants a simple RHQ environment to manage his half-dozen or so machines/app servers whatever. In that case, an embedded solution would work nice, assuming cassandra doesn't gobble up all resources that the RHQ Server needs.
> 
> When you say "embedded", I assume you mean inside the same VM as the RHQ Server? Perhaps we don't have it in the same VM? Perhaps just have rhq-server.sh start two things - the RHQ Server itself and its "embedded" cassandra node (where "embedded" just means its running on the same machine as RHQ Server and its started/stopped by the same script as the RHQ Server (i.e. rhq-server.sh)
> 



More information about the rhq-devel mailing list