[jboss-user] [JBoss Messaging] - Re: JMS HA/Clustering without shared DB

timfox do-not-reply at jboss.com
Tue Sep 23 07:34:29 EDT 2008

"testrot" wrote : Nobody?
  | Is this an uncommon request, or did I miss something when searching for a HA solution without without a clustered DB?
  | Greetings,
  | Martin
  | P.S.: Even a negative response like "there is not possibility and this is not planned" would be appreciated.

A polite word of advice. Don't be too pushy in demanding free advice in a forum in which you're new.

It's only likely to get people's backs up and make it more unlikely they'll actually want to give you the advice you want.

Remember, we offer advice in this forum on a "best effort" basis. There are no guarantees how, when or if your post will get answered.

If you want a guaranteed response within a set time, buy a support contract.

Most forum answering is done by team members in their own time, at their own expense, so be nice when asking a stranger directions.

Now, you're lucky I'm a nice person. Many others here would have flamed you immediately.

The question you asked has been discussed many times on this forum, and in JIRA tasks, and in the user guides.

Since you are clearly such a busy person, and have not time for your own searching here's the short answer:

JBM 1.x uses a shared database for persistence. If that database itself is not clustered, yes it's a single point of failure. That's why we recommend running JBM 1.4 against a clustered database such as Mysql Cluster or Oracle RAC.

JBM 2.0 does not use a database for persistence. it uses a high performance journal and supports shared nothing replication to eliminate any SPOF.

Hope that answers your question. :)

View the original post : http://www.jboss.com/index.html?module=bb&op=viewtopic&p=4178266#4178266

Reply to the post : http://www.jboss.com/index.html?module=bb&op=posting&mode=reply&p=4178266

More information about the jboss-user mailing list