I have two nodes in a cluster; I am allowing the users to have node specific configurations like logging level, local cache settings etc; at a time, it has become really difficult to manage these settings because the user has to know or remember the configuration applied on a particular node - move node after node before finding that particular node; is there any standard or known way to manage these nodes from a single place? like, from the httpd server itself or have one node as master and remember other nodes?
0
votes
Which version of JBoss? JBoss clustering is very much changed in 7.1.x and has the concept of domain and standalone clustering. If you could provide JBoss version, application characteristics (queues, session sharing, etc.) and a brief topology, I can suggest a couple of options.
- Akber Choudhry
@akberc, JBoss 5 and we are scaling out to balance the load.
- Venkatesh Laguduva
1 Answers
0
votes
Well, there is no standard-way of doing that, but one pretty good working solution is to save the configuration in SVN, do a checkout and every time you do a modification, you can see the modifications by svn diff or svn st. We use this solution to roll out our JBoss-servers and it's working fine for us.
