Discussions

News: J2EE Clustering with JBoss: Singleton and Scheduler Services

  1. A clustered singleton is a service that is deployed on multiple nodes in a cluster, but is running on only one of the nodes. The node running the singleton service is typically called master node. When the master fails, another master is selected from the remaining nodes and the service is restarted on the new master.

    The JBoss clustered scheduler ensures that scheduling code written for a single VM will continue to work predictably in a cluster.

    How much value do you think these services add to JBoss?
    Have you had a need for them?

    Read: J2EE Clustering with JBoss: Singleton and Scheduler Services

    Read: ONJava Article on JBoss 3 Clustering
  2. One of the simplest, most efficient ways to execute a task in a server farm is to have a single scheduled task identify work to be performed, break the work into discrete tasks, and put a message for each task into a queue that all of the servers are listening to.

    The singleton scheduler service makes this architecture very easy to implement, and I will be using it shortly to transform a single-server application into a clustered application with load-balancing and failover.
  3. Clustering At Other Tiers?[ Go to top ]

    The setup described in the article just covers EJB clustering, right? That's just one piece in the puzzle.

    I'm curious about how people out there have set up front end (web) clustering on their sites e.g. with Apache and hardware/software node failover?

    And what about database clustering (either with shared storage or via replication?)

    Cheers, Peter
    RimuHosting - Java Hosting on VPS
  4. Clustering At Other Tiers?[ Go to top ]

    Although the JBoss clustering services are usually used within a J2EE application, they are not directly related to EJB.
    You can choose to use JBoss without any of its J2EE services, but with the clustering services.

    JBoss supports a clustered HTTP session for both Jetty and Tomcat.
    Although rarely a great choice, it is possible to maintain replicated HTTP Session. The JBoss Clustering Book has all the details.

    As far as database caching is concerned, the JBoss group is working on a distributed cache for its 4.0 release. You can download the first developer drop from the JBoss site.