5 Replies Latest reply on Dec 17, 2010 10:20 AM by Manik Surtani

    A question on replication

    Gerry Smith Newbie

      I have Infinispan 4.2 working on AS 4.2.3 (default config).

      Our app is pretty intense and runs up to 150 times faster with infinispan, so we're excited about putting infinispan into production very soon.

      I made an MBean which starts up Infinispan on server startup and initiates population of the cache - we need to maintain a representation of the full data set in the cache (no eviction), and the cache survives redeployment of the app.

      I imagine we would have compatibility problems with the 'all' AS config, so for now we will just conigure HA with AJP, and replicate the cache across the cluster.

      As soon as possible we'll migrate to AS 6, but we will have to do some refactoring, particularly because RUN-AS seems to have been broken since AS 5 for all but the simplest of apps.


      My question is, when the cache manager is started on server startup, is there a way for my MBean to determine whether the cache will populate from the cluster, or if it's on it's own and must obtain it from the rdbms ?