1 Reply Latest reply on Dec 12, 2005 3:04 PM by vegecat

    Clustering for SFSB in case of network failure..

    ehskaga

      I have a cluster having Stateful session bean (SFSB) deployed. While trying to emulate a failover, i simulated a failure by taking out the network chord (which for me is a case of network failure) of one node which was servicing a request. The default behaviour is that of client waiting endlessly for something to happen whereas the server has gone down.
      How to resolve this issue, would it be resolved by specifying the time out (if yes then where is that specified), else what needs to be done.

        • 1. Re: Clustering for SFSB in case of network failure..
          vegecat

          I experienced similar problem as I posted in another thread. In my case, my client did not wait endlessly. It waited for about 1 minute and showed socket time out exception and then continued to wait for about 5 minutes and started to work.

          Did you find any solution to this? I also tried to solve this by configuring the clientsocket time out configuration. It did not work. I am not sure the problem is on the client smart proxy or on the JBoss server.

          "ehskaga" wrote:
          I have a cluster having Stateful session bean (SFSB) deployed. While trying to emulate a failover, i simulated a failure by taking out the network chord (which for me is a case of network failure) of one node which was servicing a request. The default behaviour is that of client waiting endlessly for something to happen whereas the server has gone down.
          How to resolve this issue, would it be resolved by specifying the time out (if yes then where is that specified), else what needs to be done.