Search code examples
websphere

Websphere web plug-in to automatically propagate cluster node shutdown

Does the WebServer web server plug-in automatically propagate the new configuration due to a manual shutdown of a node in the application server cluster? I've been going through the documentation and it looks like the only way for the web server to act on this is by detecting the node state by itself. Is there any workaround?

Solution

  • By default, the WAS Plug-in only detects that a JVM is down by failing to send it a request or failing to establish a new TCP connection.

    If you use the "Intelligent Management for WebServers" features available in 8.5 and later, there is a control connection between the cell and the Plug-in that will proactively tell the Plugin that a server is down.

    Backing up to the non-IM case, here's what happens during an unplanned shutdown of a JVM (from http://publib.boulder.ibm.com/httpserv/ihsdiag/plugin_questions.html#failover)

    1. If an application server terminates unexpectedly, several things unfold. This is largely WebSphere edition independent.
    2. The application servers operating system closes all open sockets.
    3. WebServer threads waiting for the response in the WAS Plug-in are notified of EOF or ECONNRESET.
    4. If the error occurred on a new connection to the application server, it will be marked down in the current webserver process. This server will not be retried until a configurable interval expires (RetryInterval).
    5. If the error occurred on a an existng connection to the application server, it will not be marked down.
    6. Retryable requests that were in-flight are retried by the WAS Plug-in, as permitted.
    7. If the backend servers use memory to memory session replication (ND only), the WLM component will tell the WAS Plug-in to use a specific replacement affinity server.
    8. If the backend servers use any kind of session persistence, the failover is transparent. Session persistence is available in all websphere editions.
    9. New requests, with or without affinity, are routed to remaining servers..
    10. After the RetryInterval expires, the WAS plug-in will try to establish new connections to the server. If it remains down, failure will be relatively fast, and put the server back into the markd down state.