Uploaded image for project: 'mod_cluster'
  1. mod_cluster
  2. MODCLUSTER-97

httpd should remove workers who crashed

    Details

    • Type: Feature Request
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Done
    • Affects Version/s: 1.0.1.GA
    • Fix Version/s: 1.1.0.Beta1
    • Component/s: None
    • Labels:
      None

      Description

      When worker W crashes, httpd will not remove it from its worker tables and still try to send requests to W.

      We should implement something like ping_mode=I (interval), ping_timeout and connection_ping_interval in mod_jk [1].

      It should work as follows:

      • W crashes
      • httpd waits for connection_ping_interval of idle time, then starts checking the connection to W every ping_timeout seconds (or ms)
      • When the connection to W is broken, httpd should remove W from its worker table and stop the connection checker
      • When (if !) W is started again, mod-cluster will register it again with httpd

      The configuration above should be done entirely on the AS side, e.g. in HAModClusterConfig. maybe we can find just 1 attribute rather than 3.

      [1] http://tomcat.apache.org/connectors-doc/reference/workers.html

        Gliffy Diagrams

          Attachments

            Activity

              People

              • Assignee:
                jfclere Jean-Frederic Clere
                Reporter:
                belaban Bela Ban
              • Votes:
                0 Vote for this issue
                Watchers:
                1 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: