Uploaded image for project: 'JikesRVM'
  1. JikesRVM
  2. RVM-307

Serve cattrack using a cluster of mongrel instances

    Details

    • Type: Improvement
    • Status: Closed
    • Priority: Medium
    • Resolution: Fixed
    • Affects Version/s: None
    • Fix Version/s: None
    • Labels:
      None

      Description

      Currently we are serving cattrack using a single instance of mongrel. This serializes all access to the service which up until the current time has not been an issue as the server is not under a large load and none of the requests take too long to serve. As of r13810 the ability to query the output of test_case_executions was added and serving this request can take a long time. While that request is being served no other requests can be served which means the server can appear to hang.

      We should serve the website with a cluster of mongrel instances so the website does not appear to be hung while running the queries. It should be a relatively easy configuration change (see [1] & [2]).

      [1] http://mongrel.rubyforge.org/wiki/Apache
      [2] http://mongrel.rubyforge.org/wiki/MongrelCluster

        Attachments

          Issue Links

            Activity

            dgrove David Grove created issue -
            Hide
            dgrove David Grove added a comment -

            I've just finished doing this for the cattrack instance I'm running internally in watson for use by X10.

            It was fairly easy, so at some point in the next month or two I'd like to try it for jikes rvm. As part of that, I'd also like to change
            the deploy story over to using capistrano. I've got the basic setup for that in the branch I'm using for x10 development.

            Show
            dgrove David Grove added a comment - I've just finished doing this for the cattrack instance I'm running internally in watson for use by X10. It was fairly easy, so at some point in the next month or two I'd like to try it for jikes rvm. As part of that, I'd also like to change the deploy story over to using capistrano. I've got the basic setup for that in the branch I'm using for x10 development.
            Hide
            dgrove David Grove added a comment -

            Daniel and I did this yesterday; up and running with a cluster of 5 mongrels behind a proxy load balancer.

            Show
            dgrove David Grove added a comment - Daniel and I did this yesterday; up and running with a cluster of 5 mongrels behind a proxy load balancer.
            dgrove David Grove made changes -
            Field Original Value New Value
            Workflow jira [ 17689 ] X10 Workflow [ 18435 ]
            dgrove David Grove made changes -
            Link This issue is depended upon by RVM-687 [ RVM-687 ]
            dgrove David Grove made changes -
            Priority Major [ 6 ] Medium [ 3 ]

              People

              • Assignee:
                dgrove David Grove
                Reporter:
                pdonald Peter Donald
              • Votes:
                0 Vote for this issue
                Watchers:
                0 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: