Uploaded image for project: 'XNAT'
  1. XNAT
  2. XNAT-4503

Event Handlers' list of WorkflowStatusEvents is unmanageable

    Details

    • Type: Bug
    • Status: Open
    • Priority: Blocker
    • Resolution: Unresolved
    • Affects Version/s: None
    • Fix Version/s: 1.7.1
    • Component/s: None
    • Labels:
    • Environment:

      xnat-dev05

    • Sprint:
      Sprint 1, Sprint 4, 1.7 Release Sprint
    • Rank:
      0|0hzzkh:zzzzzzzzzz
    • Sprint:
      Sprint 1, Sprint 4, 1.7 Release Sprint

      Description

      To recreate:

      1. Go to Administer > Automation
      2. Click Add Event Handler
      3. Select Event Type > Workflow Status Event.
      4. Expand the Event ID dropdown; it's filled with way too many workflows like "Added user r8...", "Invited user to project(xnatselenium@gmail.com)".

      This is basically the same as XNAT-3747.

        Issue Links

          Activity

          Hide
          hodgem@wustl.edu Mike Hodge added a comment -

          Is cnda-dev-daily built from a current 1.7 war? This commit should be limiting that list:

          https://bitbucket.org/xnatdev/xnat-web/commits/5bbc46889237503d177d80351540cb9ff178ab05

          See the attached screenshot from the latest build of my IntraDB DEV machine, which also has tons of workflows:

          Show
          hodgem@wustl.edu Mike Hodge added a comment - Is cnda-dev-daily built from a current 1.7 war? This commit should be limiting that list: https://bitbucket.org/xnatdev/xnat-web/commits/5bbc46889237503d177d80351540cb9ff178ab05 See the attached screenshot from the latest build of my IntraDB DEV machine, which also has tons of workflows:
          Hide
          moore.c@wustl.edu Charlie Moore added a comment -

          cnda-dev-daily was from the original 1.6.5 bug. I'm seeing this on xnat-dev05, which does have current 1.7 code.

          Show
          moore.c@wustl.edu Charlie Moore added a comment - cnda-dev-daily was from the original 1.6.5 bug. I'm seeing this on xnat-dev05, which does have current 1.7 code.
          Hide
          hodgem@wustl.edu Mike Hodge added a comment -

          I logged into xnat-dev05. That list should clean itself up over time. It shouldn't get indefinitely long. At the project level, I have it keeping track of the number of occurrences of workflow event IDs. It should only display the top 20, so as the site is fully utilized, workflows like the addition of specific users will work their way out. At the site level, the list is comprised of a unique set of all the top 20 from the projects, so the list can get longer than 20, but it will work its way down as the site is fully utilized. The one-off ones that include user names should work their way out as the site is fully utilized. That's what should be happening. At first look, it doesn't look like xnat-dev05 is utilized heavily enough (lots of projects, with relatively little activity on each) to cause the operations that include individual users to drop off. I thought about, but decided against hard-coded rules that would automatically try to lop off user names and such things. Looking at the workflows in HCP, there should be enough relatively standard ones that these shouldn't persist over time. I'd be happy to discuss further, if you see behavior different than I'm describing.

          Show
          hodgem@wustl.edu Mike Hodge added a comment - I logged into xnat-dev05. That list should clean itself up over time. It shouldn't get indefinitely long. At the project level, I have it keeping track of the number of occurrences of workflow event IDs. It should only display the top 20, so as the site is fully utilized, workflows like the addition of specific users will work their way out. At the site level, the list is comprised of a unique set of all the top 20 from the projects, so the list can get longer than 20, but it will work its way down as the site is fully utilized. The one-off ones that include user names should work their way out as the site is fully utilized. That's what should be happening. At first look, it doesn't look like xnat-dev05 is utilized heavily enough (lots of projects, with relatively little activity on each) to cause the operations that include individual users to drop off. I thought about, but decided against hard-coded rules that would automatically try to lop off user names and such things. Looking at the workflows in HCP, there should be enough relatively standard ones that these shouldn't persist over time. I'd be happy to discuss further, if you see behavior different than I'm describing.

            People

            • Assignee:
              jrherrick@wustl.edu Rick Herrick
              Reporter:
              moore.c@wustl.edu Charlie Moore
            • Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

              • Created:
                Updated:

                Agile