Uploaded image for project: 'Data Management'
  1. Data Management
  2. DM-21874

Create PipelineTask version of backward compatible ImageDifferenceTask

    Details

    • Story Points:
      0
    • Sprint:
      AP F19-6 (November), AP S20-1 (December), AP S20-2 (January), AP S20-3 (February), AP S20-4 (March), AP S20-5 (April)
    • Team:
      Alert Production

      Description

      Create one or more PipelineTasks that allow us to run any image-differencing functionality we need to continue to support in Gen3.  My understanding is that this means we need all calexp-template differencing functionality, but not necessarily pairwise calexp-calexp differencing (I'm hoping John Swinbank and/or Eric Bellm can confirm).

      I'm creating this now and assigning it to the AP team for middleware planning purposes, with a wild guess at SPs.  We'll of course have some DRP team members helping out with the work (probably Yusra AlSayyad and/or Nate Lust).  People doing the work are welcome to turn this into an umbrella ticket in the future with no SPs or branches of its own, but please link blocker tickets to this one so the middleware team can easily track progress towards deprecating Gen2.

        Attachments

          Issue Links

            Activity

            Hide
            swinbank John Swinbank added a comment -

            All children complete; marking this done as of 2020-04-29.

            Show
            swinbank John Swinbank added a comment - All children complete; marking this done as of 2020-04-29.
            Hide
            swinbank John Swinbank added a comment -

            Hey Gabor Kovacs

            Looking at this ticket, I think it makes more sense to allocate the story points to each of the child tickets — where the work is actually being done — rather than collecting them all on the parent. That means we can use those tickets to meaningfully track progress towards the goal.

            Given that, I have set the SPs on this ticket to 0, and allocated points to each of its children. You may wish to check whether those point totals seem reasonable to you.

            Show
            swinbank John Swinbank added a comment - Hey Gabor Kovacs — Looking at this ticket, I think it makes more sense to allocate the story points to each of the child tickets — where the work is actually being done — rather than collecting them all on the parent. That means we can use those tickets to meaningfully track progress towards the goal. Given that, I have set the SPs on this ticket to 0, and allocated points to each of its children. You may wish to check whether those point totals seem reasonable to you.
            Hide
            swinbank John Swinbank added a comment -

            This is a follow-up to work that's already being done on DM-21302. We expect Gabor Kovacs to take the lead on this effort.

            As Jim suggests, we've already agreed this won't include pairwise calexp-calexp differencing in this version. We've also encouraged Gabor Kovacs to think a little more broadly about the design of the code, so (depending on the results of DM-21302 and subsequent discussion), this may be more than a direct port of the existing codebase to new middleware.

            Show
            swinbank John Swinbank added a comment - This is a follow-up to work that's already being done on DM-21302 . We expect Gabor Kovacs to take the lead on this effort. As Jim suggests, we've already agreed this won't include pairwise calexp-calexp differencing in this version. We've also encouraged Gabor Kovacs to think a little more broadly about the design of the code, so (depending on the results of DM-21302 and subsequent discussion), this may be more than a direct port of the existing codebase to new middleware.
            Hide
            ebellm Eric Bellm added a comment -

            Confirming we can omit pairwise calexp-calexp differencing at least in the immediate term.

            Show
            ebellm Eric Bellm added a comment - Confirming we can omit pairwise calexp-calexp differencing at least in the immediate term.

              People

              • Assignee:
                gkovacs Gabor Kovacs
                Reporter:
                jbosch Jim Bosch
                Watchers:
                Jim Bosch, John Swinbank
              • Votes:
                0 Vote for this issue
                Watchers:
                2 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: