Uploaded image for project: 'Request For Comments'
  1. Request For Comments
  2. RFC-22

remove old DB ingest capabilities prior to completion of new ones

    XMLWordPrintable

Details

    • RFC
    • Status: Implemented
    • Resolution: Done
    • DM
    • None
    • this issue page

    Description

      We're approaching the end of the conversion to the new measurement framework, and the last feature that needs to be implemented is database ingest. We're already working on that, but we have other work that's being blocked by the fact that we can't remove the old measurement framework until it's done. So I'm proposing here that we go ahead and remove the old measurement framework (and its associated old ingest scripts) now, resulting in a brief period in which we wouldn't have any ingest scripts at all.

      My understanding is that currently no one is using the old ingest scripts, unless they're doing so through an older tagged release, which wouldn't be affected, but I'd like to have that confirmed here, especially by the jbecla and boutigny.

      If this RFC is approved, we'd then proceed immediately with removing the old measurement framework in meas_algorithms, and removing both ap and datarel from the stack entirely, as currently I don't believe they provide any functionality besides ingest. At this point, I'm just proposing we remove them from buildbot and that the repos will be migrated to wherever completely dead repos go; the discussion on where exactly that should be should happen elsewhere.

      Attachments

        Issue Links

          Activity

            jbosch Jim Bosch added a comment -

            I fixed the ticket link above to avoid future confusion.

            The sense in which I think ingest is alive is just that you can still switch back to the old measurement framework via config, and that should still produce outputs compatible with the old ingest. I don't think anyone has actually done that, though.

            jbosch Jim Bosch added a comment - I fixed the ticket link above to avoid future confusion. The sense in which I think ingest is alive is just that you can still switch back to the old measurement framework via config, and that should still produce outputs compatible with the old ingest. I don't think anyone has actually done that, though.
            jbosch Jim Bosch added a comment -

            ktl my sense is that you're still slightly uncomfortable with this, and at least at this stage, it's not looking as urgent as we'd originally thought (in that we still have some more work to before we can remove the old measurement framework anyway, so the work on ingest may catch up), so my inclination right now is to just Withdraw this RFC. If I've misread that, let me know.

            jbosch Jim Bosch added a comment - ktl my sense is that you're still slightly uncomfortable with this, and at least at this stage, it's not looking as urgent as we'd originally thought (in that we still have some more work to before we can remove the old measurement framework anyway, so the work on ingest may catch up), so my inclination right now is to just Withdraw this RFC. If I've misread that, let me know.

            No, I'm more uncomfortable that the ingest updates haven't been progressing along with the schema revisions, which has nothing to do with this RFC, really. But given that things are all up in the air right now (the prior-to-ingest transformation mechanisms being under review as well), if it makes your life easier to go ahead with this, I'm OK with it.

            ktl Kian-Tat Lim added a comment - No, I'm more uncomfortable that the ingest updates haven't been progressing along with the schema revisions, which has nothing to do with this RFC, really. But given that things are all up in the air right now (the prior-to-ingest transformation mechanisms being under review as well), if it makes your life easier to go ahead with this, I'm OK with it.
            jbosch Jim Bosch added a comment -

            It may make our lives slightly easier down the road to have this flexibility, even if we don't need it just yet, and K-T has indicated that he's not opposed to it, so I'm accepting this.

            jbosch Jim Bosch added a comment - It may make our lives slightly easier down the road to have this flexibility, even if we don't need it just yet, and K-T has indicated that he's not opposed to it, so I'm accepting this.
            jbosch Jim Bosch added a comment -

            Implemented on DM-2928.

            jbosch Jim Bosch added a comment - Implemented on DM-2928 .

            People

              jbosch Jim Bosch
              jbosch Jim Bosch
              Dominique Boutigny, Jim Bosch, John Swinbank, Kian-Tat Lim, Serge Monkewitz
              Votes:
              0 Vote for this issue
              Watchers:
              5 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:
                Planned End:

                Jenkins

                  No builds found.