Uploaded image for project: 'WiredTiger'
  1. WiredTiger
  2. WT-7534

Investigate FTDC drops during checkpoint prepare

XMLWordPrintableJSON

    • Type: Icon: Improvement Improvement
    • Resolution: Done
    • Priority: Icon: Major - P3 Major - P3
    • None
    • Affects Version/s: None
    • Component/s: None
    • None
    • 5
    • Storage - Ra 2021-08-09

      I noticed that FTDC keeps dropping (along with performance) when a checkpoint starts. This happens mostly when prepare stage of the checkpoint starts.

      Attached FTDC from a run, and here is a screenshot:

      Notice that even when there is no workload running, FTDC drops out. It will be useful to investigate why, and possibly fix FTDC drop, so that we can continue to get signal on what is going on with WiredTiger/Workload during the prepare stage.

        1. table locks.png
          200 kB
          Etienne Petrel
        2. no clear no scale.png
          66 kB
          Etienne Petrel
        3. mongodb_many_coll_sleep_30s.log
          30.24 MB
          Etienne Petrel
        4. image-2021-08-03-12-58-12-205.png
          174 kB
          Etienne Petrel
        5. image-2021-07-30-17-11-31-662.png
          200 kB
          Etienne Petrel
        6. image-2021-07-28-16-52-56-300.png
          164 kB
          Etienne Petrel
        7. image-2021-07-28-16-52-27-046.png
          173 kB
          Etienne Petrel
        8. image-2021-07-28-13-31-58-846.png
          190 kB
          Etienne Petrel
        9. image-2021-07-28-13-30-02-899.png
          180 kB
          Etienne Petrel
        10. image-2021-07-27-17-27-16-727.png
          62 kB
          Etienne Petrel
        11. ftdc-drops.png
          168 kB
          Sulabh Mahajan

            Assignee:
            etienne.petrel@mongodb.com Etienne Petrel
            Reporter:
            sulabh.mahajan@mongodb.com Sulabh Mahajan
            Votes:
            0 Vote for this issue
            Watchers:
            7 Start watching this issue

              Created:
              Updated:
              Resolved: