Uploaded image for project: 'Cassandra'
  1. Cassandra
  2. CASSANDRA-2419

Risk of counter over-count when recovering commit log

    Details

    • Type: Bug
    • Status: Resolved
    • Priority: Major
    • Resolution: Fixed
    • Fix Version/s: 0.8.0
    • Component/s: None
    • Labels:

      Description

      When a memtable was flush, there is a small delay before the commit log replay position gets updated. If the node fails during this delay, all the updates of this memtable will be replay during commit log recovery and will end-up being over-counts.

        Attachments

        1. 0001-Record-and-use-sstable-replay-position.patch
          32 kB
          Sylvain Lebresne
        2. 0001-Record-CL-replay-infos-alongside-sstables-v2.patch
          63 kB
          Sylvain Lebresne
        3. 2419-v3.txt
          61 kB
          Jonathan Ellis
        4. 2419-v4.txt
          62 kB
          Jonathan Ellis
        5. 2419-v6.txt
          62 kB
          Jonathan Ellis

          Activity

            People

            • Assignee:
              slebresne Sylvain Lebresne
              Reporter:
              slebresne Sylvain Lebresne
              Reviewer:
              Jonathan Ellis
            • Votes:
              1 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved:

                Time Tracking

                Estimated:
                Original Estimate - 8h
                8h
                Remaining:
                Remaining Estimate - 8h
                8h
                Logged:
                Time Spent - Not Specified
                Not Specified