Uploaded image for project: 'Debezium'
  1. Debezium
  2. DBZ-8978

Reduce frequency of data dictionary build steps for redo_log_catalog strategy

XMLWordPrintable

    • Icon: Enhancement Enhancement
    • Resolution: Unresolved
    • Icon: Major Major
    • 3.4.0.Alpha1
    • None
    • oracle-connector
    • None
    • False
    • Hide

      None

      Show
      None
    • False

      In all LogMiner implementations, we know the current list of logs and what of those will be mined on the next iteration. We should look into whether we can inspect the current log list that include or come after the low watermark SCN and if those logs have the dictionary begin/end markers set to Yes. If both markers exist, that means the entire data dictionary is present, and we could theoretically skip the next build step.

      This has several advantages, which include:

      • Less log switches generated by the redo_log_catalog strategy
      • Better stability with the redo_log_catalog strategy with multi-connector deployments
      • Faster mining session start-ups reading fewer dictionary extracts

              Unassigned Unassigned
              ccranfor@redhat.com Chris Cranford
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

                Created:
                Updated: