Uploaded image for project: 'Debezium'
  1. Debezium
  2. DBZ-349

Better support for large append-only tables by making the snapshotting process restartable

    XMLWordPrintable

    Details

    • Type: Feature Request
    • Status: Closed (View Workflow)
    • Priority: Major
    • Resolution: Done
    • Affects Version/s: 0.5.2
    • Fix Version/s: 0.7.0
    • Component/s: mysql-connector
    • Labels:
      None

      Description

      Currently, the snapshotting process is non-interruptible to ensure that snapshotting results in creating a consistent dump of the existing data in the tables before the CDC phase begins.

      This is appropriate in the general case, but for very large log-like tables, which are append-only, snapshotting can take a long time (several days in our case for some larger tables), and if this phase is interrupted, it has to be restarted from scratch.

      I propose to introduce the option to override the select statement issued during snapshotting. This way it is possible to resume a snapshot that has progressed to a specific point (e.g. by adding a custom WHERE clause with a condition on the record ID).

      I have implemented a patch for this feature, for which I'd like to create a pull request.

        Gliffy Diagrams

          Attachments

            Issue Links

              Activity

                People

                • Assignee:
                  Unassigned
                  Reporter:
                  ainagy Andras Istvan Nagy
                • Votes:
                  0 Vote for this issue
                  Watchers:
                  2 Start watching this issue

                  Dates

                  • Created:
                    Updated:
                    Resolved: