Uploaded image for project: 'Debezium'
  1. Debezium
  2. DBZ-3347

Schema change SourceRecords have null partition

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done
    • Icon: Major Major
    • 1.5.0.Final
    • 1.5.0.Beta1
    • core-library
    • None
    • False
    • False
    • Undefined
    • Hide

      Start a connector that does an initial snapshot that includes the schema, then look at the kafka topic for connector offsets.

      Show
      Start a connector that does an initial snapshot that includes the schema, then look at the kafka topic for connector offsets.

      I noticed that the "my_connect_offsets" topic has two entries for the Debezium Oracle connector

      ["oracle-logminer-connector",null]    {"snapshot":true,"scn":"7898027","snapshot_completed":false}
      ["oracle-logminer-connector",{"server":"oracle-test"}]    {"commit_scn":null,"transaction_id":null,"scn":"7903065"}
      

      where I would have expected only one.

      Scanning through the code, it looks like the one with the null partition is emitted in EventDispatcher.java:schemaChangeEvent().

      Why is it correct to not use the usual partition there? I don't see a place where this offset is read?

              ccranfor@redhat.com Chris Cranford
              christiankamm Christian Kamm (Inactive)
              Votes:
              0 Vote for this issue
              Watchers:
              4 Start watching this issue

                Created:
                Updated:
                Resolved: