Uploaded image for project: 'Debezium'
  1. Debezium
  2. DBZ-2825

Archive Log mining does not work with Logminer

    XMLWordPrintable

Details

    • Bug
    • Resolution: Done
    • Major
    • 1.4.0.Final
    • 1.4.0.Alpha2
    • oracle-connector
    • None
    • False
    • False
    • Undefined
    • Hide
      1. Start Debezium connector using Logminer, receive snapshot and some online updates.
      2. Stop connector.
      3. Make some updates in DB so that SCN stored in connect-offsets move from REDO to Archive logs.
      4. Start connector and observe the error demanding to take new snapshot.
      Show
      Start Debezium connector using Logminer, receive snapshot and some online updates. Stop connector. Make some updates in DB so that SCN stored in connect-offsets move from REDO to Archive logs. Start connector and observe the error demanding to take new snapshot.

    Description

      Archive Log mining seems to be non working - I'm using Alpha2 version and periodically getting errors like this:

      [2020-11-30 09:49:00,279] ERROR Mining session stopped due to the {} (io.debezium.connector.oracle.logminer.LogMinerHelper:515)
       io.debezium.DebeziumException: Online REDO LOG files do not contain the offset scn. Please perform a new snapshot.
       at io.debezium.connector.oracle.logminer.LogMinerStreamingChangeEventSource.execute(LogMinerStreamingChangeEventSource.java:137)
       at io.debezium.pipeline.ChangeEventSourceCoordinator.streamEvents(ChangeEventSourceCoordinator.java:140)
       at io.debezium.pipeline.ChangeEventSourceCoordinator.lambda$start$0(ChangeEventSourceCoordinator.java:113)
       at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
       at java.util.concurrent.FutureTask.run(FutureTask.java:266)
       at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
       at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
       at java.lang.Thread.run(Thread.java:748)
       [2020-11-30 09:49:00,279] ERROR Producer failure (io.debezium.pipeline.ErrorHandler:31)
       io.debezium.DebeziumException: Online REDO LOG files do not contain the offset scn. Please perform a new snapshot.
       at io.debezium.connector.oracle.logminer.LogMinerStreamingChangeEventSource.execute(LogMinerStreamingChangeEventSource.java:137)
       at io.debezium.pipeline.ChangeEventSourceCoordinator.streamEvents(ChangeEventSourceCoordinator.java:140)
       at io.debezium.pipeline.ChangeEventSourceCoordinator.lambda$start$0(ChangeEventSourceCoordinator.java:113)
       at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
       at java.util.concurrent.FutureTask.run(FutureTask.java:266)
       at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
       at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
       at java.lang.Thread.run(Thread.java:748)

      At the same time Oracle Flashback Query with same SCN works well:

      SELECT * FROM TEST AS OF scn 501194374

      Also I checked that relevant arch file exists at server: 

      select * from v$archived_log WHERE 501194374 BETWEEN "FIRST_CHANGE#" AND "NEXT_CHANGE#"

      Attachments

        Activity

          People

            ccranfor@redhat.com Chris Cranford
            ant0nk Anton Kondratev (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: