8.11. Known Issues for Flume

  • BUG-16771: Hive table returns multiple copies of streaming data when testing the Hive Server restart scenario.

    Problem: When running the Hive restart test where the Hive metastore is bounced while Flume is treaming data to Hive, 3 duplicate copies were observed for each row in the Hive table. (Expected: 200 rows; observed: 800 rows, or 3 complete copies of the expected set of 200.)