Michael Stack created HBASE-25349: ------------------------------------- Summary: [fl Key: HBASE-25349 URL: https://issues.apache.org/jira/browse/HBASE-25349 Project: HBase Issue Type: Bug Reporter: Michael Stack
We look for old replication source to be non-zero long after it has been processed. The test gets stuck waiting. See here is where the table becomes available, after which we start expecting the old source replication queue to be non-zero. {{ 2020-12-01 20:02:52,400 INFO [Listener at localhost/56996] regionserver.TestRefreshRecoveredReplication(136): Available testReplicationRefreshSource}} But see here where the replication source has been removed before we get to the 'available' line above: 2020-12-01 20:02:50,768 INFO [ReplicationExecutor-0.replicationSource,2-kalashnikov.attlocal.net,56950,1606881738045.replicationSource.shipperkalashnikov.attlocal.net%2C56950%2C1606881738045,2-kalashnikov.attlocal.net,56950,1606881738045] regionserver.ReplicationSourceManager(463): Done with the recovered queue 2-kalashnikov. attlocal.net,56950,1606881738045 -- This message was sent by Atlassian Jira (v8.3.4#803005)