[HDFS-14029] Sleep in TestLazyPersistFiles should be put into a loop

• Type: Improvement
• Status: Resolved
• Priority: Trivial
• Resolution: Fixed
• Fix Version/s: 3.0.4, 3.1.2, 3.3.0, 3.2.1
• Component/s: hdfs
• Labels: newbie
• Hadoop Flags: Reviewed
• Assignee: Adam Antal Assign to me
• Reporter: Adam Antal
• Created: 25/Oct/18 15:34
• Updated: 26/Oct/18 08:38
• Resolved: 25/Oct/18 22:17

Description

TestLazyPersistFiles#testFileShouldNotDiscardedIfNNRestarted test should be improved.

The test sleeps for 6000 at once, it could at least sleep in a loop checking for the corrupt block to be reported.

    cluster.shutdownDataNodes();

    cluster.restartNameNodes();

    // wait for the redundancy monitor to mark the file as corrupt.
    Thread.sleep(2 * DFS_NAMENODE_REDUNDANCY_INTERVAL_SECONDS_DEFAULT * 1000);

    Long corruptBlkCount = (long) Iterators.size(cluster.getNameNode()
        .getNamesystem().getBlockManager().getCorruptReplicaBlockIterator());

Thanks Kitti Nanasi for the suggestion.