[HDFS-5579] Under construction files make DataNode decommission take very long hours

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.2.0, 2.2.0
• Fix Version/s: 2.3.0
• Component/s: namenode
• Hadoop Flags: Reviewed
• Assignee: yunjiong zhao
• Reporter: yunjiong zhao
• Created: 28/Nov/13 06:18
• Updated: 01/Apr/15 07:32
• Resolved: 14/Jan/14 02:24

Description

We noticed that some times decommission DataNodes takes very long time, even exceeds 100 hours.
After check the code, I found that in BlockManager:computeReplicationWorkForBlocks(List<List<Block>> blocksToReplicate) it won't replicate blocks which belongs to under construction files, however in BlockManager:isReplicationInProgress(DatanodeDescriptor srcNode), if there is block need replicate no matter whether it belongs to under construction or not, the decommission progress will continue running.
That's the reason some time the decommission takes very long time.

Issue Links

• is duplicated by: HDFS-7306 can't decommission w/under construction blocks