[HDFS-15210] EC : File write hanged when DN is shutdown by admin command.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.1.1
• Fix Version/s: 3.3.1, 3.4.0, 3.2.3
• Component/s: ec
• Assignee: Surendra Singh Lilhore
• Reporter: Surendra Singh Lilhore
• Created: 06/Mar/20 12:40
• Updated: 10/Jun/21 07:44
• Resolved: 29/Apr/20 06:40

Description

EC Blocks : blk_-9223372036854291632_10668910, blk_-9223372036854291631_10668910, blk_-9223372036854291630_10668910, blk_-9223372036854291629_10668910, blk_-9223372036854291628_10668910

 

Two block DN restarted : blk_-9223372036854291630_10668910 & blk_-9223372036854291632_10668910

2020-03-03 18:12:17,074 DEBUG hdfs.DataStreamer: DFSClient seqno: -2 reply: OOB_RESTART downstreamAckTimeNanos: 0 flag: 8
2020-03-03 18:13:39,469 DEBUG hdfs.DataStreamer: DFSClient seqno: -2 reply: OOB_RESTART downstreamAckTimeNanos: 0 flag: 8 

 

Restarted streams are stuck in below stacktrace :

java.util.concurrent.LinkedBlockingQueue.take(LinkedBlockingQueue.java:442) at org.apache.hadoop.hdfs.DFSStripedOutputStream$MultipleBlockingQueue.take(DFSStripedOutputStream.java:110) at org.apache.hadoop.hdfs.StripedDataStreamer.setupPipelineInternal(StripedDataStreamer.java:140) at org.apache.hadoop.hdfs.DataStreamer.setupPipelineForAppendOrRecovery(DataStreamer.java:1540) at org.apache.hadoop.hdfs.DataStreamer.processDatanodeOrExternalError(DataStreamer.java:1276) at org.apache.hadoop.hdfs.DataStreamer.run(DataStreamer.java:669) at org.apache.hadoop.hdfs.StripedDataStreamer.run(StripedDataStreamer.java:46)