[HDFS-13359] DataXceiver hung due to the lock in FsDatasetImpl#getBlockInputStream

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.7.1
• Fix Version/s: 3.3.0, 3.2.1, 3.1.3
• Component/s: datanode
• Target Version/s: 3.3.0
• Assignee: Yiqun Lin
• Reporter: Yiqun Lin
• Created: 28/Mar/18 08:36
• Updated: 05/Nov/21 02:03
• Resolved: 10/Aug/19 01:42

Description

DataXceiver hung due to the lock that locked by
FsDatasetImpl#getBlockInputStream (have attached stack).

  @Override // FsDatasetSpi
  public InputStream getBlockInputStream(ExtendedBlock b,
      long seekOffset) throws IOException {

    ReplicaInfo info;
    synchronized(this) {
      info = volumeMap.get(b.getBlockPoolId(), b.getLocalBlock());
    }
    ...
  }

The lock synchronized(this) used here is expensive, there is already one AutoCloseableLock type lock defined for ReplicaMap. We can use it instead.

Issue Links

• relates to: HDFS-10682 Replace FsDatasetImpl object lock with a separate lock object