[HDFS-8423] Add additional trace to DataNode#shutdown to track progress

• Type: Sub-task
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 2.6.0
• Assignee: Xiaoyu Yao
• Reporter: Xiaoyu Yao
• Created: 18/May/15 21:10
• Updated: 18/May/15 21:10

Description

We have seen Datanode#shutdown wait infinitely during rolling upgrade but there is no progress information logged in datanode log.

This ticket propose to add additional tracing after completion of sub tasks such as:

    // Terminate directory scanner and block scanner
    shutdownPeriodicScanners();
   
   // TODO: log completion

    // wait reconfiguration thread, if any, to exit
    shutdownReconfigurationTask();

  // TODO: log completion

    if (this.dataXceiverServer != null) {
      // wait for dataXceiverServer to terminate
      try {
        this.dataXceiverServer.join();
      } catch (InterruptedException ie) {
      }
    }
   
    // TODO: log completion 
    
    if (this.localDataXceiverServer != null) {
      // wait for localDataXceiverServer to terminate
      try {
        this.localDataXceiverServer.join();
      } catch (InterruptedException ie) {
      }
    }
    
    // TODO: log completion

    ...