Uploaded image for project: 'ZooKeeper'
  1. ZooKeeper
  2. ZOOKEEPER-2380

Deadlock between leader shutdown and forwarding ACK to the leader

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Closed
    • Blocker
    • Resolution: Fixed
    • None
    • 3.5.2, 3.6.0
    • server
    • None
    • Reviewed

    Description

      Zookeeper enters into deadlock while shutting down itself, thus making zookeeper service unavailable as deadlocked server is a leader. Here is the thread dump:

      "QuorumPeer[myid=1](plain=/0:0:0:0:0:0:0:0:2181)(secure=disabled)" #25 prio=5 os_prio=0 tid=0x00007fbc502a6800 nid=0x834 in Object.wait() [0x00007fbc4d9a8000]      java.lang.Thread.State: WAITING (on object monitor)      at java.lang.Object.wait(Native Method)      at java.lang.Thread.join(Thread.java:1245)      - locked <
      0x00000000feb78000> (a org.apache.zookeeper.server.SyncRequestProcessor)      at java.lang.Thread.join(Thread.java:1319)      at org.apache.zookeeper.server.SyncRequestProcessor.shutdown(SyncRequestProcessor.java:196)      at org.apache.zookeeper.server.quorum.ProposalRequestProcessor.shutdown(ProposalRequestProcessor.java:90)      at org.apache.zookeeper.server.PrepRequestProcessor.shutdown(PrepRequestProcessor.java:1016)      at org.apache.zookeeper.server.quorum.LeaderRequestProcessor.shutdown(LeaderRequestProcessor.java:78)      at org.apache.zookeeper.server.ZooKeeperServer.shutdown(ZooKeeperServer.java:561)      - locked <
      0x00000000feb61e20> (a org.apache.zookeeper.server.quorum.LeaderZooKeeperServer)      at org.apache.zookeeper.server.quorum.QuorumZooKeeperServer.shutdown(QuorumZooKeeperServer.java:169)      - locked <
      0x00000000feb61e20> (a org.apache.zookeeper.server.quorum.LeaderZooKeeperServer)      at org.apache.zookeeper.server.quorum.LeaderZooKeeperServer.shutdown(LeaderZooKeeperServer.java:102)      - locked <
      0x00000000feb61e20> (a org.apache.zookeeper.server.quorum.LeaderZooKeeperServer)      at org.apache.zookeeper.server.quorum.Leader.shutdown(Leader.java:637)      at org.apache.zookeeper.server.quorum.Leader.lead(Leader.java:590)      - locked <
      0x00000000feb781a0> (a org.apache.zookeeper.server.quorum.Leader)      at org.apache.zookeeper.server.quorum.QuorumPeer.run(QuorumPeer.java:1108)
      
      
      "SyncThread:1" #46 prio=5 os_prio=0 tid=0x00007fbc5848f000 nid=0x867 waiting for monitor entry [0x00007fbc4ca90000]      java.lang.Thread.State: BLOCKED (on object monitor)      at org.apache.zookeeper.server.quorum.Leader.processAck(Leader.java:784)      - waiting to lock <0x00000000feb781a0> (a org.apache.zookeeper.server.quorum.Leader)      at org.apache.zookeeper.server.quorum.AckRequestProcessor.processRequest(AckRequestProcessor.java:46)      at org.apache.zookeeper.server.SyncRequestProcessor.flush(SyncRequestProcessor.java:183)      at org.apache.zookeeper.server.SyncRequestProcessor.run(SyncRequestProcessor.java:113)
      

      Leader.lead() calls shutdown() from the synchronized block, it acquired lock on Leader.java instance

      while (true) {
                      synchronized (this) {
                      long start = Time.currentElapsedTime();
      				.....
      

      In the shutdown flow SyncThread is trying to acquire lock on the same Leader.java instance.

      Leader thread acquired lock and waiting for SyncThread shutdown. SyncThread waiting for the lock to complete its shutdown. This is how ZooKeeper entered into deadlock

      Attachments

        1. ZOOKEEPER-2380-fail.out
          170 kB
          Chris Nauroth
        2. ZOOKEEPER-2380-07.patch
          16 kB
          Mohammad Arshad
        3. ZOOKEEPER-2380-06.patch
          16 kB
          Mohammad Arshad
        4. ZOOKEEPER-2380-05.patch
          15 kB
          Mohammad Arshad
        5. ZOOKEEPER-2380-04.patch
          15 kB
          Mohammad Arshad
        6. ZOOKEEPER-2380-03.patch
          14 kB
          Mohammad Arshad
        7. ZOOKEEPER-2380-02.patch
          2 kB
          Mohammad Arshad
        8. ZOOKEEPER-2380-01.patch
          2 kB
          Mohammad Arshad

        Activity

          People

            arshad.mohammad Mohammad Arshad
            arshad.mohammad Mohammad Arshad
            Votes:
            1 Vote for this issue
            Watchers:
            7 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: