[KAFKA-3693] Lost highwatermark at broker start-up

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 0.9.0.1
• Component/s: core
• Labels: reliability
• Assignee: Unassigned
• Reporter: Maysam Yabandeh
• Created: 10/May/16 23:33
• Updated: 30/Nov/16 16:14

Description

Upon broker start-up, a race between highwatermark-checkpoint thread to write replication-offset-checkpoint file and handleLeaderAndIsrRequest thread reading from it causes the highwatermark for some partitions to be reset to 0. In the good case, this results the replica to truncate its entire log to 0 and hence initiates fetching of terabytes of data from the lead broker, which sometimes leads to hours of downtime. We observed the bad cases that the reset offset can propagate to recovery-point-offset-checkpoint file, making a lead broker to truncate the file. This seems to have the potential to lead to data loss if the truncation happens at both follower and leader brokers.

This is the particular faulty scenario manifested in our tests:

1. The broker restarts and receive LeaderAndIsr from the controller
2. LeaderAndIsr message however does not contain all the partitions (probably because other brokers were churning at the same time)
3. becomeLeaderOrFollower calls getOrCreatePartition and updates the allPartitions with the partitions included in the LeaderAndIsr message

     def getOrCreatePartition(topic: String, partitionId: Int): Partition = {
       var partition = allPartitions.get((topic, partitionId))
       if (partition == null) {
         allPartitions.putIfNotExists((topic, partitionId), new Partition(topic, partitionId, time, this))
   

4. replication-offset-checkpoint jumps in taking a snapshot of (the partial) allReplicas' high watermark into replication-offset-checkpoint file

     def checkpointHighWatermarks() {
       val replicas = allPartitions.values.map(_.getReplica(config.brokerId)).collect{case Some(replica) => replica}

   hence rewriting the previous highwatermarks.

5. Later becomeLeaderOrFollower calls makeLeaders and makeFollowers which read the (now partial) file through Partition::getOrCreateReplica

             val checkpoint = replicaManager.highWatermarkCheckpoints(log.dir.getParentFile.getAbsolutePath)
             val offsetMap = checkpoint.read
             if (!offsetMap.contains(TopicAndPartition(topic, partitionId)))
               info("No checkpointed highwatermark is found for partition [%s,%d]".format(topic, partitionId))
   

We are not entirely sure whether the initial LeaderAndIsr message including a subset of partitions is critical in making this race condition manifest or not. But it is an important detail since it clarifies that a solution based on not letting the highwatermark-checkpoint thread jumping in the middle of processing a LeaderAndIsr message would not suffice.

The solution we are thinking of is to force initializing allPartitions by the partitions listed in the replication-offset-checkpoint (and perhaps recovery-point-offset-checkpoint file too) when a server starts.

Thoughts?

Issue Links

• is related to: KAFKA-2178 Loss of highwatermarks on incorrect cluster shutdown/restart