[FLINK-18641] "Failure to finalize checkpoint" error in MasterTriggerRestoreHook

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 1.11.0
• Fix Version/s: 1.11.2, 1.12.0
• Component/s: Runtime / Checkpointing
• Labels: pull-request-available
• Assignee: Jiangjie Qin
• Reporter: Brian Zhou
• Created: 20/Jul/20 07:10
• Updated: 09/Sep/20 06:34
• Resolved: 09/Sep/20 01:33

Description

https://github.com/pravega/flink-connectors is a Pravega connector for Flink. The ReaderCheckpointHook[1] class uses the Flink `MasterTriggerRestoreHook` interface to trigger the Pravega checkpoint during Flink checkpoints to make sure the data recovery. The checkpoint recovery tests are running fine in Flink 1.10, but it has below issues in Flink 1.11 causing the tests time out. Suspect it is related to the checkpoint coordinator thread model changes in Flink 1.11

Error stacktrace:

2020-07-09 15:39:39,999 30945 [jobmanager-future-thread-5] WARN  o.a.f.runtime.jobmaster.JobMaster - Error while processing checkpoint acknowledgement message
org.apache.flink.runtime.checkpoint.CheckpointException: Could not finalize the pending checkpoint 3. Failure reason: Failure to finalize checkpoint.
         at org.apache.flink.runtime.checkpoint.CheckpointCoordinator.completePendingCheckpoint(CheckpointCoordinator.java:1033)
         at org.apache.flink.runtime.checkpoint.CheckpointCoordinator.receiveAcknowledgeMessage(CheckpointCoordinator.java:948)
         at org.apache.flink.runtime.scheduler.SchedulerBase.lambda$acknowledgeCheckpoint$4(SchedulerBase.java:802)
         at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
         at java.util.concurrent.FutureTask.run(FutureTask.java:266)
         at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
         at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
         at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
         at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
         at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.flink.util.SerializedThrowable: Pending checkpoint has not been fully acknowledged yet
         at org.apache.flink.util.Preconditions.checkState(Preconditions.java:195)
         at org.apache.flink.runtime.checkpoint.PendingCheckpoint.finalizeCheckpoint(PendingCheckpoint.java:298)
         at org.apache.flink.runtime.checkpoint.CheckpointCoordinator.completePendingCheckpoint(CheckpointCoordinator.java:1021)
         ... 9 common frames omitted

More detail in this mailing thread: http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/Pravega-connector-cannot-recover-from-the-checkpoint-due-to-quot-Failure-to-finalize-checkpoint-quot-td36652.html
Also in https://github.com/pravega/flink-connectors/issues/387

Issue Links

• links to: GitHub Pull Request #13044