Hello, I have the following situation: We upgraded our application code on a flink 1.11.1 cluster. We use rocksdb as the state backend. The upgrade used a savepoint from the prior app version. We added a few MapStates to an existing KeyedProcessWindowFunction. This function used to have a single valueState before. We also started using a different WindowTrigger function.
At some point, we had to downgrade the application code. We took a savepoint and restored the old app version using this savepoint. This action itself succeeds. However none of the periodic checkpoints succeed! Every checkpoint fails on this particular KeyedProcessWindowFunction with a Null Pointer Exception. This looks interesting ( https://issues.apache.org/jira/browse/FLINK-11094). However, the bug is old. As far as i understand, this should work. Am I missing something? How can I debug this? This is the stack Trace: 2022-05-16 08:02:19 java.io.IOException: Could not perform checkpoint 1175 for operator XXX -> (Sink: kinesis-sink-1, Sink: kinesis-sink-2, Sink: kinesis-sink-3) (2/2). at org.apache.flink.streaming.runtime.tasks.StreamTask.triggerCheckpointOnBarrier(StreamTask.java:863) at org.apache.flink.streaming.runtime.io.CheckpointBarrierHandler.notifyCheckpoint(CheckpointBarrierHandler.java:113) at org.apache.flink.streaming.runtime.io.CheckpointBarrierAligner.processBarrier(CheckpointBarrierAligner.java:198) at org.apache.flink.streaming.runtime.io.CheckpointedInputGate.pollNext(CheckpointedInputGate.java:93) at org.apache.flink.streaming.runtime.io.StreamTaskNetworkInput.emitNext(StreamTaskNetworkInput.java:158) at org.apache.flink.streaming.runtime.io.StreamOneInputProcessor.processInput(StreamOneInputProcessor.java:67) at org.apache.flink.streaming.runtime.tasks.StreamTask.processInput(StreamTask.java:345) at org.apache.flink.streaming.runtime.tasks.mailbox.MailboxProcessor.runMailboxStep(MailboxProcessor.java:191) at org.apache.flink.streaming.runtime.tasks.mailbox.MailboxProcessor.runMailboxLoop(MailboxProcessor.java:181) at org.apache.flink.streaming.runtime.tasks.StreamTask.runMailboxLoop(StreamTask.java:558) at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:530) at org.apache.flink.runtime.taskmanager.Task.doRun(Task.java:721) at org.apache.flink.runtime.taskmanager.Task.run(Task.java:546) at java.lang.Thread.run(Thread.java:748) Caused by: org.apache.flink.runtime.checkpoint.CheckpointException: Could not complete snapshot 1175 for operator XXX -> (Sink: kinesis-sink-1, Sink: kinesis-sink-2, Sink: kinesis-sink-3) (2/2). Failure reason: Checkpoint was declined. at org.apache.flink.streaming.api.operators.StreamOperatorStateHandler.snapshotState(StreamOperatorStateHandler.java:215) at org.apache.flink.streaming.api.operators.StreamOperatorStateHandler.snapshotState(StreamOperatorStateHandler.java:156) at org.apache.flink.streaming.api.operators.AbstractStreamOperator.snapshotState(AbstractStreamOperator.java:314) at org.apache.flink.streaming.runtime.tasks.SubtaskCheckpointCoordinatorImpl.checkpointStreamOperator(SubtaskCheckpointCoordinatorImpl.java:614) at org.apache.flink.streaming.runtime.tasks.SubtaskCheckpointCoordinatorImpl.buildOperatorSnapshotFutures(SubtaskCheckpointCoordinatorImpl.java:540) at org.apache.flink.streaming.runtime.tasks.SubtaskCheckpointCoordinatorImpl.takeSnapshotSync(SubtaskCheckpointCoordinatorImpl.java:507) at org.apache.flink.streaming.runtime.tasks.SubtaskCheckpointCoordinatorImpl.checkpointState(SubtaskCheckpointCoordinatorImpl.java:266) at org.apache.flink.streaming.runtime.tasks.StreamTask.lambda$performCheckpoint$5(StreamTask.java:892) at org.apache.flink.streaming.runtime.tasks.StreamTaskActionExecutor$1.runThrowing(StreamTaskActionExecutor.java:47) at org.apache.flink.streaming.runtime.tasks.StreamTask.performCheckpoint(StreamTask.java:882) at org.apache.flink.streaming.runtime.tasks.StreamTask.triggerCheckpointOnBarrier(StreamTask.java:850) ... 13 more Caused by: java.lang.NullPointerException Thanks Sudharsan
