[
https://issues.apache.org/jira/browse/FLINK-10380?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16642359#comment-16642359
]
ASF GitHub Bot commented on FLINK-10380:
----------------------------------------
satybald commented on issue #6724: [FLINK-10380] Add precondition for
assignToKeyGroup method
URL: https://github.com/apache/flink/pull/6724#issuecomment-427952541
@StephanEwen thank you for feedback! From what I see,
`KeyGroupRangeAssignment#assignToKeyGroup` is utility class. It has 36 usages
in different classes: state backends, operators, and tests.
`AbstractKeyedStateBackend#setCurrentKey`, which uses the utilty method has 296
usages.
I'm bit afraid that catching and testing NPE earlier in the chain requires a
significant effort to rewrite everything without much benefit. As try/catch
doesn't make code cleaner and adding the same error message will violate DRY
principle.
Also, determine if the key is null earlier in the chain without
preconditions with try/catch might lead to false positives as there's another
reason for child method throwing NPE.
From what I see `computeKeyGroupForKeyHash(key.hashCode(), maxParallelism);`
is the place where key group index calcluated and NPE is thrown in case key is
null. Would you agree that catching NPE in utility class and overriding with a
specific message is viable alternative?
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
> Check if key is not nul before assign to group in KeyedStream
> -------------------------------------------------------------
>
> Key: FLINK-10380
> URL: https://issues.apache.org/jira/browse/FLINK-10380
> Project: Flink
> Issue Type: Task
> Affects Versions: 1.6.0
> Reporter: Sayat Satybaldiyev
> Priority: Minor
> Labels: pull-request-available
>
> If a user creates a KeyedStream and partition by key which might be null,
> Flink job throws NullPointerExceptoin at runtime. However, NPE that Flink
> throws hard to debug and understand as it doesn't refer to place in Flink job.
> *Suggestion:*
> Add precondition that checks if the key is not null and throw a descriptive
> error if it's a null.
>
> *Job Example*:
>
> {code:java}
> DataStream<String> stream = env.fromCollection(Arrays.asList("aaa", "bbb"))
> .map(x -> (String)null)
> .keyBy(x -> x);{code}
>
>
> An error that is thrown:
>
> {code:java}
> Exception in thread "main"
> org.apache.flink.runtime.client.JobExecutionException:
> java.lang.RuntimeException
> at
> org.apache.flink.runtime.minicluster.MiniCluster.executeJobBlocking(MiniCluster.java:623)
> at
> org.apache.flink.streaming.api.environment.LocalStreamEnvironment.execute(LocalStreamEnvironment.java:123)
> at org.myorg.quickstart.BuggyKeyedStream.main(BuggyKeyedStream.java:61)
> Caused by: java.lang.RuntimeException
> at
> org.apache.flink.streaming.runtime.io.RecordWriterOutput.pushToRecordWriter(RecordWriterOutput.java:110)
> at
> org.apache.flink.streaming.runtime.io.RecordWriterOutput.collect(RecordWriterOutput.java:89)16:26:43,110
> INFO org.apache.flink.runtime.rpc.akka.AkkaRpcService - Stopped Akka RPC
> service.
> at
> org.apache.flink.streaming.runtime.io.RecordWriterOutput.collect(RecordWriterOutput.java:45)
> at
> org.apache.flink.streaming.api.operators.AbstractStreamOperator$CountingOutput.collect(AbstractStreamOperator.java:689)
> at
> org.apache.flink.streaming.api.operators.AbstractStreamOperator$CountingOutput.collect(AbstractStreamOperator.java:667)
> at
> org.apache.flink.streaming.api.operators.StreamMap.processElement(StreamMap.java:41)
> at
> org.apache.flink.streaming.runtime.io.StreamInputProcessor.processInput(StreamInputProcessor.java:202)
> at
> org.apache.flink.streaming.runtime.tasks.OneInputStreamTask.run(OneInputStreamTask.java:105)
> at
> org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:300)
> at org.apache.flink.runtime.taskmanager.Task.run(Task.java:711)
> at java.lang.Thread.run(Thread.java:748)
> Caused by: java.lang.NullPointerException
> at
> org.apache.flink.runtime.state.KeyGroupRangeAssignment.assignToKeyGroup(KeyGroupRangeAssignment.java:59)
> at
> org.apache.flink.runtime.state.KeyGroupRangeAssignment.assignKeyToParallelOperator(KeyGroupRangeAssignment.java:48)
> at
> org.apache.flink.streaming.runtime.partitioner.KeyGroupStreamPartitioner.selectChannels(KeyGroupStreamPartitioner.java:63)
> at
> org.apache.flink.streaming.runtime.partitioner.KeyGroupStreamPartitioner.selectChannels(KeyGroupStreamPartitioner.java:32)
> at
> org.apache.flink.runtime.io.network.api.writer.RecordWriter.emit(RecordWriter.java:104)
> at
> org.apache.flink.streaming.runtime.io.StreamRecordWriter.emit(StreamRecordWriter.java:81)
> at
> org.apache.flink.streaming.runtime.io.RecordWriterOutput.pushToRecordWriter(RecordWriterOutput.java:107)
> {code}
> ... 10 more
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
