beam-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <>
Subject [jira] [Work logged] (BEAM-4798) IndexOutOfBoundsException when Flink parallelism > 1
Date Mon, 06 Aug 2018 07:47:00 GMT


ASF GitHub Bot logged work on BEAM-4798:

                Author: ASF GitHub Bot
            Created on: 06/Aug/18 07:46
            Start Date: 06/Aug/18 07:46
    Worklog Time Spent: 10m 
      Work Description: aromanenko-dev commented on issue #5996: [BEAM-4798] Fix IndexOutOfBoundsException
in Flink runner
   @aljoscha @aaltay Sorry for delay with response, I was on vacation for last two weeks.
Now I'm back and, in the first order, I'd like to find out why already existed tests were
not failed because they are supposed to cover this test case (when the number of IO readers
is less than value of Flink parallelism value).

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:

Issue Time Tracking

    Worklog Id:     (was: 131329)
    Time Spent: 2h 10m  (was: 2h)

> IndexOutOfBoundsException when Flink parallelism > 1
> ----------------------------------------------------
>                 Key: BEAM-4798
>                 URL:
>             Project: Beam
>          Issue Type: Bug
>          Components: runner-flink
>    Affects Versions: 2.4.0, 2.5.0
>            Reporter: Alexey Romanenko
>            Assignee: Alexey Romanenko
>            Priority: Major
>          Time Spent: 2h 10m
>  Remaining Estimate: 0h
> Running job on Flink in streaming mode and get data from a Kafka topic with parallelism
> 1 causes an exception:
> {noformat}
> Caused by: java.lang.IndexOutOfBoundsException: Index: 0, Size: 0
> 	at java.util.ArrayList.rangeCheck(
> 	at java.util.ArrayList.get(
> 	at
> 	at
> 	at
> 	at
> 	at
> 	at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(
> 	at
> 	at
> {noformat}
> It happens when number of Kafka topic partitions is less than value of parallelism (number
of task slots).
> So, workaround for now can be to set parallelism <= number of topic partitions, thus
if parallelism=2 then number_partitions >= 2

This message was sent by Atlassian JIRA

View raw message