[ 
https://issues.apache.org/jira/browse/FLINK-5372?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16357276#comment-16357276
 ] 

Tzu-Li (Gordon) Tai edited comment on FLINK-5372 at 2/8/18 5:40 PM:
--------------------------------------------------------------------

[~srichter] what is the status of this issue? Can we move this to be fixed for 
1.4.2?
Since AFAICT this is a test instability, I will change the fix version for this 
to be 1.4.2 for now. If you disagree, please let me know.


was (Author: tzulitai):
[~srichter] what is the status of this issue? Can we move this to be fixed for 
1.4.2?
I will change the fix version for this to be 1.4.2 for now. If you disagree, 
please let me know.

> Fix RocksDBAsyncSnapshotTest.testCancelFullyAsyncCheckpoints()
> --------------------------------------------------------------
>
>                 Key: FLINK-5372
>                 URL: https://issues.apache.org/jira/browse/FLINK-5372
>             Project: Flink
>          Issue Type: Bug
>          Components: DataStream API
>    Affects Versions: 1.5.0
>            Reporter: Aljoscha Krettek
>            Assignee: Stefan Richter
>            Priority: Blocker
>              Labels: test-stability
>             Fix For: 1.5.0, 1.4.2
>
>
> The test is currently {{@Ignored}}. We have to change 
> {{AsyncCheckpointOperator}} to make sure that we can run fully 
> asynchronously. Then, the test will still fail because the canceling 
> behaviour was changed in the meantime.
> {code}
> public static class AsyncCheckpointOperator
>     extends AbstractStreamOperator<String>
>     implements OneInputStreamOperator<String, String> {
>     @Override
>     public void open() throws Exception {
>         super.open();
>         // also get the state in open, this way we are sure that it was 
> created before
>         // we trigger the test checkpoint
>         ValueState<String> state = getPartitionedState(
>                 VoidNamespace.INSTANCE,
>                 VoidNamespaceSerializer.INSTANCE,
>                 new ValueStateDescriptor<>("count",
>                         StringSerializer.INSTANCE, "hello"));
>     }
>     @Override
>     public void processElement(StreamRecord<String> element) throws Exception 
> {
>         // we also don't care
>         ValueState<String> state = getPartitionedState(
>                 VoidNamespace.INSTANCE,
>                 VoidNamespaceSerializer.INSTANCE,
>                 new ValueStateDescriptor<>("count",
>                         StringSerializer.INSTANCE, "hello"));
>         state.update(element.getValue());
>     }
>     @Override
>     public void snapshotState(StateSnapshotContext context) throws Exception {
>         // do nothing so that we don't block
>     }
> }
> {code}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to