You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Konstantin Knauf (Jira)" <ji...@apache.org> on 2021/07/29 07:30:10 UTC
[jira] [Reopened] (FLINK-16365) awaitTermination() result is not
checked
[ https://issues.apache.org/jira/browse/FLINK-16365?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Konstantin Knauf reopened FLINK-16365:
--------------------------------------
Re-opening in accordance with https://issues.apache.org/jira/browse/FLINK-23206.
> awaitTermination() result is not checked
> ----------------------------------------
>
> Key: FLINK-16365
> URL: https://issues.apache.org/jira/browse/FLINK-16365
> Project: Flink
> Issue Type: Improvement
> Components: Connectors / Google Cloud PubSub
> Reporter: Roman Leventov
> Priority: Minor
> Labels: auto-closed
>
> There are three places in production code where awaitTermination() result is not checked: BlockingGrpcPubSubSubscriber (io.grpc.ManagedChannel), PubSubSink (ManagedChannel), and FileCache (ExecutorService).
> Calling awaitTermination() without checking the result seems to make little sense to me.
> If it's genuinely important to await termination, e. g. for concurrency reasons, or because we are awaiting heavy resource release and if the resource is not released we have a resource leak, then it seems reasonable to at least check the result of awaitTermination() and log a warning if the result is negative, allowing to debug potential problem in the future.
> Otherwise, if we don't really care about awaiting termination, then maybe it's better to not call awaitTermination() at all.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)