[
https://issues.apache.org/jira/browse/FLINK-2523?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15115235#comment-15115235
]
ASF GitHub Bot commented on FLINK-2523:
---------------------------------------
Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/1546#issuecomment-174518267
Thank you for this patch. The idea is good, but I think it needs a bit of
improvement:
- The value is pulled in a fairly complicated way though all the layers
and abstractions. That is not really necessary, as soon as it is in the
`ExecutionConfig`, the `Task` can access the values. No need to change the
TaskDeploymentDescriptior, JobGraph, ExecutionGraph, etc. This probably needs
changes in 2 files (otherwise, our abstractions are messed up).
- `taskCancellationDelay` is not a very intuitive name, in my opinion. It
suggests that cancellation is delayed, which is not the case. Tasks are
actually canceled a single time, the blocked threads are only repeatedly
interrupted.
- We should not try to show off such parameters in the examples. The
examples need to stay simple, or we confuse people with parameters they need
usually not worry about.
> Make task canceling interrupt interval configurable
> ---------------------------------------------------
>
> Key: FLINK-2523
> URL: https://issues.apache.org/jira/browse/FLINK-2523
> Project: Flink
> Issue Type: Improvement
> Components: TaskManager
> Affects Versions: 0.10.0
> Reporter: Stephan Ewen
> Assignee: Kostas
> Fix For: 1.0.0
>
>
> When a task is canceled, the cancellation calls periodically "interrupt()" on
> the task thread, if the task thread does not cancel with a certain time.
> Currently, this value is hard coded to 10 seconds. We should make that time
> configurable.
> Until then, I would like to increase the value to 30 seconds, as many tasks
> (here I am observing it for Kafka consumers) can take longer then 10 seconds
> for proper cleanup.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
