[
https://issues.apache.org/jira/browse/FLINK-16069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17045531#comment-17045531
]
Zhu Zhu commented on FLINK-16069:
---------------------------------
>> Is there anything in the ShuffleDescriptor that would be different depending
>> on who reads the intermediate result?
I think not. ShuffleDescriptor does not depend on the consumer. It is created
from PartitionDescriptor and ProducerDescriptor when the producer is assigned a
slot and the partition is registered to the ShuffleMaster.
That also means we cannot generate {{ShuffleDescriptor}} before scheduling the
tasks. (this answers "In general, can we not create the shuffle descriptor when
the task scheduled?" if I understand correctly)
> Creation of TaskDeploymentDescriptor can block main thread for long time
> ------------------------------------------------------------------------
>
> Key: FLINK-16069
> URL: https://issues.apache.org/jira/browse/FLINK-16069
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Coordination
> Reporter: huweihua
> Priority: Major
>
> The deploy of tasks will take long time when we submit a high parallelism
> job. And Execution#deploy run in mainThread, so it will block JobMaster
> process other akka messages, such as Heartbeat. The creation of
> TaskDeploymentDescriptor take most of time. We can put the creation in future.
> For example, A job [source(8000)->sink(8000)], the total 16000 tasks from
> SCHEDULED to DEPLOYING took more than 1mins. This caused the heartbeat of
> TaskManager timeout and job never success.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
