[
https://issues.apache.org/jira/browse/SPARK-26760?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16764803#comment-16764803
]
shahid commented on SPARK-26760:
--------------------------------
Yes. Writing store too frequently maybe a costly operation. So, after a
particular time interval only task info write in the store. So, for a running
job, task info may not update immediately in the UI.
!Screenshot from 2019-02-11 15-09-09.png!
Regarding, how many task actually running for a running job, we need to see the
logs or console, as UI may show slightly more or less depends on the update to
the store.
> [Spark Incorrect display in SPARK UI Executor Tab when number of cores is 4
> and Active Task display as 5 in Executor Tab of SPARK UI]
> -------------------------------------------------------------------------------------------------------------------------------------
>
> Key: SPARK-26760
> URL: https://issues.apache.org/jira/browse/SPARK-26760
> Project: Spark
> Issue Type: Bug
> Components: Spark Core
> Affects Versions: 2.4.0
> Environment: Spark 2.4
> Reporter: ABHISHEK KUMAR GUPTA
> Priority: Major
> Attachments: SPARK-26760.png, Screenshot from 2019-02-11 15-09-09.png
>
>
> Steps:
> # Launch Spark Shell
> # bin/spark-shell --master yarn --conf spark.dynamicAllocation.enabled=true
> --conf spark.dynamicAllocation.initialExecutors=3 --conf
> spark.dynamicAllocation.minExecutors=1 --conf
> spark.dynamicAllocation.executorIdleTimeout=60s --conf
> spark.dynamicAllocation.maxExecutors=5
> # Submit a Job sc.parallelize(1 to 10000,116000).count()
> # Check the YARN UI Executor Tab for the RUNNING application
> # UI display as Number of cores 4 and Active Tasks column shows as 5
> Expected:
> It Number of Active Tasks should be same as Number of Cores.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
