You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@kudu.apache.org by "Grant Henke (Jira)" <ji...@apache.org> on 2020/02/18 13:40:00 UTC
[jira] [Commented] (KUDU-3056) kudu-spark HdrHistogramAccumulator
is too big, and make spark job failed
[ https://issues.apache.org/jira/browse/KUDU-3056?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17039087#comment-17039087 ]
Grant Henke commented on KUDU-3056:
-----------------------------------
I have not seen this reported before. It could be useful to disable the histogram accumulators if this is an ongoing issue. There may also be an opportunity to mak% them smaller.
As a workaround could you increase spark.driver.maxResultSize?
> kudu-spark HdrHistogramAccumulator is too big, and make spark job failed
> --------------------------------------------------------------------------
>
> Key: KUDU-3056
> URL: https://issues.apache.org/jira/browse/KUDU-3056
> Project: Kudu
> Issue Type: Bug
> Components: spark
> Affects Versions: 1.9.0
> Reporter: caiconghui
> Priority: Major
> Fix For: NA
>
> Attachments: heap1.png, heap2.png, heap3.png
>
> Original Estimate: 12h
> Remaining Estimate: 12h
>
> in production envrinment, we use kudu-spark to read kudu tablel but even we don't use the
> HdrHistogramAccumulator, the HdrHistogramAccumulator stored in an array is stiil so big,
> totoal of them are almost 2 MB, so that when the number of kudu-spark task(for read kudu data and shuffle) is more than 900, the spark job failed, and the follwing error occured,
>
> Job aborted due to stage failure: Total size of serialized results of 1413 tasks (3.0 GB) is bigger than spark.driver.maxResultSize (3.0 GB)
--
This message was sent by Atlassian Jira
(v8.3.4#803005)