[
https://issues.apache.org/jira/browse/TEZ-4059?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Chenren Shao resolved TEZ-4059.
-------------------------------
Resolution: Not A Problem
> TezSplitGrouper doesn't respect tez.grouping.max-size if the originalSplit
> has bigger splits
> --------------------------------------------------------------------------------------------
>
> Key: TEZ-4059
> URL: https://issues.apache.org/jira/browse/TEZ-4059
> Project: Apache Tez
> Issue Type: Bug
> Affects Versions: 0.9.1
> Environment: Tez 0.9.1, Hive 2.3.4, Hadoop 2.8.5 (EMR 5.20.0),
> Reporter: Chenren Shao
> Priority: Minor
> Attachments: split-generation-log.txt
>
>
> In using Tez 0.9.1, Hive 2.3.4, Hadoop 2.8.5 (EMR 5.20.0), I found if we have
> below scenario, TezSplitGrouper is unable to respect tez.grouping.max-size.
> On line 238, Let us say, a customer-defined inputformat generate a M number
> of splits, but its split size is greater than tez.grouping.max-size, then
> line 248 is executed, and a newDesiredNumSplits is calculated based on
> tez.grouping.max-size, N. (N is greater than M), desiredNumSplits is set to
> be N.
> However, when it goes to line 273, since the N is greater than M now, it
> executes line 274, it uses originalSplits M anyway and basically ignores the
> N setting.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
