You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@hudi.apache.org by "Raymond Xu (Jira)" <ji...@apache.org> on 2023/01/03 02:40:00 UTC

[jira] [Updated] (HUDI-5321) Fix Bulk Insert ColumnSortPartitioners

     [ https://issues.apache.org/jira/browse/HUDI-5321?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Raymond Xu updated HUDI-5321:
-----------------------------
    Reviewers: Raymond Xu

> Fix Bulk Insert ColumnSortPartitioners
> --------------------------------------
>
>                 Key: HUDI-5321
>                 URL: https://issues.apache.org/jira/browse/HUDI-5321
>             Project: Apache Hudi
>          Issue Type: Bug
>    Affects Versions: 0.12.1
>            Reporter: Alexey Kudinkin
>            Assignee: Jonathan Vexler
>            Priority: Blocker
>              Labels: pull-request-available
>             Fix For: 0.13.0
>
>
> Currently, all of the Custom Bulk Insert ColumnSortPartitioner impls incorrectly return "true" from the "arePartitionRecordsSorted" method, even though records might not necessarily be sorted by the partition-path columns as is required by this method.
> In case when such Partitioner is used and the data is NOT sorted by the list of columns that start w/ partition ones, this could lead to a Parquet writers being closed prematurely when writing files creating a LOT of small files.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)