You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hudi.apache.org by "Vinoth Chandar (Jira)" <ji...@apache.org> on 2020/08/05 04:09:00 UTC
[jira] [Updated] (HUDI-1054) Address performance issues with
finalizing writes on S3
[ https://issues.apache.org/jira/browse/HUDI-1054?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vinoth Chandar updated HUDI-1054:
---------------------------------
Status: Closed (was: Patch Available)
> Address performance issues with finalizing writes on S3
> -------------------------------------------------------
>
> Key: HUDI-1054
> URL: https://issues.apache.org/jira/browse/HUDI-1054
> Project: Apache Hudi
> Issue Type: Sub-task
> Components: bootstrap, Common Core, Performance
> Reporter: Udit Mehrotra
> Assignee: Udit Mehrotra
> Priority: Blocker
> Labels: pull-request-available
> Fix For: 0.6.0
>
>
> I have identified 3 performance bottleneck in the [finalizeWrite|https://github.com/apache/hudi/blob/master/hudi-client/src/main/java/org/apache/hudi/table/HoodieTable.java#L378] function, that are manifesting and becoming more prominent with the new bootstrap mechanism on S3:
> * [https://github.com/apache/hudi/blob/5e476733417c3f92ea97d3e5f9a5c8bc48246e99/hudi-client/src/main/java/org/apache/hudi/table/HoodieTable.java#L425] is a serial operation performed at the driver and it can take a long time when you have several partitions and large number of files.
> * The invalid data paths are being stored in a List instead of Set and as a result the following operation becomes N^2 taking significant time to compute at the driver: [https://github.com/apache/hudi/blob/5e476733417c3f92ea97d3e5f9a5c8bc48246e99/hudi-client/src/main/java/org/apache/hudi/table/HoodieTable.java#L429]
> * [https://github.com/apache/hudi/blob/5e476733417c3f92ea97d3e5f9a5c8bc48246e99/hudi-client/src/main/java/org/apache/hudi/table/HoodieTable.java#L473] does a recursive delete of the marker directory at the driver. This is again extremely expensive when you have large number of partitions and files.
>
> Upon testing with a 1 TB data set, having 8000 partitions and approximately 190000 files this whole process consumes *35 minutes*. There is scope to address these performance issues with spark parallelization and using appropriate data structures.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)