You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hudi.apache.org by "Manoj Govindassamy (Jira)" <ji...@apache.org> on 2022/01/11 03:44:00 UTC
[jira] [Updated] (HUDI-3144) Make metadata table getRecordsByKeys() operations more performant by doing range reads
[ https://issues.apache.org/jira/browse/HUDI-3144?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Manoj Govindassamy updated HUDI-3144:
-------------------------------------
Summary: Make metadata table getRecordsByKeys() operations more performant by doing range reads (was: Parallelize metadata table getRecordsByKeys() operations)
> Make metadata table getRecordsByKeys() operations more performant by doing range reads
> --------------------------------------------------------------------------------------
>
> Key: HUDI-3144
> URL: https://issues.apache.org/jira/browse/HUDI-3144
> Project: Apache Hudi
> Issue Type: Task
> Reporter: Manoj Govindassamy
> Assignee: Manoj Govindassamy
> Priority: Blocker
> Fix For: 0.11.0
>
>
> When the metadata index lookups is done for several thousands of keys, the keys are looked up in metadata table partitions in a serial fashion key by key, leading to the overall delay in the index lookup.
> # When the indexes are laid out in multiple file groups, the lookup can be parallelized at the filegroup level
> # Even within a single filegroup, sorted keys can be split and the lookup can be done in parallel.
>
>
--
This message was sent by Atlassian Jira
(v8.20.1#820001)