You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hudi.apache.org by "Danny Chen (Jira)" <ji...@apache.org> on 2023/04/07 03:14:00 UTC
[jira] [Updated] (HUDI-6045) Support append mode by default for MOR table with INSERT operation
[ https://issues.apache.org/jira/browse/HUDI-6045?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Danny Chen updated HUDI-6045:
-----------------------------
Summary: Support append mode by default for MOR table with INSERT operation (was: Adjust HoodieTableSink for sink operator generation)
> Support append mode by default for MOR table with INSERT operation
> ------------------------------------------------------------------
>
> Key: HUDI-6045
> URL: https://issues.apache.org/jira/browse/HUDI-6045
> Project: Apache Hudi
> Issue Type: Improvement
> Components: writer-core
> Reporter: zhuanshenbsj1
> Priority: Minor
> Labels: pull-request-available
> Fix For: 0.14.0
>
>
> # Insert scenario in the COW table, if both online sync-clustering and online async-clustering(plan generate || plan execute) are configured simultaneously, sometimes sync-clustering takes effect, sometimes async-clustering takes effect:
> sync-clustering=true & generate async-clustering=true & execute async-clustering=true:sync-clustering takes effect
> sync-clustering=true & generate async-clustering=false & execute async-clustering=true: async-clustering takes effect
> # Insert scenario in the MOR table,sometimes generate log files, sometimes generate parquet files.
> async-compaction=true & generate async-clustering=false : generate log files
> async-compaction=false & generate async-clustering=true: generate parquet files
> This will cause confusion for users
> After modification:
> # Insert scenario in the COW table, sync-clustering has higher priority than online async-clusering.
> # Insert scenario in the MOR table will always generate parquet files
--
This message was sent by Atlassian Jira
(v8.20.10#820010)