[
https://issues.apache.org/jira/browse/HUDI-6252?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ethan Guo updated HUDI-6252:
----------------------------
Component/s: cleaning
> Deduce archival config based on hours derived from cleaner hours based config
> -----------------------------------------------------------------------------
>
> Key: HUDI-6252
> URL: https://issues.apache.org/jira/browse/HUDI-6252
> Project: Apache Hudi
> Issue Type: Improvement
> Components: archiving, cleaning, table-service
> Reporter: sivabalan narayanan
> Priority: Major
>
> We have hours based policy/configs for cleaning, but archival only has
> num_commits based configs. So, as of now, user/dev has to ensure they both
> align well together. We should derive archival configs automatically from
> cleaner configs.
>
> for eg,
> if cleaner configs is 48 hours,
> we can add 10% more for archival min commits and 20% for max commits.
> So, archival window can be 53 hours -> 58 hours. we can expose configs for
> these percentage if need be.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
