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

Posted to commits@hudi.apache.org by "sivabalan narayanan (Jira)" <ji...@apache.org> on 2024/10/01 01:34:00 UTC

[jira] [Commented] (HUDI-8189) Introduce a `.abort` state for abandoned commits and table services

    [ https://issues.apache.org/jira/browse/HUDI-8189?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17886055#comment-17886055 ] 

sivabalan narayanan commented on HUDI-8189:
-------------------------------------------

We have proposed a design here [https://github.com/apache/hudi/pull/11555/] Will be flushing out the details w/ more finer details and alternate design options. If anyone wants to review in the mean time. 

> Introduce a `.abort` state for abandoned commits and table services
> -------------------------------------------------------------------
>
>                 Key: HUDI-8189
>                 URL: https://issues.apache.org/jira/browse/HUDI-8189
>             Project: Apache Hudi
>          Issue Type: Improvement
>            Reporter: Ethan Guo (this is the old account; please use "yihua")
>            Assignee: sivabalan narayanan
>            Priority: Blocker
>             Fix For: 1.0.0
>
>   Original Estimate: 16h
>          Time Spent: 0.5h
>  Remaining Estimate: 15.5h
>
> We should support a new state `.abort` so that the commits and table service instants can be cancelled on the timeline (right now after clustering is scheduled, it cannot be cancelled).  In this way, we can filter out such instants without having to do rollbacks.  This ticket also involves making the new state `.abort` used by all writers and readers wrt timeline and FileSystem view so it works end-to-end.



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