You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "Caizhi Weng (Jira)" <ji...@apache.org> on 2022/02/24 02:58:00 UTC
[jira] [Commented] (FLINK-26247) Introduce a better expire strategy
[ https://issues.apache.org/jira/browse/FLINK-26247?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17497132#comment-17497132 ]
Caizhi Weng commented on FLINK-26247:
-------------------------------------
Why is this solving the problem? With the snapshot id, we can only know when this file/manifest is added but we can't know when it is safe to be deleted.
I'd suggest another method. Just write down all removed files/manifests in a separate file (a file just like the manifest list) and we just need to look at that file when expiring a snapshot.
> Introduce a better expire strategy
> ----------------------------------
>
> Key: FLINK-26247
> URL: https://issues.apache.org/jira/browse/FLINK-26247
> Project: Flink
> Issue Type: Sub-task
> Components: Table Store
> Reporter: Jingsong Lee
> Priority: Major
> Fix For: table-store-0.1.0
>
>
> We can add the snapshot id in which the file/manifest was added to the table. With this snapshot id, we can have better expire strategy. Instead of scanning all files of the snapshot.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)