[jira] [Commented] (IGNITE-20209) Catch-up rebalance triggers on node restart

["Alexander Lapin (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IGNITE-20209?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17809485#comment-17809485 ] 

Alexander Lapin commented on IGNITE-20209:
------------------------------------------

[~kgusakov] LGTM, Thanks!

> Catch-up rebalance triggers on node restart
> -------------------------------------------
>
>                 Key: IGNITE-20209
>                 URL: https://issues.apache.org/jira/browse/IGNITE-20209
>             Project: Ignite
>          Issue Type: Improvement
>            Reporter: Alexander Lapin
>            Assignee: Kirill Gusakov
>            Priority: Major
>              Labels: ignite-3
>          Time Spent: 40m
>  Remaining Estimate: 0h
>
> h3. Motivation
> Please check https://issues.apache.org/jira/browse/IGNITE-20187 for more context, that is about catching-up assignments.pending meta storage keys, whether given one is about catching-up its triggers:
>  * Replica factor updates.
>  * Partitions count updates. Immutable for now.
>  * Data nodes updates.
>  * Replica storage addition/removal. !By the way, is it possible to remove replica storage.
> For all aforementioned cases, it's required to update distributed assignments pending (planned) keys if it's not yet done. And the only difficulty here is precisely in understanding whether this was done or not.
> h3. Definition of Done
> Updated distributed assignments pending(planned) keys if necessary according to the current triggers state.
>  
> Notes:
> 1) -Add to metastorage starting revision- ({{{}metaStorageMgr.recoveryFinishedFuture(){}}} returns long with the maximal recovered revision)



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