[
https://issues.apache.org/jira/browse/IGNITE-23602?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17895566#comment-17895566
]
Kirill Tkalenko commented on IGNITE-23602:
------------------------------------------
It turns out that it might happen that, when a node is joining (and it acts as
a Metastorage group follower), its local compaction revision is ahead of the
leader's compaction revision (returned by GetCurrentRevision command), but it's
not a problem because because this can also happen during normal cluster
operation, for example, when on one of the nodes we can quickly finish reading
operations from mtastorage and already start compaction on it.
> Add metastorage compaction revision check to divergency validation
> ------------------------------------------------------------------
>
> Key: IGNITE-23602
> URL: https://issues.apache.org/jira/browse/IGNITE-23602
> Project: Ignite
> Issue Type: Improvement
> Reporter: Kirill Tkalenko
> Assignee: Kirill Tkalenko
> Priority: Major
> Labels: ignite-3
> Fix For: 3.0
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Since there is a revision check in the divergency validation, it is necessary
> to add for the compaction revision as well. Since in theory a node could
> become a leader and promote the compaction revision for itself.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
