[
https://issues.apache.org/jira/browse/HDFS-16691?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17576938#comment-17576938
]
Steve Vaughan commented on HDFS-16691:
--------------------------------------
We can work with just a quorum as long as other JournalNodes can catch up.
Syncing handles many issues, but it didn't handle the case where a JournalNode
was unformatted.
> Use quorum instead of requiring full JN set for NN format
> ---------------------------------------------------------
>
> Key: HDFS-16691
> URL: https://issues.apache.org/jira/browse/HDFS-16691
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: namenode
> Environment: Demonstrated in a Kubernetes environment running Java
> 11. Using an HA configuration:
> # Start new cluster, but short 1 JN (minimum quorum, and the missing JN
> won’t resolve). VERIFY:
> - NN formats the 2 existing JN and stabilizes
> - Messages show sync between JN-0 and JN-1, and NN -> JN
> # Scale JN stateful set to add missing JN. NOTE: Requires HDFS-16690
> Reporter: Steve Vaughan
> Assignee: Steve Vaughan
> Priority: Major
>
> Currently a format request fails if any of the JournalNodes is unresolvable.
> For dynamic cluster environments where a JournalNode may not be available
> during the initial formatting step but JournalNodes can self-heal, it makes
> sense to allow the format to succeed when a quorum of JournalNodes is
> available.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
