[
https://issues.apache.org/jira/browse/HBASE-21641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16729141#comment-16729141
]
Sean Busbey commented on HBASE-21641:
-------------------------------------
Please write a scope document. Consider emailing dev@hbase to gather feedback
once you have one.
Important bits to cover:
* What use cases this enables
* How will this impact upgrades
* How/where we'll document this
* How we'll test this feature
Some example scope documents:
* [HBase Spark
Integration|https://issues.apache.org/jira/secure/attachment/12878023/Apache%20HBase%20-%20Apache%20Spark%20Integration%20Scope%20-%20update%201.pdf]
(HBASE-18405)
* [Read Replica
Clusters|https://issues.apache.org/jira/secure/attachment/12888376/HBase%20Read-Replica%20Clusters%20Scope%20doc_v2.pdf]
(HBASE-18477)
> RegionServer Group based WAL creation & replication
> ---------------------------------------------------
>
> Key: HBASE-21641
> URL: https://issues.apache.org/jira/browse/HBASE-21641
> Project: HBase
> Issue Type: Improvement
> Components: wal
> Reporter: Nikhil Bafna
> Priority: Minor
>
> I'm looking at extending HBASE-6721 to apply it to WALs such that WALs are
> created & replicated within an RSGroup. This extends multi-tenancy to WALs
> also, and not just cover Hbase data. I was working out of 1.2.x code. The
> feedback on mailing list was to continue the discussion on Jira and target 2.x
>
> The approach I'm using is
> - Strategy interface for WAL placement on the filesystem. Default to
> delegate it to respective filesystem (which is the old behavior). FavoredNode
> strategy computes the favoured nodes from the RSGroup memberships.
> - FavoredNode strategy requires instance of hbase.Server, to get the current
> server name and an instance of Zookeeper watch to listen for changes to
> RSGroup memberships
> - The strategy is initialised in HRegionServer init and set in a static
> field in DefaultWALProvider
> - DefaultWALProvider.Writer takes the strategy in its init, and invokes it
> before output stream creation and passes the favored nodes information to
> DistributedFileSystem.create()
>
> Few questions
> - Any glaring miss in the approach?
> - I have hesitation in setting the strategy in static field in
> DefaultWALProvider. I would have preferred it to be passed in "init" itself,
> but that change seems to be too expansive.
> - Also, this introduces the dependency of server/zookeeper instance inside
> the WAL code path, which seems to be not there till now. Is that an explicit
> choice to keep them separate?
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
