[
https://issues.apache.org/jira/browse/HBASE-23697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17016382#comment-17016382
]
Michael Stack commented on HBASE-23697:
---------------------------------------
>From Duo on why hfiles for the buried ProcedureStore Region are on WALFS
>rather than RSFS: "On why we put all things on the WAL filesystem, partly
>because that I want to put all the data under a single directory, which can
>hide the internal implementation details about procedure store, and easy to
>replace the old implementation with another implementation.
And another thing is about performance. The WAL filesystem is HDFS, and the
HFile storage, if not the same one, is usually S3 or some other object
storages, which are very slow, usually. And usually the size of the procedure
store will be small, I think the cost is also fine."
> Document new RegionProcedureStore operation and migration
> ---------------------------------------------------------
>
> Key: HBASE-23697
> URL: https://issues.apache.org/jira/browse/HBASE-23697
> Project: HBase
> Issue Type: Sub-task
> Components: documentation
> Affects Versions: 2.3.0
> Reporter: Michael Stack
> Priority: Major
>
> Add a few notes to the refguide on the new RegionProcedureStore, how it works
> (A 'Region' but buried in the Master with dedicated flushing/compacting
> threads and archivers for WAL and hfile), how it differs from WALPS, and note
> it auto-migrates and there should be new issue moving on to the new store.
> Mention the configuration. Mention it is on WALFS even though it is a
> 'Region', etc.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
