[
https://issues.apache.org/jira/browse/HDFS-8888?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14703624#comment-14703624
]
Colin Patrick McCabe commented on HDFS-8888:
--------------------------------------------
bq. Each volume could become its own RW lock with in the NN. This would improve
parallelism within NN without much additional effort.
Given the problems we already have with large NN heaps, perhaps we would be
better off running multiple Namenode processes than trying to manage multiple
independent subtrees in a single process.
I am also worried that a lot of the changes here seem incompatible. If we are
going to break backwards compatibility, why wouldn't we push people towards
something like ozone, which does have a better horizontal scalability story.
It seems like we should have a design meeting about this before we do any work
in this direction.
> Support volumes in HDFS
> -----------------------
>
> Key: HDFS-8888
> URL: https://issues.apache.org/jira/browse/HDFS-8888
> Project: Hadoop HDFS
> Issue Type: Improvement
> Reporter: Haohui Mai
>
> There are multiple types of zones (e.g., snapshottable directories,
> encryption zones, directories with quotas) which are conceptually close to
> namespace volumes in traditional file systems.
> This jira proposes to introduce the concept of volume to simplify the
> implementation of snapshots and encryption zones.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
