[
https://issues.apache.org/jira/browse/HDFS-7702?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14307895#comment-14307895
]
Ray commented on HDFS-7702:
---------------------------
[~zhz], our design for the distributed namenode framework should be considered
orthogonal to HDFS federation, in theory it could be extended from either HDFS
single-namenode or HDFS federation. Moreover, as [~raviprak] pointed out, HDFS
federation only uses static partitioning and we ultimately aim to dynamically
partition namespace across name nodes. However, we still choose HDFS federation
as the source distribution for this work because it provides nice federated
communication models between data nodes and name nodes, and could serve as a
good starting point to accommodate large distributed file systems that don't
fit into a single namenode to begin with. So far me and [~mengmeng] are the
main contributors of this work. We certainly welcome anyone who'd like to join
or comment or give us valuable feedback along the way. We'll provide more
design documents and feel free to discuss it from the high level and we can
dive into the details later. Appreciate your excitement and hope that we can
make this ambitious goal happen.
> Move metadata across namenode - Effort to a real distributed namenode
> ---------------------------------------------------------------------
>
> Key: HDFS-7702
> URL: https://issues.apache.org/jira/browse/HDFS-7702
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Ray
> Assignee: Ray
>
> Implement a tool can show in memory namespace tree structure with
> weight(size) and a API can move metadata across different namenode. The
> purpose is moving data efficiently and faster, without moving blocks on
> datanode.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
