[
https://issues.apache.org/jira/browse/HDFS-6469?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Konstantin Shvachko updated HDFS-6469:
--------------------------------------
Attachment: CNodeDesign.pdf
In a nutshell.
Namespace updates are initiated on one of the instances of ConsensusNode
(CNode), which submits each update to the Coordination Engine as a proposal.
The Coordination Engine agrees on the order of the updates and guarantees that
the updates are delivered to all CNodes as agreements in the same order. The
actual updates to the namespace are carried over as the result of executing the
corresponding agreement. This gives clients a consistent view of the namespace
across all replicas.
The replication technique provides active-active model of high availability for
HDFS with which namespace requests (read or write) can be load balanced between
multiple NameNode replicas.
Coordination Engine (CE) can use different algorithms and have different
implementations. To me it would make sense to track the APIs and reference
implementation(s) for CE in hadoop-common, so that it could be used as a
pluggable component in different projects, including HBase.
This jira is primarily to track the ConsensusNode development.
> Coordinated replication of the namespace using ConsensusNode
> ------------------------------------------------------------
>
> Key: HDFS-6469
> URL: https://issues.apache.org/jira/browse/HDFS-6469
> Project: Hadoop HDFS
> Issue Type: New Feature
> Components: namenode
> Affects Versions: 3.0.0
> Reporter: Konstantin Shvachko
> Assignee: Konstantin Shvachko
> Attachments: CNodeDesign.pdf
>
>
> This is a proposal to introduce ConsensusNode - an evolution of the NameNode,
> which enables replication of the namespace on multiple nodes of an HDFS
> cluster by means of a Coordination Engine.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
