[
https://issues.apache.org/jira/browse/HDFS-1623?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12993315#comment-12993315
]
Sanjay Radia commented on HDFS-1623:
------------------------------------
There are various steps one can take to improve the availability of the HDFS
Namenode (NN) including reducing startup time, refreshing configuration without
restarting the cluster, reducing upgrade time, and providing an manual or
automatic failover of the NN, etc. This jira focuses on failover of the NN to
address the issue of NN as a single point of failure.
There are various ways to provide failure of NN. Some of these include the use
of shared storage while others do not. Some include the use of IP failover
while others use a smart client side library. One could use Zookeeper for
leader election or some other off the shelf scheme such as Linux HA. These
different solutions can share some framework components as building blocks. The
purpose of this jira is to define these framework components for building
failover solution towards improving availability. The avatar node could use
some of these framework components. Some of the code from avatar node can be
used towards building the framework components.
Sub jiras will be created to address each of the framework components.
> High Availability Framework for HDFS NN
> ---------------------------------------
>
> Key: HDFS-1623
> URL: https://issues.apache.org/jira/browse/HDFS-1623
> Project: Hadoop HDFS
> Issue Type: New Feature
> Reporter: Sanjay Radia
> Assignee: Sanjay Radia
>
--
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
