hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Kihwal Lee (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-5585) Provide admin commands for data node upgrade
Date Wed, 15 Jan 2014 15:59:25 GMT

    [ https://issues.apache.org/jira/browse/HDFS-5585?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13872214#comment-13872214

Kihwal Lee commented on HDFS-5585:

bq. The nice thing about going through NN is NN has the state and is able to decide the order
in which DNs are restarted to minimize the impact on write and read operation.

After getBlockLocations() or addAdditionalBlock(), NN can only guess what is happening between
clients and datanodes. We thought about NN-driven approaches, but decided not to do it because,
- DNs can make better decisions as they have more up-to-date and complete state
- Not a good idea to overload the less-than-ideal include/exclude mechanism. (We need to fix
this for HA and various hadoop daemons.)
- The quick restart logic and command can also be used for non-upgrade cases.
- Less dependency is better.

I think it will also be easier for existing upgrade/management tools.

> Provide admin commands for data node upgrade
> --------------------------------------------
>                 Key: HDFS-5585
>                 URL: https://issues.apache.org/jira/browse/HDFS-5585
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: datanode, ha, hdfs-client, namenode
>            Reporter: Kihwal Lee
> Several new methods to ClientDatanodeProtocol may need to be added to support querying
version, initiating upgrade, etc.  The admin CLI needs to be added as well. This primary use
case is for rolling upgrade, but this can be used for preparing for a graceful restart of
a data node for any reasons.

This message was sent by Atlassian JIRA

View raw message