You are viewing a plain text version of this content. The canonical link for it is here.

Posted to common-user@hadoop.apache.org by F21 <f2...@gmail.com> on 2015/12/14 23:38:49 UTC

Minimize downtime when changing hdfs cluster topology

Are there any strategies fro minimizing downtime when changing the 
cluster topology of my hdfs cluster?

For example:

- adding and removing nameservices in an HA cluster to introduce or 
remove namespaces.
- adding or removing journal nodes from qjm
- adding or removing zookeeper addresses from core-site.xml

Is there anyway to ask the namenode, zkfc, journalnode and datanode to 
reload the configuration (rather than having to restart the daemon) for 
zero-downtime changes to the cluster?


---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@hadoop.apache.org
For additional commands, e-mail: user-help@hadoop.apache.org