Node Replacement Steps
In this chapter
If in case you want to change the host machine of DataNode or NameNode or you want to delete a node from the host machine. Node Replacement steps are explained for :
NameNode Replacement Steps
- In non-HA cluster mode, you can not delete a NameNode as it may cause cluster unavailability.
- In HA-Enabled mode, If an active NameNode is deleted, failover process will switch standby NameNode to active NameNode. .
- Deletion of standby NameNode will change cluster to non-HA mode.
- To delete a NameNode, select the check box against the node and click Delete on NameNode page.
DataNode Replacement Steps
- Before a node can be deleted, it must be first decommissioned.
- When a DataNode is decommissioned, HDFS will copy all its data to other DataNodes(if any) and update the NameNode data.
Decommissioned node will no longer be valid for storing data, but will be a part of cluster until it is deleted.
- If DataNode is deleted directly, then HDFS will wait for 15 minutes (configurable) for that DataNode to come back and if it doesn't, it will re-replicate its blocks on other nodes(if available).
This may cause damage to data because there would be a possibility of some blocks to be present exclusively on one DataNode.
- To decommission a node, select the DataNode and click Decommission on DataNode page.
- To delete a DataNode from the cluster, select the check box against the the DataNode and click Delete on DataNode page.
Copyright © 2017 QueryIO Corporation. All Rights Reserved.
QueryIO, "Big Data Intelligence" and the QueryIO Logo are trademarks
of QueryIO Corporation. Apache, Hadoop and HDFS are trademarks of The Apache Software Foundation.