How to remove a node in a Hadoop cluster?
To remove a node from the Hadoop cluster, you can follow these steps:
- Ensure that the node to be deleted has stopped the Hadoop service and there are no running jobs or tasks.
- Edit the Hadoop configuration files on the main node (NameNode) of the Hadoop cluster by removing the relevant configuration information of the node to be deleted, such as hdfs-site.xml and core-site.xml.
- Execute the command on the master node to remove a node from the Hadoop cluster’s master node.
hadoop dfsadmin -refreshNodes
- Execute the command on the main node of the Hadoop cluster to remove the data blocks associated with the node to be deleted.
hadoop balancer -threshold 0
- Execute the command on the main node of the Hadoop cluster to remove the node from the cluster.
hadoop dfsadmin -refreshNodes
- Execute commands on the main node of the Hadoop cluster to check the status of the cluster nodes and ensure that the node to be removed has been successfully deleted.
hadoop dfsadmin -report
- Finally, on the node to be removed, run the command to stop the Hadoop service and uninstall the node from the cluster.
hadoop-daemon.sh stop datanode
After completing the above steps, the node will be successfully removed from the Hadoop cluster.