In this article, we will explain how to remove Datanode(s) from the Hadoop cluster using Ambari in Hortonworks/Cloudera.

Decommissioning Datanode(s) from Hadoop cluster with Ambari:

Here we provide simple steps to remove Datanode(s) from the entire Hadoop cluster with Ambari in Hortonworks/Cloudera distributions in the Big Data environment.

Step 1: Login Ambari web-UI:

Basically, default username as admin and password also admin.

• Note: Before you need to do stop all services on that node.

Step 2: Once login Ambari server web-based. Goto dashboard then click on HDFS (Hadoop Distributed File System)

Step 3: Goto summary in HDFS dashboard then click on Datanodes.

Step 4: Select which Datanode you need to delete from the cluster.




Step 5: On the right side top, showing the “Host actions” button. Click on the Host action button then see the “Delete Host”.

Step 6: Click on “Delete Host”. Now it’s completely deleted from the Hadoop cluster.

The above steps are very simple to delete the Datanodes from the cluster for Big Data/Hadoop administrators.

Summary:

In the Hadoop eco-system, Datnodes commissioning/decommissioning is common. Sometimes if the Data node is corrupt, we need to delete Datanode from the cluster. Here we are using Hortonworks distributions for storage/processing large data. Normally, Hortonworks provides all services like HDFS, Hive, Spark, Sqoop, Kafka, and etc. In this scenario, we need to delete Datanode with simple steps for users/clients. First, login Ambari UI then clicks on HDFS after that select the Datanode. After selecting the Datanode, goto Host action then click on “Delete Host” from the cluster.




In the Big Data environment, adding/deleting the nodes are quite common. If you have large data you need to add additional data node, In case no external data or less data comes into the picture then we need to delete the data node from the cluster for the Hadoop administrator level.