adding new datanode to running cluster

Question

If I am using a Hadoop cluster that consists of three nodes with Apache Hadoop version 2.7.1

(active name node nn1,stand by name node nn2 and data node n1)

and we decide to add a new data node to the cluster while it's running dn2 (we call that adding data node on the fly)

What is the best way to achieve that without stopping services on name node

Of course first of all we have to install Hadoop files on this new data node or we can copy them from name node, don't we ?

Possible duplicate of Is there a way to add nodes to a running Hadoop cluster? — jedijs
if you are using Ambari to manage your cluster, this is fairly straightforward (Hosts->Actions) — tbone
thanks jedijs ,yes it is duplicate of this question stackoverflow.com/questions/13159184/… — oula alshiekh
Ambari is a web UI to manage Hadoop clusters. As mentioned, adding a datanode is only a few clicks — OneCricketeer

shahin shahin · Accepted Answer · 2017-05-08T12:03:30

Install and configure the new hadoop node. Make sure that you mention the namenode IP correct in core-site.xml.j2

Then you can start the datanode using the following command

/sbin/hadoop-daemon.sh start datanode

This will add the node to the cluster.