Managing Multiple Clusters
Cloudera Manager can manage multiple clusters. Once you have successfully installed your first cluster, you can add additional clusters, running the same or a different version of CDH. You can then manage each cluster and its services independently.
Security (Kerberos) can also be implemented independently for each managed cluster.
To install an additional cluster to be managed by Cloudera Manager, see Adding a Cluster.
Cloudera Manager automatically creates the cluster name and the names of the services running in the cluster – for example, hdfs1 for the HDFS service instance in Cluster 1, hdfs2 for the HDFS service instance in Cluster 2. Role names are typically created by combining the role name and the host name on which it is running. For example, a DataNode role instance running on host node1.xyz.com would be named datanode (node1).
It is important to note the names of the services and roles, because in some areas within Cloudera Manager they are not organized by cluster, but rather just by service or role name. For example, on the Logs or Events pages, services are listed by name and are not organized by cluster, so you need to know which service instance is the one you want to look at.
On the Services page, if you view All Services, you will see each cluster in its own section, with the Cloudera Management services separately below. The Reports tab also shows the set of reports separately for each cluster. The Hosts page lists all hosts under management by Cloudera Manager, but also indicates the cluster to which each belongs.
The Activities tab lets you select which cluster's MapReduce service you want to see.