Edge node is the access point for the cluster and it is always good to have multiple nodes to load balance and also to maintain high availability in case one node goes down.
This can be done in multiple ways, by either using a dedicated hardware load balancer in front of Edge nodes or by setting a DNS round robin with a health check.
Make sure that the user has a running cluster with at least two Edge nodes with Hadoop installed. Refer to Chapter 1, Hadoop Architecture and Deployment, for Edge node and DNS configurations. It is presumed that the users are aware about the working of DNS and is from Linux Administration background.
Connect to the
client1.cyrus.com
Edge node.Check the DNS resolution of the node using the following command:
$ nslookup client1.cyrus.com
The preceding command will return an IP address, something like
10.0.0.11
.Check the resolution on the other Edge node
client2.cyrus.com
as well. It will also return an IP...