I am new to Hadoop, and learning Hadoop.
My question is:
Let say i have one master (data node) and 3 slave (name nodes), Hadoop is configured and running perfectly with no problem.
After few months, if i want to configure another 2 name nodes, how that work.
My understanding is:
1 Make ready name node with linux OS
2 and there may be way to configure 2 namenodes in Hadoop (How i dont know yet)
3 After complition of task #2, hadoop will start using other 2 namenodes, means now hadoop is using 5 namenodes in total.
I will highlight the steps for configuring High availablity
we have to modify core-site.xml and hdfs-site.xml
Installation of Journal nodes
installation of ZooKeeper nodes and ensemble
We need zookeeper failover controller
we have to initialize the shared edits directory
bootstraping the new namenodes.
There is no secondarynamenode the standby namenode will perform the checkpointing.
hope this helps.