Multiple hive nodes in a cluster?

What is the use of having multiple hive nodes in a cluster as hive just acts a query engine over Hadoop

It can be installed in one of our client node, right ?

i tried reading docs but there was no information


  • Metastore and HiveServer2 are two different nodes already.

    Beyond that, the answer is simply for failover / high-availability, typically behind a load-balancer like HAProxy / Nginx, and using Zookeeper for a client quorum.

    None of these should be on client nodes, no, only beeline / JDBC drivers are needed on client machines.