HIVE 2.1 in HDP 2.5

[#1]

I learned in the community that can enable HIVE 2.1 in HDP2.5 by turning on "Interactive Query" in Ambari web UI. My questions are:

1. Once enabled HIVE 2.1, can the cluster support HIVE 1.2 and 2.1 simultaneously?

2. Can Hive 1.2 and Hive 2.1 runs well in one cluster?

3. How Can I disable HIVE 2.1 later?

Thanks

[#2 · Accepted solution]

1. Both "Hives" will run simultaneously with different jdbc URLs (using zookeeper for discovery)

jdbc:hive2://node1:2181,node2:2181,node3:2181/;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2
jdbc:hive2://node1:2181,node2:2181,node3:2181/;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2-hive2

and Hive Server ports (10000, 10500)

2. Technically "Hive" and "Hive Interactive" (2.1) run well on one cluster (different ports and folders). Be aware that Hive 2.1 is Tech Preview and that of course the machines need to be powerful enough (e.g. memory)

3. You can turn it off where you turned it on (under Hive Config, Interactive Query)

[#3]

Thanks @Bernhard Walter I still have two questions

1. Is there shell command of HIVE 2.1 to connect to HIVE2.1 like hive shell in HIVE 1.2?

2. Can I setup SQuirrel SQL Client to connect to HIVE 2.1? Is there any reference ?

Thanks

[#4]

Hello @Huahua Wei yes, you can connect through beeline e.g: jdbc:hive2://yournode.yourdomain.office:10500/default