Support Questions

Find answers, ask questions, and share your expertise

How to use HBase after HBase HA

avatar
Super Collaborator

HBase HA documentation says:

"Once running two or more HBase Masters, HBase uses ZooKeeper for coordination of the active Master."

Can you give me more detail on how to use Zookeeper to coordinate two HBase master servers? Also what changes a client program needs to make, which was referring to the original HBase Master earlier?

For example: After making NN HA, we always use nameservice, instead of nn1 or nn2.

1 ACCEPTED SOLUTION

avatar

Manish Gupta in Hbase HA zookeeper cordinate to make one of the Hbase master active , and rest of them are standby. and we need to configure following properties in case of HA at client side too:

hbase.zookeeper.quorum

hbase.zookeeper.property.clientPort

View solution in original post

4 REPLIES 4

avatar

Manish Gupta in Hbase HA zookeeper cordinate to make one of the Hbase master active , and rest of them are standby. and we need to configure following properties in case of HA at client side too:

hbase.zookeeper.quorum

hbase.zookeeper.property.clientPort

avatar
Super Collaborator

Thanks @Deepak Sharma

How about Hive HA? Does Zookeeper do the same for Multiple Hive Server2 & Hive-Metastore? What changes do we need to make at client side for Hive HA?

avatar

Manish Gupta in case of hive ha also you can connect to HiveServer2 JDBC URL that contains zookeeper for eg: jdbc:hive2://<zk1>:2181,<zk2>2181/;serviceDiscoveryMode=zooKeeper;zooKeeperNamespace=hiveserver2,

this is explained very well in this doc , you can refer the same under Query Execution Path With ZooKeeper topic

https://docs.hortonworks.com/HDPDocuments/HDP2/HDP-2.3.6/bk_hadoop-ha/content/ha-hs2-service-discove...

avatar
Super Collaborator

Awesome. Thanks for the link.