Configuring HMS for high availability
To provide failover to a secondary Hive metastore if your primary instance goes down, you need to know how to add a Metastore role in Cloudera Manager and configure a property.
Multiple HMS instances run in active/active mode. Load balancing is done at the Hive
client side (like HiveServer or Spark) as the HMS client picks an HMS instance
randomly. By default, the
hive.metastore.uri.selection property is
set to RANDOM. If that HMS instance is down, then the client randomly picks another
instance from the list of HMS instances specified through the
- In Cloudera Manager, click .
Take one of the following actions:
- If you have a cluster secured by Kerberos, search for Hive Delegation Token Store, which specifies storage for the Kerberos token as described below.
- If you have an unsecured cluster, skip the next step.
org.apache.hadoop.hive.thrift.DBTokenStore, and save the change.Storage for the Kerberos delegation token is defined by the
hive.cluster.delegation.token.store.classproperty. The available choices are Zookeeper, the Metastore, and memory. Cloudera recommends using the database by setting the
- In Assign Roles, in Metastore Server, click Select Hosts.
- In Hosts Selected, scroll and select the host that you want to serve as the backup Metastore, and click OK.
- Click Continue until you exit the wizard.
Start the Metastore role on the host from the Actions
hive.metastore.urisproperty is updated automatically. To verify, go to
/etc/hive/configdirectory in your cluster node and look for the updated property in the
To check or to change the
hive.metastore.uri.selectionproperty, go to in Cloudera Manager, and search for 'Hive Service Advanced Configuration Snippet (Safety Valve) for hive-site.xml'.
- Add the property and value (SEQUENTIAL or RANDOM).