You can replicate data on-premise to Google cloud with a single cluster. The
metastore must be running on the cloud. There is no requirement to run the HiveServer 2
on the cloud environment. You must have Infra Admin or
DLM Admin role to perform this set of tasks.
-
Select Policies and click Add
Policy. Select HIVE as the service in the
Create Replication Policy page.
-
Enter the replication policy name and description.
-
Click SELECT SOURCE and choose
Type, Source Cluster, and
Select Database.
-
Click SELECT DESTINATION and choose
Type and Destination Cluster.
-
Enter the Destination Database.
-
Provide the Hive External Table Base Directory path:
GCS://bucket_name/path
The external table base directory path cannot be changed
once the policy is created.
-
Select Cloud Credential from the drop-down.
 | Important |
|---|
If the target dataset is non-empty, a warning message
appears - Target dataset directory /xxxx/xxx is not empty.
You can proceed by selecting the supressWarnings check-box.
Opting to select the check-box overwrites the target location, considering
the conflict resolution between HDFS location and Hive External Table base
location directory. |
-
Click VALIDATE.
-
Once the validation is successful, click SCHEDULE.
-
Configure the job settings for the replication policy.
-
Click ADVANCED SETTINGS to set up the policy
queue.
-
Click CREATE POLICY.
The data replication process is enabled.
View job status from the policies page. Verify that the job starts and runs
as expected.
