Timeout in hive (hadoop) after swoitching to use private endpoint storage
We have an aks cluster that has access to a storage account (we use managed identities to allow it to access the storage account)
The storage account has selected network access, we limit access from the aks node subnets. We have hive installed in the cluster and it is able to connect to the storage account.
When we disable the public network access (this is done via terraform) - we also create a private endpoint for the storage with an IP address from the same subnet as the aks nodes - hive is unable to access the storage account and we see timeout errors like the following:
9:45:40.260 [pool-6-thread-183] ERROR org.apache.hadoop.hive.metastore.RetryingHMSHandler - Retrying HMSHandler after 2000 ms (attempt 6 of 10) with error: javax.jdo.JDODataStoreException: HikariPool-1 - Connection is not available, request timed out after 30000ms.