Skip to main content

Set up Hive

Who can do this?

You need your Hadoop administrator to run these commands.

Configure authentication and permissions for Atlan to connect to your Hive database and extract metadata.

Choose authentication mechanism

Atlan supports the following authentication mechanisms for Hive. Choose one and configure it according to the steps below.

Use basic authentication for development environments or when Kerberos infrastructure isn't available. For production environments, Kerberos authentication is recommended.

Create user

Hive doesn't provide a built-in command to create users. User creation depends on how HiveServer2 is configured (operating system, LDAP, or custom authentication). When HiveServer2 uses OS-level authentication, create a Linux user to connect to Hive:

sudo useradd -m atlan_hive_user
sudo passwd atlan_hive_user

Replace atlan_hive_user with the username for your Atlan connection. For other authentication backends (for example, LDAP), create the user in the appropriate directory or identity store. For details, see Setting up HiveServer2 in the Apache Hive documentation.

Grant read permission on objects

Grant read permission on objects with the following commands:

GRANT SELECT ON DATABASE <database_name> TO USER <username>;

Atlan requires read permission for all the objects you want to crawl in Hive.

Verify configuration

Before proceeding to crawl metadata:

  1. Confirm the Hive user has access to the necessary database objects (for example, SELECT on the databases and tables you want to crawl).
  2. Validate authentication using the keytab file: run kinit -kt /path/to/your.keytab principal@REALM and then connect to HiveServer2 with a Hive client to confirm the credentials work.

Next steps

Now that you've configured authentication for Hive, you're ready to: