Setting up access for Model Registry in a RAZ-enabled environment

In a RAZ-enabled environment you need to set up the S3-Ranger policy by manually adding the machine user name in the S3 Ranger policy.

To set up the S3-Ranger policy, complete the following:

1. On the Models Registry Details page, find and copy the Machine User Workload User Name in the Machine User Workload User Name field.

   For example, in the following screenshot, the Machine User Workload User Name field contains srv_cml_env_machine_user_82a49. Copy the Machine User Workload User Name which is 82a49.

   

2. Go to the Ranger UI in the Datalake of the environment.

   

3. Depending on your environment, select cm_s3 (AWS) or cm_adls (Azure).

   

4. Go to the policy named all - bucket, path which controls the access to the object store bucket.

   

5. Enter the Machine User Workload User Name in the Select User field in the allow conditions section.

   For example, using the Machine User Workload User Name from Step 2, add the value which is 82a49.