Tableau Cloud vs On-Premise: Metadata Access Comparison

On-Prem vs Cloud: Limitations and Purposes

Before diving into specifics, it’s crucial to clarify that on-premise Tableau Server and Tableau Cloud have different methods of how we can work with metadata:

Tableau Server (on-premise)

• PostgreSQL repository: you can directly query many historical events, metadata, configurations and other data through a standard relational SQL database. It is not available in the Online version.
• REST API: fully supported for programmatic tasks (creating/deleting users, listing workbooks, downloading dashboard screenshots, etc.).
• GraphQL: offers a more flexible way to query data, allowing to retrieve only the required fields and nested relationships in a single request.

Tableau Cloud

• No direct repository access, but it has pre-aggregated published data sources having some of data from the repository available.
• REST API and GraphQL are still available for all the same tasks.

Since I manage an on-prem server, I had a chance to experiment with all three: repository queries for deep analysis of our data, REST API calls for automations, and GraphQL for some specific requests.