Environments and Deployments

We recommend to use a single production deployment of the Data Mesh Manager. That deployment can either be the managed SaaS offering or a self-hosted deployment.

Environments

Within that single production deployment, you can manage multiple environments (e.g., prod, preprod, test) by two ways: creating one organization per environment, or creating one organization and using the environment field at the output port.

1. One Organization per Environment: Organizations are the tenants in the Data Mesh Manager, and are highly isolated from each other. This is useful if you want to have strict separation between environments, and make sure that data products, teams, and users cannot request access to data outside of that environment. If you have the environments prod, preprod, and test, we recommend to create three organizations: acme, acme-preprod, and acme-test. Users can be members of all three environments, and switch between these organizations.
2. One Organization for All Environments: This is useful if you want to allow requesting access across different environments. If you have the environments prod, preprod, and test, we recommend to create a single organization with your company name, e.g., acme and use the environment field at the output port to distinguish between environments with the values prod, preprod, and test.

We do not recommend to create one deployment per environment, as the Data Mesh Manager supports multi-tenancy via its organizations.

Test Deployment

If you want to test new releases of the Data Mesh Manager before rolling it out in your enterprise, consider an additional deployment that is available for the data platform team. That test deployment acts as a playground for testing new features, and can be used to validate new configurations before applying them to the production deployment.

We do not offer a test deployment for the managed SaaS offering.