Environments, Roles or Something Else?

We’ve got quite a bunch of machines here to which we want to deploy software. Usually those machines are not only associated with environments (QA/PROD/…) and “high level” roles (web server/database server), but are also associated with customers, specific programs we run for the customers or even a specific piece of software that we operate as part of a specific program that we run for our customers.
What would be the recommended way of setting this up in Octopus Deploy? Would we set up separate environments (e.g. QA for customer A, QA for customer B)? Or would we set up more fine grained roles (“web server for program X of customer A”)? Are there any other alternatives?

Hi MIchael,

Thanks for getting in touch.

The feature you are looking for is called Multi-tenancy, and its one of our main features for our next major release. It’ll besically add Tenants as another dimension to which you’ll be able to scope variables.

You can find out more about it in the “Multi-tenant deployments in Alpha 1” section of the link below

For the time being (Octopus 3.3), the best approach would be setting up separate Environments per customer.

Best regards,