The goal of Azure Stack is not Infrastructure as a Service (IaaS), as there are better solutions in the market just running VMs. If you just want to run some VMs somewhere and not in your data center, you may even think of a Hyper-V cluster with Virtual Machine Manager as the management solution.

A cloud solution always starts to invoice itself and start gaining money if you think about more than a lift and shift scenario (which means migrating each VM on premise 1:1 to the cloud). Often, it even costs more than running the same VMs in your data center. As most of the cloud solutions provide a solution to power off VMs when you are not using them and power them on when you really need them, but it is that one you are really willing to use? Do you run VMs in your data center and power them on only if you need them? From a pay as you go model perspective, this will reduce the costs and drive your cloud solution to a more economical one from the business perspective. Costs will start decreasing when you start using multi-tenant solutions on a PaaS basis.

Platform as a service (PaaS) means that in regard to shared responsibility in cloud environments, you are not responsible for the VM and the OS; you are running your services (for example, a SQL database) on an existing SQL Server cluster. You are responsible for the database and the cloud provider is responsible for running the VM plus OS plus basis application (for example, SQL Server) itself:

The example SQL Server is a good way to explain even the business aspects a little more. In general, everybody knows how expensive a SQL Server license is and what resources you really need for a powerful database of an application. If you could just use an existing SQL Server with your database, that would cost lesser per database for sure, and if you think about your IT infrastructure, how many SQL Server are actually running with only a single database on it? How many resources could be set free if you would start using these databases on a single SQL Server cluster?

If we take a step forward and move this story to a private cloud environment with hybrid connectivity, you will have to rethink the story a little bit. If you are running the Azure Stack environment, it means that you will have to run the SQL Server environments yourself. If you are using an Azure Stack environment at a service provider, you would not have to run the environment as your responsibility. This needs to be considered when planning your hybrid cloud environment.

As Azure Stack owns the concept of resource providers, you will note that each additional resource provider will follow the same concept to interact with the ARM technology. In general, you will see that each new PaaS solution from Azure Stack will bring its own virtual machines (a minimum of one) with its own web service and some ARM integrations.

For GA, you will find three additional resource providers with PaaS solutions; these are as follows:

• Microsoft SQL resource provider
• MySQL resource provider
• App Service resource provider

On the following pages, we will describe these resource providers and discuss their installation, requirements, and features in depth. You will learn how to add them to your Azure Stack environment and which administrative processes are required.