Not all clouds are the same. In this video, you’ll learn about the differences between SaaS, PaaS, and IaaS.
When you’re deploying a cloud-based application, you also need to consider who will be accessing that application and from where. If you’re creating an application that anyone on the internet would need to access, then you would probably want to include it in a public cloud so that everyone can gain access. But if you’re building out a cloud-based application that is for internal use only, you may want to deploy this to your own virtualized local data center, where you have a private cloud. In many organizations, we often use both of these. And we need a hybrid configuration of both public and private clouds for all of these applications.
One type of cloud model is Software as a Service, or SaaS. This is commonly associated with software that is on demand. You have a login screen on your browser. You log in. And now you have access to the application. Someone else has written the application. Someone else manages the application. Your data is stored somewhere within that application. And you don’t have to do anything else to be able to use that particular app.
You don’t have a local installation of software. You don’t manage or upgrade any of the software itself. Instead, you simply log in and use the app. This allows for centralized management of the application and the data that you’re putting into the application, which means everything is being stored on a third-party cloud platform. This is a complete application offering. And you don’t have to do any type of development work to be able to use that app. A good example of software as a service would be an app like Google Mail or Office 365, where you simply log in and use the app.
There may be times when you would like to build your own application and deploy and manage your own data. And all you really need are the computing resources to be able to run that application. In that model, you want to use Infrastructure as a Service, or IaaS. Sometimes you’ll see this referred to as Hardware as a Service, or HaaS, because you’re effectively using hardware in the cloud and then using your own software on top of that.
This means that you would be responsible for installing your software and managing all of the data associated with that software. If the software needs to be upgraded, you would be the one to upgrade it on your infrastructure as a service model. The application that you’re using and the data that you’re putting into that application is still in the cloud. But you have much more control over the access and security of that data when you’re using infrastructure as a service. A good example of infrastructure as a service is with web service providers, where you can buy time on a particular web server. And they make that server available to everyone on the internet.
A middle ground between software as a service and infrastructure as a service is Platform as a Service, or PaaS. This is still in the cloud. But you’re the one who manages all of the development efforts to customize the application to your specifications.
The cloud provider gives you all of the tools necessary to build this particular app. You can think of this as a series of building blocks provided by the platform as a service infrastructure. And then you build out those blocks to create your own application.
The underlying engine that runs the application is the responsibility of the platform as a service provider. And your responsibility is to build and maintain the apps that are running in platform as a service. A good example of a platform as a service infrastructure would be something like salesforce.com, where they give you all of the tools necessary to create a customized application running on the salesforce.com platform.
Here’s an example of a cloud responsibility matrix that you might find with any cloud service provider. In this chart, I’ve separated out software as a service, platform as a service, infrastructure as a service, and then an on-prem device that you would manage yourself. Obviously, with on prem, everything from the data to the applications to the physical data center are all the responsibility of the customer.
But with infrastructure as a service, you have the physical data center, the physical network, and the physical host all provided by the cloud provider. And then you handle all of the other pieces of that particular application. With platform as a service, you can see that network controls, applications, and the identity and directory infrastructure may be shared between the customer and the cloud provider. But you’d still be responsible for the accounts, devices, and the information and data.
With software as a service, much more of these are provided by the provider themselves. For example, everything associated with the applications is clearly in the responsibility of the provider. Everything above that would still be the responsibility of the customer. So depending on the type of infrastructure you want and the type of responsibility that you’re willing to take on, you’ll need to pick the right cloud model for your application deployment.