This post: Cloud Computing Part 4: PaaS (Platform as a Service)
In Part 3 we said that IaaS and Virtual Server providers basically take hardware out of the equation for the end user – it’s the IaaS provider’s responsibility to purchases, physically install and pay for. The hardware is then provided as a service – a virtual machine, with specifications that can be chosen, changed and billed by the minute. Your responsibility as the end user, is from the Operating System up. PaaS build on this, not only providing the hardware as a service, but the operating system and run time environment as well, which could include certain programs, including programming languages and libraries. The benefit of PaaS is that not only do you not have to worry about hardware purchase and maintenance, neither do you have to be concerned about administering an operating system, security controls or updates. This means you can simply write an application in a programming language and deploy it to the PaaS, enabling developers to concentrate on writing the code not system management. The downsides are that PaaS generally cost more, and they are less flexible, since you may not have access to the entire operating system that is running. Furthermore, you may have to write your code in a specific way for the PaaS, meaning it is harder to switch providers at a later time.
Below aare some PaaS providers, there is a good comparison of these at this page.
- Openshift by Redhat
- Google App Engine (note this is different to Google Compute Engine, its IaaS offering)
- AWS’s Elastic Bean Stalk
Returning to our example of a MS Word to PDF converter, the program could be the same as deployed in the virtual server example, without the hassle of setting up and maintaining the server itself.
In the final post in this series we’ll look at Software as a Service and consider the differences between, IaaS, PaaS and SaaS.