From the course: Introduction to Cloud Computing for IT Pros

Cloud infrastructure

From the course: Introduction to Cloud Computing for IT Pros

Start my 1-month free trial Buy for my team

Cloud infrastructure

“

- [Instructor] An integral piece to understanding how the cloud works is the topic of infrastructure, more specifically, full virtualization and paravirtualization. Let's check it out. Cloud computing isn't the same for everyone, and there are many different ways the infrastructure portion can be deployed. The actual infrastructure will depend on the application and how the provider of that application has decided to build that particular cloud solution. So, there may be big advantages to using the cloud from a business perspective. Maybe your business needs massive processing power, but you don't have the budget nor the desire to run it in-house. And then on the other side of the coin, if you require very little processing power, it might make sense to purchase and run your own dedicated server. The cloud is ideal for both scenarios. Let's begin with something called full virtualization. In this setup, a complete installation of software is run on a server, but is displayed on the clients. With this type of deployment, you can be running all sorts of unique applications, even different operating systems. This is one way to access services on the cloud where a remote data center delivers your services in a fully virtualized format. Now, full virtualization has been very successful where multiple users will be sharing a computer system, where users need to be isolated from each other and the control program, and where one type of hardware needs to be emulated on another machine. Now, paravirtualization differs from full virtualization in that it allows multiple operating systems to run on a single device at the same time. It does this by using system resources like processors and memory more efficiently. Now, unlike full virtualization where the entire system is emulated, paravirtualization works with an operating system that's been adjusted to work on a virtual machine. Not every element needs to be emulated with the paravirtualization model. And because of this, paravirtualization allows for better scaling than the full virtualization model, because it requires less processor utilization. So there are certain deployments where paravirtualization works best, like disaster recovery. If there's an equipment failure, for example, guest instances can be moved to another hardware until the equipment can be fixed. Migration is simplified and faster with paravirtualization, because guest instances can be removed from the underlying hardware, and because of easier migrations, capacity management is also easier to implement. Adding processing power and hard drive capacity is just so much simpler in a virtualized environment.

Contents