Microsoft Azure and What it Means for Your Business
Microsoft Azure has made waves in the virtualization and cloud computing space since its inception. It’s not the only player in the game, but it has quickly become one of the biggest and one of the best for many small and medium businesses. Azure rivals Amazon’s AWS offering in terms of scalability and scope, but occupies a different market space in some respects.
Companies or entities who primarily use Microsoft products will benefit from looking into Azure. If you run an MSP, this is going to be most of your clients. Most businesses leverage Microsoft products in some capacity, so there are advantages for many technical and non-technical companies in general. Licensing costs can eat the financial advantage some platforms have over Azure.
The purpose of this article is to explain the very basics of Microsoft Azure, what makes cloud computing work, and how it can be a boon to a business. This document will be somewhat technical, but parts of it will be accessible to end-users and clients to help explain some of the core concepts in a way that makes sense. Let’s go over cloud computing, cloud infrastructure and virtualization, Windows virtual desktop, serverless computing and more, the general cost, and what resellers offer.
Cloud Computing in a Nutshell
Almost everyone has heard of the cloud at this point. It’s not even correct to say that the cloud is the future; the cloud is essential for scalability. The cloud is a case study in the economics of scale in computing. The cloud is a culmination of computing knowledge, computing abilities, and the hardware and bandwidth to make it all work.
Pretty much anything you can think of in standard computing exists in the cloud. If you need a server, you can virtualize one. If you need a data store, you can spin one up. As long as it is something which can be done conventionally, there’s probably a cloud provider which does it. The joke in IT is that the cloud is just someone else’s computer, which while cynical, isn’t wrong either.
Cloud computing is just someone else abstracting the data center infrastructure required to run different infrastructures or systems into an offering which plays off of their scale. You’re paying a little more regularly to not have to deal with matching CPU architectures, balancing racks, and having to maintain equipment. Someone else handles that so you get to pay for exactly what you need and so you can scale without having to deal with the hassle. You trade risk for an amortized and averaged out cost. We’ll get more into these factors with the cost impact section.
You can think of it as looking to rent for a bit more per month than you’d pay to own in exchange for a less unpredictable experience and without having the same degree of maintenance. When something breaks, it’s the owners problem to fix it. The cloud hosting provider is responsible for maintaining the infrastructure, the backend that lets you work, and resolving internal issues. Tasks that used to take a whole team to maintain can be relegated to a single individual or even outsourced to an MSP to manage since the vendor handles the more painful parts.
Cloud Infrastructure and Virtualization
While most cloud platforms started with basic virtualization, they had to be able to adapt to and incorporate multiple types of infrastructure to be effective. Eventually the focus shifted from just being able to run “stuff” in the cloud to the cloud having to become an extension of your computing infrastructure. To put it simply, to have SaaS, you have to build the pieces which can allow PaaS, IaaS, etc. which can be repackage and resold as XaaS. Azure has this entire angle nailed, arguably better than basically any other provider (for most use cases).
Basics of Virtualization and Cloud Infrastructure
Feel free to skip this section if you’re already familiar with virtualization and cloud infrastructure in general.
Virtualization is the practice of running the equivalent of a specific computer system on a virtual machine (VM). Instead of installing the OS onto a physical computer, you install it into a controlled subsystem (which emulates a physical machine) running on a hypervisor. The hypervisor (HV) is the system which runs the operating system which maintains the virtual machine and the internal infrastructure to make it functional, but its health is independent of its virtual machines. If a virtual machine has an issue, the hypervisor shouldn’t be impacted (there are exceptions, ring -1 exploits, etc., but those are far beyond the scope of this article).
A VM should be a compartmentalization of the features it is supposed to abstract. The fewer features you have, the fewer moving pieces there are, and the less likely it is for something to go wrong. The HV shouldn’t run essential roles for a domain or similar (in most cases), and the individual VM’s can afford to do less. With a hardware setup, you have a limit based on the hardware, the cloud tends to be infinite for all practical intents and purposes.
To make a VM useful, there has to be infrastructure in place to make it function. The idea to virtualize came first, but the infrastructure is what makes it all work. You have to handle VPNs, routing, network setups, etc. to make a virtualized machine do anything. The work had been done, so it was only natural to expand the offerings to do more and more as it became more practical and refined.
The Advantage of Azure with Virtualization and Cloud Infrastructure
The easiest way to understand this growth is to look at the question: “Why do companies use virtualization?” Virtualization is a way to reduce risk and (usually) costs; you trade absolute efficiency for something which is less prone to incidents and accidents, but is easier to maintain. The easier the process, the easier the client will be sold on the whole system. This is the angle Azure has taken in general.
Azure isn’t the cheapest provider by any stretch, but they’re one of the easiest to migrate to (for Microsoft solutions), and one of the easiest to get going with. The basic Azure console is easy enough the average technician can figure it out with a little help from Google (mainly for differences between options). Resellers make it even easier by restricting options while simplifying controls.
Microsoft Azure is basically a one stop shop for setting up a virtualized environment. You can set up VM’s, set up the infrastructure they need, and so much more. Data stores, VLANs, etc. are all trivial to get working with Azure. Networking setup is a matter of clicking the right things instead of handling firewall appliances like some older providers. Azure even offers backups, advanced desktop options (DaaS), etc. If you can do it in a data center or with standard hardware, you can do it with Azure.
Windows Virtual Desktop
Azure’s primary Desktop as a Service (DaaS) solution is Windows Virtual Desktop. It is the spiritual successor to traditional terminal servers allowing each individual use to have their environment imaged out of a standard base system. You pick the applications and similar, the rest is handled by the system itself.
Windows Virtual Desktop allows you to cut down on resources for what kind of system you or your clients need to work. Pretty much any basic laptop or desktop is going to be good enough to work off of no matter what the client needs to do. You get the advantages of a terminal server environment without having the resources shared (unless you want to). Most services offering this type of virtual desktop system are going to use Azure as the backend due to licensing costs among other factors.
Windows Virtual Desktop is the culmination of all of the advanced features of Azure distilled into an offering. The process appears completely transparent from the infrastructure down to the end-user’s login. It takes the pain away from IT departments and MSPs while offering an improved user experience which can be quite cost effective.
Serverless Computing and More
Azure leverages the cloud to provide even more types of offerings which fit in where standard cloud offerings don’t. Serverless computing isn’t anything new, but it is much easier with platforms like Azure. Consumer AI and other offerings exist as well which can be used to build all sorts of new technologies, all in the same place.
Serverless computing is the practice of abstracting a program beyond the confines of where it will run. The system itself isn’t lacking of a server, but the design process and implementation is; you think outside of the traditional paradigm of being at the whims of the operating system and local elements in favor of using standard interfaces and abstract implementations of traditional operating system functions. Basically, your programmers are focused on programming and an authority (for the system employed) and the vendor (where it is hosted) maintains the overall security and other nuts and bolts for the rest of the process.
Azure isn’t just a platform for developing an IT infrastructure, but it can serve as a way to run applications and services, and tie them all together as necessary. If you’re dealing with .NET or C#, they’re one of the best places to move these applications. The initial migration can be a little painful, but the reduction in maintenance (especially with a Windows environment) is worth the cost.
Microsoft Azure Cost
Azure tends to be one of the more expensive cloud providers for many use cases, but it makes up for that cost with what it offers. When you’re calculating the cost of a cloud provider, you can’t just look at the cost of equipment versus the cost of the cloud. What all goes into the instantiation, maintenance, and service of your infrastructure?
The Total Cost of Ownership (TCO) is what really matters when shopping these solutions. For some IT shops, this means running everything on Azure just won’t make sense internally, and that’s okay. For most other companies which use Microsoft products, Azure has a lot to offer. The biggest issue with looking at Azure is figuring out how to compare the apples and oranges with traditional offerings.
Azure works on a basis of cost of computing. The more you use it, the more it costs. For virtual desktop environments, you can look at their estimates to get an idea of what it takes to run a virtual desktop environment. They offer a standard calculator as well to get an idea of what you’re looking at for each item. Keep in mind though, licensing and similar is included with most offerings.
How much does it cost you to set up and maintain a server? What happens when a drive blows or the power surges? How do you back it all up? How many people do you or your client keep on the payroll just to do basic administration of the system? Who do you have on call, and how much do they cost to support it around the clock? What is the cost of down time, and how long does it take you to get back up? Are these resources split between multiple departments or tasks which have their own unpredictability?
The deeper you dig, the more favorable the cloud is going to be for most business purposes. The cloud functions on the economy of scale, and while a huge MSP with thousands and thousands of endpoints might be able to compete, the profit margin to liability ratio may not be worth it. As cloud providers get bigger and better, the thin margins on cloud computing services get thinner for smaller shops.
Resellers
Resellers such as Nerdio and Crayon work to make administering Azure easier without necessarily costing more. They use their volume to get competitive rates which allows them to help you help your client or yourself. Azure offers so many offerings it can feel daunting without someone to show you what you need first.
Nerdio dives deeper into Azure and combines their expertise with the raw backend that Azure provides. They offer IaaS, DaaS, and ITaaS (among other things). You’re paying a bit more to get assistance doing the harder parts of Microsoft Azure. While Azure itself is straightforward, it can be overwhelming with the sheer volume of options and configurability it offers. Services like Nerdio aim to reduce the complexity for you and your client.
Crayon is similar to Nerdio in what they offer, but the devil is in the details and the implementations. Crayon can allow an MSP to offload many of the more mundane duties. The exact differences that make them each offer unique value is a bit outside the scope for this article. We work with both since both are subtly different with their implementations and the other offerings they provide.