1. How and where you provision a virtual machine can have a major impact on your costs
There are many factors that go into decisions regarding where to place Azure resources. One factor that should always be considered is price. Azure data centers are built in different places, which logically have different capacities and operational costs. Azure pricing reflects these differences. For example, a virtual machine provisioned in the East U.S. region is approximately 18% less expensive than one provisioned in the West U.S. region. In addition to savings that can be realized based on location, Azure offers significant discounts for committing to their platform with “reserved” virtual machines. A reserved virtual machine is one that you commit to running for a certain number of years in Azure. Provisioning a three-year reserved virtual machine can result in a savings of between 60% and 80%. There is a cost to de-provisioning a reserved virtual machine early, but it isn’t excessive.
2. You can likely use the same network appliances in Azure that you use on premises
There are many vendors hosting their products in Azure. In the East U.S. region, for example, there are currently 2,460 virtual machine images provided by 745 different vendors. If your enterprise uses network appliances from Cisco, F5, Palo Alto, Barracuda Networks or many other vendors, it is likely that the same appliance is already available in Azure. In fact, many vendors have procedures for extending your on-premises network into Azure with their products.
3. Several Azure resources support high availability zones to protect against data center outages
When you commit to a cloud provider, you are putting your business in their hands. You want to be sure that the cloud provider is doing everything possible to take care of your resources. To that end, Microsoft has recently introduced availability zones for several services, including virtual machines, managed disks, SQL databases and more. Availability zones allow you to seamlessly host multiple instances of a supported service in different data centers. This provides availability even in the event of total data center failure. These zones are currently available only in select regions, but keep an eye on this vital feature as its availability expands.
4. Containers are everywhere in Azure
I started researching (okay, playing around with) containers about two years ago, and within a week I didn’t see too many reasons not to use them for everything. Apparently, I’m not the only one who feels this way, as Azure has container-based options for pretty much everything. Whether you want to manage your own hosts, let Azure manage a cluster of hosts, or just deploy a container to a fully-managed web app and let Azure handle the rest, Azure has a container solution for that. This is one of those technologies that have me thinking, “this can’t be this easy,” but so far, it really is. I am by no means a container expert, but I have built a custom image, provisioned a Docker host and launched my image on the host in under 30 minutes. Microsoft has done a lot of work with Docker, and it shows.
5. Azure has a great way to migrate virtual machines
It isn’t difficult to see the value in migrating workloads to the cloud. Reduced facilities costs and increased scalability and availability are fairly easy to quantify. One difficulty that typically arises is in the migration process itself. Migrating virtual machines can be especially challenging — moving with no data loss and minimal downtime is critical and difficult to plan and execute. Fortunately, Azure Site Recovery can dramatically simplify the process. Site Recovery can replicate entire virtual machines (as well as “bare metal” servers) into Azure and then seamlessly fail over to the Azure instance. This is typically seen as a disaster recovery capability, but it works well for migration scenarios as well. Site Recovery facilitates test failovers, so you can make sure the migration will succeed without interrupting normal operations.
Azure is rightly praised for its flexibility and scalability. With time and experience, people will discover other great features of the product, features that aren’t quite so obvious.