The course is part of this learning path
Azure Pricing and Support
Microsoft Azure is a comprehensive collection of cloud-based services and features, ranging from Infrastructure as a Service virtual machines to Software as a Service offerings, such as Office 365. Using Azure subscriptions, businesses can choose which Azure services and features they want to deploy in the cloud. Prior to deployment to Azure, it is important to know how Microsoft bills for Azure services as well as what levels of support are provided.
This course provides an explanation of Azure subscriptions, the different types of subscriptions that are available, and the options available that can be used to save money in an Azure deployment. The course then focuses on Azure services, how they are priced and, where applicable, what metering costs are associated with the service, and how it is metered. This course also covers the Service Level Agreements that are available for some of the Azure services and the different levels of support that can be purchased. Lastly, it covers the Azure service lifecycle, including public and private previews of new services and features, and how you can be notified when they become available for preview.
- Understand Azure subscriptions and subscription types
- Understand Microsoft’s Azure pricing
- Plan and manage costs associated with subscriptions
- Understand Azure support options
- Understand Microsoft’s Service Level Agreements
- Understand Azure’s service lifecycle
- Sellers or purchasers of cloud-based solutions and services
- Individuals preparing for Microsoft’s Azure AZ-900 exam
- Basic understanding of cloud computing
- Experience with Azure, while not required, will be helpful
There are three purchasing options for Azure products and services. Enterprise, web direct, and cloud solution provider.
The enterprise option (https://azure.microsoft.com/en-us/pricing/purchase-options/enterprise-agreement/) is when a customer enters into an enterprise agreement with Microsoft. The enterprise agreement essentially is an agreement to purchase a negotiated amount of Azure services. Typically paid on an annual basis, but there are other flexible payment options. In order to qualify for an enterprise agreement, a private business must have at least 500 users and devices, while public organizations must have at least 250 users and devices. Enterprise agreement customers enjoy savings of between 15 and 45% over the public Azure prices.
When customers sign up for Azure using the Azure website, this is known as Web Direct (https://azure.microsoft.com/en-us/pricing/purchase-options/pay-as-you-go/). The Web Direct method is a pay as you go plan billed monthly. Web Direct customers pay public general prices for Azure services. When using the Web Direct plan, review the additional options for savings, such as the Azure Reserved VM Instances, Azure Hybrid Benefit, and dev/test pricing. These features will be explained later in this course.
Cloud solution providers, or CSPs, typically Microsoft partner companies, are businesses that develop Azure based solutions for their customers. CSPs can enter into a cloud solution provider agreement with Microsoft. Billing and payments for Azure usage by CSP customers is then billed directly to the CSP.
While there are some resources in Azure that incur no monthly flat rate charges, like virtual networks, many resources do incur a charge based on their usage. This type of cost is called metering. Therefore, these are considered meter resources. There are a few factors that impact the cost for a given resource.
The first factor is the type of resource. The type of resource will determine the type of metering that applies to it. Metering tracks a type of a usage like bandwidth for network utilization, running time for IAAS virtual machines, or storage capacity for storage accounts. For example, Azure virtual machine metering costs depend upon the size of the virtual machine. The size determines the number of CPUs, memory, network adapters, and other resources in that virtual machine. Metering for virtual machines is time based, and charges occur only when the virtual machine is allocated resources. A virtual machine in the stopped deallocated mode would not incur virtual machine charges.
Azure is globally available, allowing customers to place their cloud-based resources where needed. Usually, close to the consumers. Usage costs may vary slightly between regions. As you can see from this chart, the cost for virtual machines per hour varies between just under 50 cents for West US two, to about 78 cents in Brazil South and East Asia. While it is recommended to host your resources in the region closest to the users of those resources, it may still be possible to use a nearby region that offers a slightly lower cost.
Network-based resources such as site to site VPNs and ExpressRoute incur bandwidth charges (https://azure.microsoft.com/en-us/pricing/details/bandwidth/). In most cases, the charges are for egress traffic. That is, outbound data. The exception being VNet Peering and virtual machines deployed in availability zones. Outbound data incurs charges after the first five gigabytes per month. And is charged per gigabyte on a tiered rate. As you can see from this table, the lowest tier for up to 10 gigabytes is 8.7 cents per gigabyte, and the cost decreases per gigabyte for larger bandwidth usage. Note that this table is for zone one. Outbound data transfer rates vary slightly based on the zone. There are currently four zones. Zone one through three, and DE zone one. To learn what zone your data center is in select a region from the drop-down box. The zone number in the table will change to reflect the zone number of the selected region, and will then display the outbound data transfer rates for that region and zone.
For the planning phase of an Azure deployment the price calculator tool is an excellent online tool to use in estimating your Azure costs (https://azure.microsoft.com/en-us/pricing/calculator/). This tool allows you to select and estimate the cost of deploying resources to Azure. Of course to generate the best cost estimate, you'll need to know exactly what resources you plan to deploy to Azure, and their compute and storage requirements. For example, Azure virtual machines are available in a range of sizes based on the number of CPU cores and memory. So if you plan to deploy IAAS virtual machines you'll need to choose from these sizes based on current demands of the workloads that will be hosted on the Azure virtual machines, and on projected growth. The bottom line is that a cost-efficient Azure deployment requires careful and extensive planning.
To use the price calculator, we'll begin by navigating to the price calculator website. We'll begin by scrolling down, and we'll add virtual machines to our estimate. I'll click Virtual Machines, and then we'll click View. I'll leave the region at west US. And the type set to the default. Windows operating system, and type OS only. But I'll change the size to reflect the size I need for my hypothetical deployment. I'll choose D4. As you can see, the cost is about 42 cents per hour. I'll change the number of VMs to four. Now I'll change the managed OS disks to standard SSD. I'll set the number of snapshots to 10. And I'll leave the storage transactions at their default of 100. Scrolling to the bottom reveals that my estimated monthly cost for these four VMs is approximately $1,747.
I'll click Purchase options. This presents me with the choices of either purchasing via the website, adding the purchase to an existing enterprise agreement, or I could purchase them through cloud service providers. I'll choose purchase through the website. Here I'm informed that I could save up to 80% of my purchase cost by using Azure Reserved VM Instances. Or Azure Hybrid Benefit, or development and testing pricing. I'll explain Azure Reserved VM Instances a little later in this course. The Azure Hybrid Benefit allows you to save up to 40% on virtual machines by using your existing Windows Server licenses. When I'm ready to make my purchase I can click Purchase now, where I'll be directed to login to my existing Azure account. If I didn't have an Azure account, I would be prompted to create one.
Another useful tool to assist in estimating the cost savings a business can realize by migrating to Azure is the total cost of ownership calculator (https://azure.microsoft.com/en-us/pricing/tco/calculator/). This tool will help you estimate your current cost for hosting your on-premises resources. Then it will display the cost savings over a three to five year period when hosting the same resources in Azure.
There are three main steps to using the TCO calculator: define workloads, adjust assumptions, and view the report. You'll start by inputting all of the servers and databases that you plan to migrate to Azure. Along with the storage and networking requirements. Then you can adjust the values of the default assumptions that the calculator makes to closely match the costs of your on-premises environment. Including storage, IT labor, hardware, and software costs. Then the TCO calculator generates a report that you can use to compare the costs of your on-premises infrastructure with that of hosting your resources in Azure, which were estimated using the price calculator.
To use the total cost of ownership calculator, I'll navigate to the URL. The first thing we need to do is define my workloads. So I'll use the total cost of ownership calculator to estimate the cost of hosting a small on-premises website consisting of four web servers, two database servers, and three file servers. I'll scroll down and I'll begin by clicking the Add Server workload. I'll enter 12 gigabytes for the amount of RAM. And I'll set the number of servers to four. Additionally, I'll set the number of processors to two for each server. And the number of cores per processor to four. This is to reflect the four physical web servers that I have in my on-premises environment.
Now I'll scroll down to databases and I'll click Add database. From here I'll choose Microsoft SQL Server. I'll enter four for the number of processors. And four for the number of cores per processor. And I'll set the number of servers to two to reflect my two node cluster. I'll set the maximum database size to 200 gigabytes. And the maximum number of concurrent logins to 400. Now I'll click Add storage. I'll set the storage type to SAN, of the three types. And I'll set the capacity to six terabytes. And the backup to four terabytes. And my archive to eight terabytes.
I'll scroll down to the networking section. And I'll set the network bandwidth to 100 gigabytes per month. Now I'll click Next. Now I can adjust my assumptions. I'll leave the software assurance coverage on for both Windows and SQL Servers, indicating that I have licenses for both of these products.
I'll scroll down. And I'll set my electricity costs to 15 cents per kilowatt. I'll leave the storage costs at their default, as well as my IT labor costs. Except I'll set the hourly rate for IT administrators at 35 an hour. Note the other assumptions. Hardware costs, software costs, electricity costs, and the like.
I'll click Next, leaving all the others at their default. Now I can view my report, and I can see that over five years my cost savings could be as much as 1.5 million if I moved all of my resources to Azure.
About the Author
Jeff is a technical trainer and developer residing in Arizona, USA. He has been a Microsoft Certified Trainer for the past 18 years, providing in-house development and training on Microsoft server operating systems, PowerShell, SQL Server and Azure. When he’s not developing and delivering courses on Azure, he’s photographing galaxies, nebulae and star formations from his computer-automated observatory in Chino Valley, Arizona using a 14” Schmidt Cassegrain telescope.