SpyglassMTG Blog

  • Blog
  • Azure Virtual Desktop: Simplifying Pricing

Azure Virtual Desktop: Simplifying Pricing

Azure Virtual Desktop: Simplifying Pricing

License pricing

For those of you out there thinking of making the switch to a Virtual Desktop Infrastructure environment in a large or small setting, consider Azure Virtual Desktop. While Azure Virtual Desktop (AVD) is still in its’ early years, it is considered top tier amongst VDI technologies on the market. If you are strictly a Microsoft environment or just lean towards Microsoft technologies, then there is even more reason to go with AVD. While using other technologies instead of AVD is still completely acceptable, a one-stop shop is obviously less to manage, and not to mention, the pricing may be lower depending on your current license suite.

Let’s take a common license of Office 365 E3. I see this license mostly, so we will stick with this for the example. Office 365 E3 doesn’t come with the user access licensing, so an additional license is required. Please note that if you have any of the following licenses, then you are already licensed for AVD!

  • Microsoft 365 E3/E5
  • Microsoft 365 A3/A5/Student Use Benefits
  • Microsoft 365 F3
  • Microsoft 365 Business Premium**
  • Windows 10 Enterprise E3/E5
  • Windows 10 Education A3/A5
  • Windows 10 VDA per user

If you do not have any of the licenses mentioned above, then you have the choice to change your current Office 365 E3 license or to simply add an add-on license such as Windows 10 Enterprise E3/E5.

Hardware pricing

What’s required for hardware to access AVD is a great selling point as you can use a laptop running Windows, Android OS, or MacOS using the RDP client or a browser. The devices can either be corporate owned or BYOD and can be restricted as required. AVD can also be accessed using thin clients. So, if you’re looking to cut costs, a thin client can typically help there. Hardware also requires minimal resources to access AVD since the workload will be on the virtual machines, not the physical devices.

Compute pricing

Compute pricing can be high or low. This is going to be entirely dependent on what applications are required to run and what the workloads are for each user. For instance, I can run 3 users per Standard_D4as_V4 machine accessing Microsoft suite and browsers. You may be able to push this to 4, but when you start throwing in Team’s meetings with video and screensharing, it can bog down the machines. If you have users using more resource intense programs such as database entry apps or CAD programs, then you would want to increase the compute and decrease the users. For instance, 2 users would work well on a D8s_v5 or similar. You can get more sizing guidelines from Microsoft here - https://learn.microsoft.com/en-us/windows-server/remote/remote-desktop-services/virtual-machine-recs, but I found these guidelines not to be the most reliable. Some seem to be oversizing the compute per user ratio, some under sizing. Again, this is all down to what apps and workloads each user is going to use and whether you want to do pooled or personal instances.

Once you decide on the compute types and sizes, you now have the decision of how to optimize the costs for these machines. You can implement scaling policies to shut down and startup machines based on schedules and/or usage, or you can use Azure reserved instances and keep the machines on 24/7. Both methods will save a lot of money in the end and should be decided based on what you find suitable for your environment. Just remember to follow some simple best practices:

  1. Size your VMs appropriately. Don’t oversize and pay for more compute than you need.
  2. Use scaling policies where possible keeping Azure reserved instances in mind.
  3. Use Azure reserved instances where possible keeping scaling plans in mind.
  4. Monitor all Azure resources to adjust accordingly to reduce unnecessary costs.

Storage pricing

Storage pricing can be on the low end or the high end depending on the needs of the company. If you are looking for high performance and low latency for intense workloads, Azure Netapp Files would be what you are looking for. Azure Netapp Files charges by the provisioned amount of space. The minimum provisioned space used to be 4TiBm, but early this year it was decreased to 2TiB, which makes a big difference for customers with less of a footprint in AVD. If this is still too costly, you can use Premium Azure File Shares for all your needs. Azure File Shares has less performance than Netapps on each equivalent tier, but it comes down to how AVD will be used. If you don’t require and high-performance workloads, then stick with Azure File Shares.

Network pricing

Azure Virtual Desktop networking pricing primarily revolves around two key components: data transfer and networking resources. Inbound and outbound data transfers are subject to the standard Azure data transfer pricing. Typically, inbound data transfer is free and outbound data transfers are as mentioned, subject to the standard data transfer pricing. To optimize costs, keep in mind the region being used and the location of your users. If users are on the west coast accessing a US East region, then transfer costs will be higher. Generally, though data transfer costs will be minimal. Aside from data, networking resources themselves have a cost. This includes the VNet, Private endpoints, or VPNs depending on how the environment needs to be configured.

Conclusion

While it's ideal to consider factors beyond costs when making decisions, it's undeniable that expenses play a significant role in driving business choices. Therefore, comprehending the costs associated with Azure Virtual Desktop becomes crucial when determining the appropriate technology for your VDI requirements. By the end of this read, I hope you will have gained a clearer comprehension of the expenses involved and be better prepared for what lies ahead. If you have additional questions or need help deploying AVD, please contact us today!

Related Posts
Azure Bicep Tip of the Day:  Optional parameters become conditional property assignments
Azure Azure Tip of the Day:  Hidden tags give more meaning to your Azure resources