With flying Colors let us start a new blog based on Azure WVD.
Till date we have seen how we setup horizon, creating pools desktops etc On Premise. In this series let us focus on how we create desktops based out of Azure Cloud which is called Azure Windows Virtual Desktops/ Azure WVD.
In this blog let us see both the advantages of having an Cloud based solution as well as On-Prem with estimated pricing
Desktop Virtualization: Choosing Between Cloud or
On-Premise
Desktop Virtualization:
Cloud or On-Premise
Whether you’re setting up Thin Clients for users in an
office or provisioning Zero Clients for a remote
workforce, you’re going to need access to some powerful computing to run the virtual
desktop infrastructure (VDI) that host the desktop virtualization
environments.
You have two main options: in the Cloud or On-Premise.
Essentially, the fundamental difference between the two comes down to these
initial questions –Where is the server located and Who is
responsible for server maintenance?
Cloud Computing basically means that you’re using remote
servers in someone else’s data center and that they’re responsible for
provisioning and maintaining it. Cloud offerings can range from Infrastructure
as a Service (IaaS), which provides base access to hardware in the Cloud, to
Platform as a Service (PaaS), which includes an operating system, all the way
to Software as a Service (SaaS), where the Cloud company is responsible for the
entire application and its underlying technology stack. On-Premise, on the
other hand, means that you ‘own’ the machine and are responsible for everything
from keeping the software up-to-date all the way to the HVAC system that keeps
the servers cool. (Both of these options are applicable, whether your computing
environment is desktop virtualization or not.)
Cloud /vs/ On-Premise: Pros and Cons – Desktop
Virtualization and More
Picking the right option comes with understanding the
advantages and disadvantages of each. Here are five important areas of
consideration.
COST – Price point is a good first thing to
think about. In many cases, Cloud Computing will be cheaper than On-Premise,
because you don’t need to buy related server or storage hardware, fix it, or
pay an IT admin to maintain it. Most Cloud providers use a pay-as-you-go model,
where you pay a monthly charge based on how many resources you use. For smaller
organizations who want to leverage VDI desktop virtualization, this could be a
no-brainer, but larger enterprises should really crunch the numbers to figure
out whether it’s worth running desktops in the Cloud. One last note here is
that Cloud and On-Premise fit into different parts of the budget; while
On-Premise is largely CAPEX, Cloud falls into OPEX.
SECURITY – Security should be your next
biggest concern. In almost all cases, Cloud is more secure than On-Premise.
Most major Cloud providers can afford to spend more on cybersecurity than even
the most well-funded organizations, giving you access to talent and technology
that may otherwise be out of reach. On top of that, Cloud providers will patch
vulnerabilities at any time, meaning you don’t have to call IT in the middle of
the night, over the weekend, or when they’re on vacation. As a result,
companies that leverage Cloud Computing have an extra layer of defense against ransomware and
other attacks against their desktop virtualization environment.
SCALABILITY – In Cloud also has the benefit
of scalability. Scalability in Cloud computing is the ability to quickly and
easily increase or decrease the size or power of your IT solution. A scalable
Cloud is why you can sign up and use most Cloud solutions in just a few
minutes, if not seconds – sometimes, with just the push of a button. That is a
big gain for fast-growing companies, or those that have frequently shifting
requirements, whereas On-Premise requires a lot more set up and configuration.
LATENCY – Simply put, the further away your
endpoints are from their server, the more latency they will experience. This
can cut both ways, though, depending on where your desktop virtualization
employees work from. Office workers can benefit from the low latency of
connecting to a local On-Premise server, while remote workers may also
experience lower latency if the Cloud provider has a data center that’s
geographically closer than their company’s data center.
COMPLIANCE – Compliance is a final
consideration. If you’re in a strictly regulated industry like healthcare for
example, you need to make sure that your chosen Cloud Provider is
HIPAA-compliant. Some regulations require certain types of data to be carefully
protected, and On-Premise configurations can guarantee you meet these
requirements.
Why on-premise?
• CAPEX – often viewed as a negative, but a one-off, upfront expense can actually keep operating costs down.
• Security – keeping data securely on your own infrastructure, protecting your IP.
• Customisation – owning your own hardware and having full control over your choice of processors allows more scope for provisioning of resource to users. As opposed to pre-configured, cloud instances that don’t always provide the best performance for resource hungry graphical applications.
• Deployment – more control over deployment; the what, when, where and how are in your hands.
• ISV licensing – most ISVs do not take issue with this method and is considered no different to your current workstation licensing agreement. • Data – all your data is kept in one centralised store, when running alongside your current workstation environment.
• Designed to minimise latency. Why the cloud?
• OPEX – predictable ongoing costs. No surprise, large capital expense for hardware refresh or failure.
• Security – shifting the responsibility of security to the CSP can reduce nervousness. Smaller businesses benefit from enterprise grade security.
• Ready to use instances – instances that have been pre-configured with licensing and should be ready for use immediately. • Deployment – cloud solutions typically take less time deploy.
• Project-based deployments, spin up your environment then close it down, on demand.
Both solutions offer the key benefits of mobility, flexible working and centralisation of your data.
The costs
Below outlines the hardware and software required to run 24 concurrent users using Revit and AutoCAD in an on-premise VDI solution.
Required infrastructure (1 x server with 24 x virtual desktops)
• Server, processors, RAM, 2 x Nvidia GRID GPUs, SSDs
• Licensing – Microsoft Windows Server 2016/9 + Windows 10 and Citrix/VMware licenses • SQL express and Active Directory
• Remote access SSL VPN
• Nvidia GRID vDWS licenses
Financial Estimate
• Hardware £30,000 approx.
• Software £20,000 approx.
• Consulting / deployment services £15,000 approx.
Total approx. £65,000 one off, plus an annual renewal of licensing and support.
This does represent a substantial Capital Expenditure (even without failover) but how does this differ from a cloud-based solution?
With the cloud, there is no upfront hardware cost but there are still software and consulting costs, plus additional costs for extra provisioning such as file storage, data migration / bandwidth. Setting up an environment with a public cloud provider is still complicated, so you will need an experienced partner to ensure the deployment is a success.
In the tables below you can see the per user, per month cost of an on-premise solution compared to both a Desktop as a Service (DaaS) solution and a public cloud solution over a four-year period.
Each table gives an approximate cost to run Revit and AutoCAD. The servers in the DaaS offering are provisioned exactly as the on-premise servers have been and both solutions give 24/7 access to your users.
The public cloud offering is running on instances utilising Nvidia M60 GPUs. The cost has been calculated on a per instance, per hour cost for a 40-hour week, working 50 weeks of the year. For every extra hour worked on top of this, there will be extra cost.
In Our next blog let us have a look at how to get started creating a Microsoft account and getting an Azure subscription etc
0 Comments