- Managed DaaS
- Happy End Users
If you do a simple job search for VDI engineers, you will quickly notice this position is in hot demand. Companies are throwing good money at these skilled individuals and are having a hard time with retaining them. So what makes this position so sought after and a hard position to fill?
Companies are seeking VDI engineers because they’re seeing the value in centralizing desktops in the datacenter, which minimizes breaches by greatly reducing the amount of data leakage. With workloads that touch credit card information (PCI) or n care records (HIPAA), there is an even greater concern to ensure you stay compliant. A data breach can cost organizations millions or ultimately result in businesses closing their doors. Other negative effects of your business going through a breach include the reputation impact. No one ever wants to end up on the HIPAA breach report, as it could be devastating to your client base current and future.
While desktop virtualization is popular, it’s not always easy to achieve. Go to your favorite search engine and type in ‘VDI Failures,’ this topic is one of the most prevalent phrases you will see in regards to VDI. The issue stems from a misunderstanding of VDI and the intricate balance of skills and expertise required to build and maintain high performance infrastructure designed to support this specialized solution. Many failed VDI attempts can be traced back to a company deciding that it has some talented storage, networking and systems engineers on staff and that should be enough to tackle VDI. It’s not. With VDI you are going to need an individual that is a master in all three of these categories.
I’ve worked in the desktop virtualization industry for nearly a decade, as both a VDI Engineer specializing in a specific solution and as a VDI-Specialist Sales Engineer for a desktop as a service provider (Dizzion) that lets organizations implement virtual desktops without worrying about having the specialized infrastructure or expertise in house. Over my career I’ve learned what it takes to be considered a VDI expert and why you can’t turn to just any IT professional to tackle the job.
Why would a VDI engineer need to possess a skillset that covers storage, networking and systems but three experts in these field isn’t the same? Consider this scenario:
If your Exchange server is having performance issues and delivering email to inboxes two seconds delayed, odds are users wouldn’t even notice. However, if your desktop that now lives on a server in the datacenter is delayed two seconds and you are waiting for your text to catch up with your typing, you will have end users coming your way with pitchforks and torches. In all seriousness, if users are complaining about performance issues within virtual desktop environments where do you even start to look for the root cause? Is it storage? You just bought a really expensive SAN, so that shouldn’t be the problem. Your network guy says we have a 10GB network, so that can’t be it. Is it the servers the desktops are running on? Your systems engineer just bought new hardware, so what could be the problem? You can see real quickly finger pointing will start happening.
A VDI expert who has ultimate ownership over the virtual desktop environment is able to take a more unbiased look at the situation to pinpoint the issue without being the one “at fault.” The expert should also have a strong understanding of how virtual desktop infrastructure works and the unique performance demands on the hardware, something that is critical to maintaining high performance and troubleshooting.
To add more fuel to the fire, when you look at VDI, the application layer can be the most challenging. Which broker did you choose, was it Citrix, VMWare, Microsoft, Parallels or some other player in the space? If you have someone versed in VMware and you want them to deploy Citrix XenDesktop or vice versa, there is a steep learning curve. Each provider has various desktop delivery options, strengths and weaknesses and you’ll need not just a VDI expert but an expert in that solution to smoothly and successfully navigate the initiative.
You also have to think about persistent vs non-persistent delivery models. If you chose a non-persistent pool, how are you going to handle the profile data? Are you going to use the broker’s built-in software (VMware – UEM / Citrix – UPM) or look at a third party product like AppSense, UniDesk (now owned by Citrix), RES, Liquidware Labs, Immidio, etc… So now you have to know how to manage the profile management software, does it also need an outside database or does it use a Windows NTFS share? Did you just turn the software on so it grabs the entire Windows’ user profile or did you sit down and search every app and figure out what needs to persist and what does not? How are logins affected by this, did you just increase login times by minutes?
What about backups? Do you backup each virtual desktop or just the golden image? If you are a non-persistent shop, you will need to backup all the profile data as these profiles will bloat and get corrupted from time to time and need to be restored.
A VDI expert can properly plan for these multi-layered demands with an understanding of how everything will work together and its effect on the end user experience.
All of the above options most likely dictate which hypervisor, storage array and network protocol you will need. Most brokers work on any hypervisor, but if you want all the features, you may be stuck deploying the hypervisor from the broker vendor to take full advantage of the latest advances of the product.
Once you have chosen the hypervisor, which servers and storage array are you going to pick? Are you going with the traditional pizza box or blade server system with a SAN model, or maybe you are going with onboard hard drives in your servers and use a company like Atlantis or PernixData (now owned by Nutanix) for caching disks. Maybe you will choose a hyper converged infrastructure vendor that houses all the compute and storage built-in to their appliance. This is a big decision as you can get locked into a large capital investment and could be stuck in your current delivery model until your hardware depreciates, which for most organizations is 3-5 years.
After you have chosen your servers and storage, there are yet more decisions to make. Do you go all SSD drives, spindles or caching disks? Your desktop delivery model dictates how many reads vs writes a typical workload might be. Some VDI workloads might be 90:10 Write vs Read. Can your chosen storage array handle that type of workload? Of course then you have to figure out what protocol you are going to talk to the storage array with, NFS, iSCSI, FC, FCoE, etc.
Another thing to consider is which load balancer will sit in front of your VDI environment. If you choose Citrix’s broker, you are more than likely going to want to deploy a Citrix NetScaler for all the added integration, but you might be an F5 shop.
You can see how these decisions layer on top of one another and a lack of expertise or understanding in one area can have a massive negative impact on your overall VDI implementation. Likewise, a member of your staff that specializes in one area likely won’t have a full understanding of how that area truly effects a virtual desktop environment and may make a bad recommendation not fit for this particular use case. Or worse, you could be stuck in a constant power struggle between each of your area experts arguing over what solutions are best, stalling your VDI project at the very start.
With desktop virtualization, the job doesn’t end when the infrastructure is in place and the desktops are rolled out to end users. VDI needs constant maintenance and management.
Who’s going to make the decision when your broker has a pending update and you need to implement it or wait? You finally have everything dialed in and now you’re faced with a possible update. It’s much easier to have a single expert assess the situation to weight the new feature updates against would could break (especially when the release notes and compatibility notes are longer than your car’s owners manual!). You have to figure out all the components you are currently using and components that you are not and determine what will happen if you do upgrade. If you have included any third party products in your environment, you will also have to search to see if those products are supported or affected by the new updates.
The members of your team are busy, trying to get all the “co-owners” of the VDI together to review this potential upgrade and make a unanimous recommendation is a headache and will likely delay the decision considerably.
Finding the right skilled VDI expert that knows the ins and outs to all these various technologies will not only save time, but it will save money and prevent potential issues. For those companies that want to forgo this big headache all together, desktop as a service companies like Dizzion streamline this whole process. Outsourcing your VDI initiative means you get the benefits of desktop virtualization without the struggle, time investments and need for an (expensive) in-house VDI expert to design, manage, troubleshoot and maintain your own (also expensive) infrastructure to support VDI.
Feb 23, 2023
Chromebooks are an inexpensive end point device that are an excellent option for virtual desktop environments. ChromeOS enables IT to deliver secure devices, fast deployment, and cloud-first management. READ MORE
Apr 13, 2021
Dizzion extends AnyCloud global delivery with new Managed Desktops on IBM Cloudâ, featuring VMwareâ Horizon technology Earlier today, Dizzion announced ...READ MORE
Nov 15, 2018
As the year comes to an end, spend time auditing your virtual desktops and planning for the future so you can have a smooth, productive 2019. READ MORE