freshidea - Fotolia
As a desktop virtualization administrator, you wear a lot of hats. At any given time, you could be working on your gold image, patching operating systems or applications, installing or troubleshooting a web server, carving out virtual LANs for an expansion of your deployment, rolling out thin clients or doing any of a dozen other jobs. But there is one thing that can divert attention from all of those seemingly important tasks: the user experience.
Desktop virtualization pros often forget that the people using the virtual desktops have to be happy because, well, they actually use it. (Raise your hand if you use your own VDI all the time. I didn't think so.)
To keep an eye on overall performance, you employ various platforms that provide monitoring. These can capture information from lots of data points, such as servers, network equipment, storage, hypervisors and even the desktops themselves. All that data is important, but it's all very clinical in nature. Much like how an MRI for knee pain might show nothing is wrong even though your knee still hurts, users don't care that there's nothing technically wrong with the VDI deployment when they call and say a certain application is "being weird."
One way to get a grip on the actual user experience is to employ a desktop user experience monitoring platform. UX monitoring helps you quantify user problems that are often very subjective. These tools come in many shapes and sizes.
The agent-based approach
With agent-based UX monitoring, an agent runs inside each desktop or application server and watches user activity, cataloging it over time. It knows, for example, when a user clicks on an individual message in Microsoft Outlook and how long it takes to load. It can compare that over time to similar operations while also looking at the other, more clinical, system-level factors.
A quiz on UEM tools for VDI
Prove your knowledge of user environment management tools with this quiz, covering VMware's built-in UEM tool, UEM alternatives and more.
With this approach, you can view user experience statistics regarding specific applications, user activities, servers or even geographic locations. So when a user calls and says Outlook is slow, you can look to see what he or she is referring to and view a snapshot of the environment as it was at the time the problem occurred.
Agent-based desktop user experience monitoring works best when it comes from the same vendor as your existing server or data center systems monitoring platform, since all the data can be correlated more easily. It's not a show stopper if they're different platforms, but it could mean more work for you.
The virtual user-based approach
One of the drawbacks of the agent-based approach is that it still uses data based on clicks and system information that happens within the virtual desktop or application. Unless you find an agent-based platform that runs on the endpoint and is aware of the applications in the remote desktop, you could be missing some key data points that have a huge effect on the user experience. That's where the virtual user-based approach comes in.
Virtual user-based monitoring works by running prescribed workflows from an endpoint and watching how long it takes to complete each operation. The software does this automatically, and there is typically one endpoint for each application server, VDI host system, or geographic location. The monitoring software establishes a remote connection to the system, logs in with a typical user account and observes how long it takes to complete the operation. When it's done, the process starts again and compares the information it collects against past samples.
Armed with this information, you can see how the user experience changes throughout the day and even predict when UX issues might occur. This type of virtual desktop user experience monitoring measures behavior through a remote connection, so you can also identify problems that might be caused by latency or other external factors affecting the communication between the client and the data center. Over time, you can use this information to identify trends and predict future problems.
Boiling it down
The agent-based approach provides insight into the specific users who are having problems, when they're having the problem, but this is typically limited to in-guest operations. These tools are amazingly perceptive at operations for a variety of apps, even going so far as to identify things such as overloaded graphics processing. Still, they often ignore the possibility of external influences on the remote connection itself.
Virtual user-based platforms give a system-wide view of the real-world user experience, so you can pinpoint whether the problem is specific to the user, network connection or a certain application. While you may not get the low-level information that you can get with an agent-based approach, you can certainly get that data from your overall systems monitoring platform.
Both approaches to desktop user experience monitoring have their merits, and in a perfect world, organizations would probably use one of each. If you must choose one, look at where you have gaps in your UX coverage today and start there. Your users will thank you.
Strike a balance between UX and security
How to ensure a positive UX with VDI
Take VDI UX up a notch with monitoring