VDI Tip #1: Track all the “moving parts” in a VDI environment

Nathanael Iversen, Director of Technical Marketing, Xangati

It is essential for you to track all the moving parts in a VDI environment. It’s not enough to just look at the hypervisor or just look at the storage or even just look at the separate management console. It is crucial to have the ability to track all of those moving parts together.

The Gartner Group have basically made the observation that in VDI you essentially need to combine: network, storage, servers, the guest VMs, and various other existing IT products and the problem could be anywhere when the end user has a problem. It’s often the case that while the problem could be anywhere, VDI takes the blame as it’s the newest part of the infrastructure.

Their critical observation is really that the end user expects the performance of a VDI Desktop to be the same or better than whatever the physical devices that they used to have on their desktop (whether it be a laptop or a tower). They expect that level of performance and in order to do that, VDI has been implemented as a completely different architecture.

In terms of architecture, it is challenging as with VDI, you are actually creating a dynamic environment. As each user logs in and logs out of the system, the architecture combines in real time to make that particular session happen.

With VDI, you have a wide variety of devices – everything from thin clients and phones to dedicated partner terminals to standard PCs and you have a mix of storage, physical servers, and network devices. You need to be able to see how these various devices connect and how the traffic is going across the network.

The interesting thing is that all of these elements are combined in real time to serve a particular user. Maybe it’s a laptop or a virtual desktop, etc. and that’s how their session is maintained and their user experience is defined by that particular architecture.

Just moments later, another connection will use these same items but potentially order them in a different way, attach to a different desktop on a different host and connect to a different storage the back end and still be able to function.

When you start piling on multiple systems like this, that are all defined at the time of login, and dynamically allocated, the whole architecture is built in real time to meet the needs of your user base.

So, in this environment, how do you determine which things are taking which resources and how do you determine, when there is a contention issue, how to resolve it? How do you know when you see a storage latency spike that affects a particular VM, what are the other items that may be contributing to that dynamically in real time.

And, that is exactly what Xangati provides. Because of the transformational architecture that puts virtual desktops in a position of talking out of the network through physical servers, using resources from the cloud, and talking to a variety of end user devices and being dependent on a whole cluster of connection brokers: active directory, DHCP, DNS and other IP services, it becomes necessary to have tools that can actually monitor that entire environment. You start multiplying that out by dozens of hundreds of client devices and the complexity becomes such that you need to monitor the entire environment.

Stay tuned for tomorrow’s Tip #2 where we zoom in on storage performance.