Basic monitoring is a staple of IT management. Server monitoring, storage monitoring, network monitoring and the list goes on—monitoring is a fundamental part of operating most IT infrastructures that have grown beyond a couple of servers. As companies increasingly incorporate virtual desktop infrastructure (VDI) into their IT infrastructure, the issue of monitoring specific to VDI becomes relevant.
VDI has a tendency to lull IT administrators into a false sense of confidence. After all, there were desktops before and when there were problems there was a help desk structure to address those problems. There was back-end virtualization infrastructure before (e.g., VMware®, Hyper-V®, etc.) and VDI is just being added as an additional workload to an existing infrastructure.
Whatever has been used to monitor the virtual environment before will still be available for VDI. The problem is that the marriage of end-user desktops with back-end data center infrastructures creates an entirely new beast.
With a stand-alone desktop, a storage failure meant that one person couldn’t work so a help desk solution for that type of problem makes sense. For VDI, however, a storage failure may mean that hundreds of people can’t work. On the data center side, teams typically have fairly extensive experience with the applications running on either physical or virtual servers, so they understand the typical workloads and demands on the infrastructure.
With VDI, the virtualization team may not be as familiar with the workload of the end-user, so demand for IT resources can be less predictable. When a new workload or priority occurs in an end-user group, it can be multiplied by hundreds or thousands of end-users and, in a very short time, overload the available systems.
While the actual management processes may seem fairly similar to how they were done before, in reality VDI monitoring has a fundamentally different dynamic. Successful VDI implementations must have the ability to monitor performance, capacity, and availability from the end-user experience to the back-end virtualization and storage resources—in context of each other. Having an early warning system for problems as they develop and a way to troubleshoot and correct them when they occur can be the difference between a successful VDI project and a disaster.
This aligns to what we hear from customers. Many companies have successfully completed an initial VDI pilot. In a pilot implementation, there is often more information about the workloads, more communication with the teams regarding which activities are current or planned, and more awareness and vigilance regarding capacity and potential problems.
Most companies struggle when they roll out a new VDI to a broader, less controlled production environment without an end-to-end monitoring solution in place. In this early phase of broad adoption, the newness of the activity combined with high sensitivity on the part of the end-users can turn a minor problem into a catastrophic failure (especially if the company executives are some of the users experiencing the problem).
Knowing how to design and build an end-to-end VDI monitoring solution is difficult. To help you better understand the challenges associated with end-to-end VDI solutions, SolarWinds® has teamed with Liquidware Labs™ and VDI expert Doug Brown of DABCC to present a webinar, E2E Visibility from Virtual Desktops to the Data Center, on Wednesday, February 20 at 11:30 a.m. Eastern/ 8:30 a.m. Pacific. Please register here to attend the webinar, which will be followed by an interactive Q&A session.