Application and desktop virtualization from Citrix has been around for decades. Despite predictions that software-as-a-service or competing virtual desktop infrastructure (VDI) providers would eat away at Citrix, it has remained the de facto standard for as long as I can remember. Prior to my career as an analyst, I spent many years in corporate IT, and Citrix was actually my first-ever certification; it was known as thin client then.
[For the record: VDI is virtualization technology that hosts a desktop operating system and distributes it to multiple workstations via a centralized server in a data center. The term was coined by VMware.]
Citrix virtual desktops aren’t for every business, but many verticals have relied on them heavily to enable workers to get tasks done securely. Once such vertical is health care, where the workstations-on-wheels (WOWs) that clinicians use are thin client terminals running Citrix VDI. Nurses rely on these to be able to look up and enter information at a patient’s bedside in real time.
Troubleshooting Can Be a Challenge
Despite the popularity and longevity of Citrix, the application can be very challenging to troubleshoot because the environment can be complex. As an example, a WOW is typically connected over Wi-Fi, and when Citrix isn’t working, it’s difficult to distinguish between network and application issues. Once a problem been identified, how does one then understand which specific Citrix apps are performing poorly? Another task that’s difficult with legacy tools is getting alerted if the performance of critical applications hosted on Citrix is poor.
These were some of the problems Northeast Georgia Health Systems (NGHS) was experiencing with its Citrix implementation. NGHS is a large health-care facility with about 15,000 employees in three hospitals, each with its own Citrix environment powered by about 350 servers to deliver services to approximately 5,000 Citrix XenDesktops. Citrix XenApps is also used to provide applications to about 80 remote and specialty sites where workers can use a laptop, thin client, desktop or a local WOW.
Kristien Kramer, an NGHS network engineer and team lead, agreed to be interviewed about the problems they were facing and how Nyansa’s AI-infused Voyance IT management platform helped overcome them. He told eWEEK one of the factors that has added to the complexity of the environment is its process of migrating to wireless. Historically, most of the devices were connected via a wired connection, but now they are about 70/30 wired and wireless.
Connection Timeouts a Steady Problem
I asked Kramer what kind of problems NGHS was having with Citrix, and he told me they had a lot of connection timeout problems, which causes the application to crash. Problems were occurring daily, which obviously has a big impact on productivity and patient care. This often left Kramer and his team scratching their heads as they try and determine whether the problem is with WiFi, the endpoints, Citrix or something else. Once the source of the problem is identified, fixing the problem is often straightforward.
What NGHS is facing isn’t unique to them. ZK Research has found that 90 percent of the time taken to fix a problem is identifying the source. Wi-Fi, in particular, can be very problematic. Another interesting data point from ZK Research is that many IT pros spend at least one day a week doing nothing but Wi-Fi troubleshooting. In this era where businesses are dependent on IT to drive innovation, this kind of time can’t be wasted.
In an effort to help understand the source of problems faster, the client-server team invested in Citrix Director, and while this was effective in helping understand what was going on within the Citrix environment, it didn’t help with the network or client. For those issues, NHGS relied on Wire Shark, which is essentially a massive data dump that requires separate tools for analysis. Both Citrix Director and Wire Shark are good tools, but using them leads to a scenario where client-server and networking each require their own tools—thus any kind of correlation needs to be done manually. Given the volume of data being generated today, no engineer that can connect the dots in the data fast enough to troubleshoot.
AI to the Rescue
This is the value Nyansa’s AI-based Voyance product brings. People can’t analyze huge volumes of data as quickly as an AI platform can.
Nyansa is able to quickly pinpoint the source of Citrix issues through the ingestion of ICA streams and the data from the OData API. Its analysis of this data is correlated to the rich set of analytics produced from info gathered from endpoints and network infrastructure. Voyance is able to display both Citrix and network settings and information in a single dashboard. Nyansa recently added this capability to Voyance to specifically address the large XenApp and XenDesktop install base.
Prior to using Voyance, Kramer and team spent hours on the phone with the various vendors who comprise the Citrix environment only to have them finger-point at one another. Now they have the evidence to approach the right vendor and have them take responsibility in fixing the problem. I asked Kramer if he had an idea as to how much time Nysansa’s Voyance product was saving them, and his estimate was a whopping 20 hours per week per person. I was first skeptical of this number but Kramer explained that much of the troubleshooting time came from having to send team members out to remote locations to gather data.
IT environments have increased exponentially in complexity, and the rise of IoT, mobility and cloud isn’t making things any easier. IT professionals should take a lesson from NGHS and seek out machine-learning and AI-based tools to manage their environments better and troubleshoot faster. Manual correlation of data no longer works and AI is the game changing technology that IT has lacked for years.
Zeus Kerravala is founder of ZK Research and a regular contributor to eWEEK.