How to Use ControlUp Live View to Identify Processes Causing CPU Spikes in VDI Environments

When attempting to diagnose CPU spikes and identify the specific processes responsible using ControlUp, the most effective approach is to utilize the Live view, also known as the VDI page in the ControlUp Web UI. This view provides real-time, granular insights into process-level CPU consumption across virtual desktop infrastructure (VDI) environments. Unlike historical reports, which show aggregated past data such as CPU Utilization Max but lack immediacy, the Live view updates metrics approximately every three seconds, enabling administrators to track CPU spikes as they happen and correlate them with individual processes.

Accessing this Live view requires either appropriate licensing or configuration that may prompt for a user principal name (UPN) login depending on how ControlUp is set up in the organization. When correctly logged in, administrators can drill down into specific machines experiencing CPU spikes and see a detailed, sortable list of CPU-intensive processes. This functionality helps pinpoint the exact processes causing resource bottlenecks, which is critical since the total CPU utilization of a machine does not always equal the sum of all visible processes. This discrepancy arises because kernel interrupts and other system-level operations that consume CPU are not attributed to individual user processes.

Users noted that if access to the Live view is initially unavailable or the relevant VDI tab not visible, it may be due to missing licenses or particular configuration settings controlled by the ControlUp administrator in the environment. In such cases, reaching out to the administrator and providing the UPN for authentication often resolves access issues. Without this, troubleshooting CPU spikes becomes difficult, and users may resort to external scripting solutions like PowerShell, which can be less reliable and less integrated.

Once logged into the Live view, operators can monitor real-time CPU usage per user and process, identify the top five consumers, and respond quickly to performance incidents causing a slowdown in the VDI environment. While the Live view is potent for immediate troubleshooting, storing long-term granular process data in a live manner is impractical due to data volume and cost, so ControlUp uses historical reports for trend analysis. Overall, the Live view is regarded as an essential and powerful feature in ControlUp’s suite for understanding and resolving CPU utilization spikes in virtualized desktop environments.

For further information on configuring and accessing the Live view in ControlUp, administrators can consult the official ControlUp documentation at https://docs.controlup.com and the ControlUp Academy at https://cuacademy.controlup.com. These resources cover licensing requirements, configuration steps, and best practices for effectively using real-time monitoring to maintain optimal VDI performance.

Read the entire ‘How to Use ControlUp Live View to Identify Processes Causing CPU Spikes in VDI Environments’ thread below:

I am trying to capture CPU spikes to understand which process was taking up all the cpu during these spikes. Is it possible to follow this with control up LIVE?

Assuming you’re on VDI. You can go into processes in the webUI and track the most heavy processes in your environment by sorting on CPU.

You can also filter down on a specific machine if you want to limit the amount of data.

where can I see this view? this is what I need. This is what I see when I click on VDI

I think if I can get access to this view I will be able to identify the processes much easier

Ah you might not have VDI configured. This isn’t my area of expertise though.

why am I not able to see any data under reports for example?

if I could see the data I might possibly see the processes that takes up all the CPU and identify the culpit

its the data for CPU Utilization MAX

Reports show you historical data. That might work for you but it won’t be live. You need the VDI section for that.
Why that top consumers box is not working I don’t know. But you can also click on the machine name and then drill down into CPU util spikes

if I drill down and click on the spike, I dont feel like it gives me the correct data? how is this 92% CPU

I want to know exactly which processes during that spike

I need to be able to trust the data otherwise the troubleshooting will go in circles

another spike on a different server and I cant understand which process is doing this so that I can optimize it or disable it

There could be a couple of reasons for this. We measure process CPU and machine CPU separately.
• There simply are alot more processes on that machine and those combined take up 92%
• If you take all processes and do a sum it might still not add up to 92%. Kernel interrupts and other system work are not counted in individual processes

this is the reason why I am not able to see any issues with the hosts, the spikes last a few minutes but they slow down the env for everyone on that VM.

ok

then I would need top 5, the 92% becomes irrelevant in that case

I’m not interested in combining all the processes

Then your best bet is to get access to the live view. It’s more granular then reports

Live view = VDI page

can I get access to it temporarily if we don’t have the license? just so I can troubleshoot and see if it actually helps me

right now I am forced to run powershell scripts outside control up because the data I receive from there is not something that I can trust

the troubleshooting is not moving forward

I have no idea if it’s a license or config issue. We’d have to wait for someone more knowledgeable about this

Hey Artiom, you should be able to access the live view after logging in. It prompting for a username is a configuration decision by who-ever set up ControlUp

Like him ^

ok so all I have to do is provide my UPN?

it worked!

I thought that this was a service that is not included in our license

no one in the company even knows that this works

Enjoy, you’re special now

indeed haha, this is like one of the most useful tools in control up in my opinion

I can see things live

It is, it gives you very granular insights. Which is not something we can maintain for a very long timeframe (simply to expensive to store). But the real-time view you’re looking at is hands-down the most powerful piece we have for VDI.

yea its really nice. Right now its quite calm because not that many users are logged on but I would like to learn how to use this as efficiently as possible to be able to understand what causes the high CPU spikes

sweet found a good way, using the machine agent I can see whenever a server jumps up to 90+% cpu live

thats a good start

xenserver doesnt tell me anything when using the performance tab

this was very very powerful

I can just click on the CPU graph and I can see which users takes up the CPU

LIVE

Most metrics update every 3 seconds. So yeah can’t get much closer to live

what I would like is to see which processes takes up that CPU when I click on the CPU graph under user

right now I can only see the logon duration etc

that would be really detailed info

I found it…

its perfect

Continue reading and comment on the thread ‘How to Use ControlUp Live View to Identify Processes Causing CPU Spikes in VDI Environments’.  Not a member? Join Here!