CPU monitoring and EC

User Support Hardware Issues
1

The CPU graph can be observed either with Task Manager or in the Panel

IMG_20231115_120011
IMG_20231115_1200113024×2419 1.44 MB

  • How do users make use of this information?

  • What is “suboptimal SMT scheduling,” for example, and how can it be optimized?
    Topic on CPU pinning might be referenced here]

  • Can anyone shed light on the cause that explains the following observations?
    I have an anomaly I am trying to get more information on which is indicated by high heat and high fan speed. [“Settings gear”–> Other → Sensor Viewer - in R4.2] CPU heat is supposed to trigger an EC sensor which turns on the fans. However, CPU graph monitors display that the CPU is only at around 1% but the fans are at the highest speed and the chassis is hot. I don’t want to imagine how hot 90% would be. Automatic shutoff before melting? Certainly no fire. Why is this happening and how should I go about troubleshooting the issue?

2

Task Manager only show you the CPU usage of the dom0 processes and don’t show the CPU usage of other VMs.
You can use xentop in dom0 terminal to see CPU usage of VMs.