Open the server → Sensors tab.
Live hardware health read directly from the server's BMC over IPMI. Because it comes from the management controller rather than the operating system, it keeps working when the OS is down, hung, or not installed at all.
What's reported
| Panel | What it shows |
|---|---|
| Temperature Sensors | CPU, chassis and (where fitted) drive temperatures |
| Fan Sensors | Fan speeds in RPM, per fan |
| IPMI | Whether the management controller is reachable |
| Power | Whether the machine is powered on or off |
Reading them
Hardware rarely fails without warning, and this tab is where the warning shows up:
- A fan reading 0 RPM that used to spin is a failed fan. Temperatures around it will climb next.
- Temperatures climbing slowly over days usually mean dust or a failing fan, not workload.
- Temperatures spiking under load and throttling performance point at cooling, not at the CPU.
- Power showing Off when you expected the machine to be up answers "did it crash or did it lose power?" in one look.
TipIf a sensor looks wrong, open a ticket with a screenshot. We can dispatch a technician on site — that's the part of baremetal you can't fix remotely.
When sensors are unavailable
If the tab can't fetch data, the BMC isn't answering. Use Reset BMC on the server page to perform a cold reset of the controller.
NoteIPMI and sensor monitoring are unavailable for 1–2 minutes during the reset. The operating system and your services are not affected — resetting the BMC does not reboot the server.
If sensors stay unavailable after a BMC reset, the server may simply not expose IPMI. Open a ticket and we'll check it.