Temperature alert

[SirHaschke]

Hi,
we have two identical Fujitsu Hosts in different server rooms. On 1 ESXi host we get an alert:
Temperaturstatus from hosthardware, see screenshot.
If we look to IRMC all sensors are green. 
Is it possible to adjust this alert?

Regards

Dennis

[SirHaschke]

I found this post...
https://community.broadcom.com/vmware-cloud-foundation/discussion/host-hardware-temperature-status-1

Thats the question that i mean. Any ideas?

[Andrea Consalvi]

Hi Dennis,

If the iRMC shows all sensors in green but vCenter is still reporting a hardware temperature alert, there are a few things you can try to fix the issue.

First, check the sensors directly in vCenter under Monitor > Hardware > Sensors to see if any specific sensor is triggering the alert. Then, compare these values with those displayed in the iRMC to verify if they match.

If the iRMC data looks fine, try resetting the hardware sensors on the ESXi host. Connect via SSH and run the command:
esxcli hardware monitoring ipmi sel clear
This clears the IPMI event log. Then, restart the hardware monitoring service with:
/etc/init.d/sensord restart
After that, check if the alert disappears in vCenter.

If the issue persists, you can manually refresh the sensor data using:
esxcli hardware monitoring ipmi sel get
This will retrieve the latest sensor readings and might clear any outdated values that vCenter has stored.

If the alert is still there, you can adjust the alert settings in vCenter. Go to Configure > Alarm Definitions, find "Host hardware temperature status," and modify the thresholds to avoid false positives.

If nothing works, check for firmware updates on the Fujitsu support page, especially for the BIOS and CIM provider, since they manage how hardware data is reported to vCenter.

[SirHaschke]

Hi Andrea,
thanks for your answers.
Actually i update the Fujitsu iRMC and Bios Firmware.
Actually it look good :-)

One question.
You wrote:
If the alert is still there, you can adjust the alert settings in vCenter. Go to Configure > Alarm Definitions, find "Host hardware temperature status," and modify the thresholds to avoid false positives.

But i can't set a threshold...

Regards
Dennis

[Andrea Consalvi]

Hi Dennis,
Glad to hear that updating the Fujitsu iRMC and BIOS firmware seems to have resolved the issue!

[Andrea Consalvi]

Hi Dennis,

Glad to hear that updating the Fujitsu iRMC and BIOS firmware seems to have resolved the issue!

I see the confusion. What I originally meant is that you can modify how the alarm behaves in vCenter, but you can't set a specific numerical threshold for the temperature.

vCenter relies on the hardware's monitoring system, like iRMC, to report sensor statuses (Green, Yellow, Red). It doesn't allow you to define a custom temperature limit, like a trigger an alert at 70°C. However, you can still adjust the alarm rules to prevent false positives.

One way to do this is by modifying the alarm definition in vCenter. If the system frequently triggers warnings that aren't critical, you can disable alerts for the Yellow status and keep only the Red alerts active. This way, vCenter will only notify you when the issue is actually critical.

Another option is to adjust the threshold directly in iRMC. Since iRMC determines when a sensor transitions from Green to Yellow or Red, check if there's an option in its settings to raise the limit before it considers a temperature warning.

If vCenter's built-in alarms don't give you enough control, you could also create a custom alarm that tracks the actual sensor readings. This is a more advanced approach, but if your system provides temperature metrics within vCenter, you could use them to set a custom rule.

So, while you can't manually set a numerical threshold in vCenter, you can still influence when the alarm gets triggered by adjusting these settings.