Fan and temperature alerts are rarely random. Vendor documentation points to a short list of causes: a failed or unseated fan, a missing blank or air baffle, blocked airflow, a cover left off, or air that is simply too warm at the server inlet. This guide shows where to read the readings on each platform, what the common alerts mean and the checks Dell and HPE document, so you can separate a site cooling problem from a hardware fault.
Before you start
- Note the exact alert text and time from the iDRAC Lifecycle Log or the iLO IML.
- Check whether other servers in the same rack or room reported temperature warnings at the same time. If they did, start with the room, not the server.
- Have the service manual for your exact model: fan positions, required blanks and fan replacement steps differ by model.
Warning: Do not disable thermal protection to silence alerts. HPE cautions that a thermal event can damage server components when the Thermal Shutdown feature is disabled, and that the server should not run for extended periods without the required fans. Opening the chassis to reseat internal fans may require a planned shutdown; protect running workloads and back up first.
Step 1 — Read fan and temperature status
- Dell iDRAC9: System > Overview > Cooling shows fans and hardware temperatures; System > Overview > Cooling > Temperature overview shows the temperature probes, the inlet warning thresholds and Fresh Air compliance. From RACADM,
racadm getsensorinfodisplays the status of all system sensors (see RACADM health commands). - HPE iLO 6: Power & Thermal > Fans (called Fans & Cooling Modules on servers with liquid cooling) shows overall status, fan redundancy, minimum fan speed, thermal configuration and each fan's status and speed. Power & Thermal > Temperatures lists every sensor with its reading and its Caution and Critical thresholds. The System Information health summary also shows Fan Redundancy. On iLO 5 (Gen10) the menu names may differ; check the iLO user guide for your version.
HPE defines a hot spot as a sensor within 3°C of its Caution threshold, and notes that fan speeds can be high when a sensor is within about 10°C of Caution. HPE also says fan speeds above 60% are expected to be loud.
Step 2 — Understand the alert
| Alert | Documented meaning and action |
|---|---|
| Dell FAN0001 "Fan <n> RPM is less than the lower critical threshold." | The fan may be installed improperly or failing. Restart iDRAC to rule out a communication issue; if it persists, check for cable interference in the fan blades, reseat the fan in the same slot, swap it with a fan from another position, then replace it. |
| Dell RDU0002 "Fan redundancy is lost." | One or more fans failed, were removed, or the configuration now needs more fans. Remove and reinstall failed fans or install additional fans. |
| Dell TMP0120 "The system inlet temperature is greater than the upper warning threshold." | Ambient air is too warm. Check the event log for fan failures, verify the operating environment and site conditions, and keep iDRAC, BIOS and CPLD firmware up to date. |
| HPE IML "Fan failure (Fan %1, Location %2)" | Check that fans are seated and meet the server's fan configuration, that no ventilation issue exists, that all fan bays have fans or blanks, and replace non-functioning fans. |
| HPE IML "The system is overheating (Temperature Sensor %1, Location %2, Temperature %3)" | Verify that all air baffles and blanks (drive, heatsink, power supply) are installed, the correct heatsink is fitted and the correct fan type is installed. |
Step 3 — Check the physical causes
- Confirm the cover, air shrouds and blanks are all in place. On the R760, Dell's corrective action for the amber temperature indicator is to make sure no fan has been removed or failed, the cover, air shrouds and back filler bracket are installed, the ambient temperature is not too high and external airflow is not obstructed.
- Check that every fan bay has a fan or the required blank. On the DL380 Gen11, a single-processor configuration needs fan blanks in bays 1 and 2 and fans in bays 3 to 6; a dual-processor configuration needs six fans. One failed or missing fan causes a loss of redundancy; a second causes an orderly shutdown.
- Look for cables or other material blocking the airflow path, and for a hot spot near a specific component on the temperature page.
- Reseat the reported fan using the procedure in your model's service manual. Dell's R760 fan removal begins with the "Before working inside your system" steps (power off and disconnect), so plan downtime unless your model documents hot-swap fans. HPE notes that on servers with redundant fans, backup fans may spin up periodically to test themselves.
- Update firmware: HPE's first action for fans running faster than expected is to update iLO, system ROM and option firmware.
Step 4 — Read the front panel LEDs
- Dell (R760): the temperature indicator turns solid amber on a thermal error, such as an out-of-range ambient temperature or a fan failure.
- HPE (DL380 Gen11 Systems Insight Display): Fan LED amber with an amber health LED means one fan has failed or been removed; with a red health LED, two or more fans have failed or been removed. Over-temp LED amber with an amber health LED means a cautionary temperature level was detected; with a red health LED and amber power LED, the hardware detected a critical temperature level.
HPE describes the thermal policy as: above the Caution threshold, fans go to maximum speed; if the temperature stays above Caution for 60 seconds a graceful shutdown is attempted; at the Critical threshold the server is shut down immediately.
Step 5 — Cooling settings, and when to escalate
Both platforms let you add cooling margin, for example for third-party PCIe cards. In iDRAC9 use Configuration > System Settings > Hardware Settings > Cooling Configuration (Thermal Profile Optimization, Fan Speed Offset, Minimum Fan Speed). In iLO 6 use the Fan Settings on Power & Thermal > Fans to set Minimum Fan Speed or a Thermal Configuration (Optimal, Enhanced CPU, Increased, Maximum or Smooth Cooling); changing the thermal configuration requires an iLO reset. These settings do not fix a failed fan.
Escalate to vendor support when a fan still reports failure after reseating and swapping, when temperatures rise with normal room conditions and all blanks in place, or when the server has already shut down on a critical temperature. Collect a SupportAssist collection or an Active Health System log first.
Frequently asked questions
Why are the fans loud when nothing has failed?
Fan speed follows temperature. HPE notes speeds rise when a sensor is within about 10°C of its Caution threshold and that speeds above 60% are expected to be loud. Check missing blanks, baffles and inlet temperature.
Can I keep running with one failed fan?
On the DL380 Gen11 one failed fan only removes redundancy, but a second failure causes an orderly shutdown, so replace the fan promptly.
Checklist
- Alert text and time recorded; neighbouring servers checked for the same alert.
- Fan and temperature pages reviewed in iDRAC or iLO.
- Cover, shrouds, baffles and all blanks in place; airflow path clear.
- Reported fan reseated or swapped per the service manual; firmware current.
- Thermal protection left enabled; cooling profile changed only for a documented reason.
- Logs collected and a case opened if the fault persists.
Skyline can supply, configure and maintain Dell, HPE and Huawei servers under an annual maintenance contract (AMC). Contact us.
Comments
0 total · 0 threads