Yes, it is the drives. HPE ProLiant Gen10 servers check the firmware identity of every installed drive, and any drive that does not report an HPE firmware string trips a fail-safe cooling policy: all fans go to maximum speed. Nothing is broken and no data is at risk. Here is what causes it and what your options are.
What Actually Triggers the Fan Ramp
Gen10 ProLiant servers (DL380 Gen10, DL360 Gen10, DL385 Gen10 and the rest of the family) run a drive validation pass through iLO 5 and the system ROM during every boot. The platform reads identification data from each drive behind the Smart Array controller and from drives it can see directly. If a drive reports a firmware string HPE does not recognize, the system logs it as an unsupported drive and switches the thermal policy to maximum cooling until that condition clears.
The logic is conservative on purpose. The Gen10 fan curve is driven by telemetry: the platform knows how to interpret temperature data from drives carrying HPE firmware. When it cannot map that data, it does not guess. It cools for the worst case, which in a 2U chassis like the HPE ProLiant DL380 Gen10 8SFF barebone with P408i-a means both system fans pinned at 100% from the moment the OS hands control back to the BMC.
The iLO event log wording is unmistakable. It reads something like: unsupported drives detected, all system fans temporarily set to maximum. The same entry shows up in the Integrated Management Log (IML) and in the health summary inside iLO 5, one line per offending drive, at every boot.
How to Tell It Is the Drive Check, Not Real Heat
Before you assume the worst, match the symptoms. This specific behavior has a clear fingerprint:
- Fans hit 100% shortly after POST and stay there at idle, with a cool room and a light workload.
- The iLO 5 event log shows unsupported-drive entries at each boot, right around the time the OS starts.
- No thermal alerts anywhere else: CPU and DIMM temperatures on the iLO dashboard sit in their normal ranges.
- Remove or replace the flagged drives and reboot: the fan curve returns to normal within a minute or two.
If the fans scream at full POST but the event log is clean, or CPU temperatures are clearly out of range, you have a different problem: clogged heatsinks, failed fans, or a bad sensor. That is a separate diagnosis.
Which Drives Pass and Which Trigger It
| Drive situation | What iLO 5 reports | Fan behavior |
|---|---|---|
| Drive already carrying HPE firmware | Recognized as supported | Normal, temperature-based control |
| Third-party drive with stock firmware | Unsupported drive entry in the IML | Fans to 100% after POST |
| Third-party drive flashed with HPE firmware | Usually recognized as supported | Normal in most reported cases, unit-dependent |
| Empty bay | No entry | Normal |
One nuance worth knowing: results are firmware-dependent, not brand-dependent. Some WD, HGST and Seagate enterprise models have HPE firmware images available, and units flashed with those images generally pass the check. Other models do not. This is why any statement about a specific drive has to be verified per unit and per firmware, not per brand.
Your Options as a Used-Hardware Buyer
Option 1: buy drives that already carry HPE firmware. The used market has plenty of pull-from-server enterprise drives, and units that spent their life in HPE machines usually still carry HPE firmware. When you source 10TB 7200RPM SAS drives or a 14TB SAS 12Gb/s drive, ask what firmware each unit reports before you plug it into a Gen10 chassis.
Option 2: flash HPE firmware onto a compatible third-party drive. For certain enterprise drive models this is a known, documented practice, and it usually silences the warning. It is model-specific, it can fail on some controller combinations, and it is not something to attempt blind. Verify the exact model against an HPE firmware list first.
Option 3: live with it. The ramp does not throttle anything. If the server lives in a closed room or a rack in a garage where noise does not matter, the unsupported-drive policy costs you only extra fan power draw and a recurring log entry. Many homelab buyers make exactly this trade.
Option 4: pick a platform without the drive check. Dell PowerEdge machines such as the R640 1U barebone and the R740 8SFF do not apply the same firmware validation to drives, so mixing third-party SAS drives typically does not move the fan curve. If you are still choosing a chassis and drive uniformity matters less to you than acoustics, this is a legitimate tiebreaker.
One practical middle path: many buyers run third-party drives in a non-HPE machine and keep the HPE chassis for drives that pass the check. Our own listings, including the Dell 2.4TB 10K SAS 2.5-inch drive and both 3.5-inch WD units above, are third-party units, so for a Gen10 build you should confirm the firmware situation per part number before committing.
Does the fan ramp hurt anything? No. Throughput, IOPS, latency and data integrity are untouched; the policy only changes fan duty cycle. The real costs are acoustic (a 2U Gen10 at full fan speed is loud enough to dominate a room), thermal power draw (noticeable on an electricity bill if you run several nodes), and log noise (recurring IML entries that can hide real warnings if you never clear them). If any of those matter to your deployment, deal with the drive firmware before the machine goes on the rack, not after.
For related reading on this platform, we covered what a DL380 Gen10 barebone is still missing to boot, and on the storage side how to spot a dying CacheVault supercap on a 9361-8i before it takes your write cache with it.
How Gen10 fan control works
- Gen10 uses a more conservative fan curve than Gen9, responding to ambient temperature, drive temperature and the fan cage configuration
- The controller holds the fans high when it is not confident about the sensor set it has
- Full speed at idle is a protection behaviour, not necessarily a fault
Third-party drives as the usual trigger
- The controller may not recognise a non-HPE drive and keeps the fans high as a precaution
- Drive temperature data from an unrecognised drive is treated as unavailable, so the curve errs conservative
- Third-party SSDs in a hot NVMMe bay can push the same response without any drive-recognition issue
What you can actually change
- Firmware updates sometimes improve the fan behaviour and drive recognition
- Ambient temperature and rack position matter more than most people expect
- On a used unit, confirm the fans are healthy rather than assuming the speed indicates a problem
Related reading
- DL380 Gen10 8SFF barebone — the platform this article is about
- DL380 Gen10 barebone guide — the companion article
- 7-point verification checklist — the thermal check detail
- Servers and barebones — category hub
Frequently Asked Questions
Why are my HPE Gen10 fans running at 100 percent?
Usually because the controller does not have a confident temperature reading. Gen10 uses a conservative fan curve that reacts to ambient temperature, drive temperature and the fan cage, and if drive sensors are unavailable the curve holds the fans high as a precaution. It is a protection behaviour, but it is worth checking the drives and the firmware.
Do third-party drives cause high fan speed on Gen10 servers?
Frequently, yes. If the controller does not recognise the drive, the temperature data is treated as missing and the fan curve stays conservative. This is the most common explanation we see for a Gen10 chassis idling at full speed. It is not a fault in the fans and it does not usually mean the server is overheating.
Is high fan speed a sign of a failing fan?
Not necessarily. A healthy fan responding to a conservative curve still runs at full speed. To tell the difference you need to check whether the speed tracks temperature. On a used unit we test the chassis on our own line and report the fan behaviour and the fan condition, so you are not guessing from a used purchase.
Can I reduce the fan speed in the BIOS?
You can adjust the fan curve, but it is worth being careful. Gen10’s default is conservative for good reason, and a manual curve that is too aggressive risks thermal throttling or an early shutdown. The more productive approach is fixing the underlying cause: recognised drives, current firmware and sensible ambient temperature.
Does updating the HPE firmware help with fan behaviour?
It can, because firmware releases frequently improve drive recognition and refine the fan curve. We re-flash during testing and state the firmware level each used unit ships with, so you know whether you are starting from a current image or an old one. Updating is worth doing before you judge the fan behaviour.
How do you verify a used HPE Gen10 server before shipping it?
We clean, re-flash and bench-test on our own test line, then run the seven-point check — inspection, branding, load testing, firmware validation, error logs, thermals and a documented burn-in. On this platform we specifically report the fan behaviour and the power-supply condition. Ask for photos of the actual unit and the test notes.