ProLiant ML350p Gen8 - System Fans constantly stepping up to 100%

[StevieeeeeB]

This school's ProLiant ML350p Gen8 Server, has been a godsend for some years, and now recently, over the last 6 months has changed to deafening anyone nearby with jet engine sound effects as the fans ramp up to 100%
iLo shows absolutely no issues, and everything A-OK ... including temperatures which for everything are nowhere near 'Caution' temps}
The server lives in a small Air Conditioned room at a sustained low temperature also.
Ive updated everything i could think of and performed a physical examination inside the server case, and can see nothing that may be causing said issue...
If i reset iLo (under diagnostics) the fans return to nearer their normal speed around 10% ... however this progressively steps up 1% at a time to around 30% and from there steps up in 4% speed intervals ... ultimately resulting in a deafening 100% fan speed on all 3 internal fans..
So loud that its disrupting teaching in the IT suite that the server room is within etc.
Any advice or support would be welcomed. 

[Torsten. · Accepted solution]

Sure the ILO is OK?

Please check this:

https://support.hpe.com/hpesc/public/docDisplay?docId=emr_na-a00048622en_us

An ILO with NAND issue may no longer control the fans, hence they speed up to the max.

[StevieeeeeB]

Our server was showing none of the signs listed as having a NAND issue, all status's were showing as healthy... despite this however ... i performed the local command line NAND reset, and my prayers have been answered ... running a smooth and steady 18-19% fan speed ... no more jet engines taking off... many many thanks @tor 

[Torsten.]

What ILO firmware version are you running?

[StevieeeeeB]

iLO 4 v.2.76 

Perhaps a bit over eager to declare victory ... as after a short delay the fan speeds are now ramping back up again... sigh... are there any other hidden nuggets of knowledge you can offer @Torsten. 

Ive updated everything i can locate, but have no support agreement to download the SSP 

[Torsten.]

You won't probably need the SPP. All components are free available except BIOS.

Check the lastest BIOS, it may be free too since it is a critical fix (due to Intel security issues).

However, check IML, ILO temperature overview, etc, there must be a reason like non-supported disks, defective sensors , etc.

[StevieeeeeB]

I tried updating everything I could think of previously, bios is running the latest version, no impact on the issue...
I’ve checked the IML logs, and reviewed the temp readings in iLo and all are stable at very low temps... nothing has been changed drive wise in something like 2.5 years, but for the first 2 years it was happy as Larry?

[StevieeeeeB]

@Torsten. 
I gave up on the initial problem ... however ... curiously ... i came in today, and rather than work in a icy cold room, i left the door open and turned down the air con to low ... and 2hours later, at a slightly cool room temp, ive just realised im not completely deaf and the server fans are maintaining a constant 20% speed? ... Has anyone ever heard of a server or sensor not liking being too cold and misinterpreting it as overheating? If anything since the issue started happening, the aircon power has been turned up, not down, and had no effect.. curioussss...

Any ideas what could be checked or looked at to rectify if the air con is the issue? I cant leave the air con off completely, or the door open realistically ...

[Torsten.]

The server can operate with an inlet temp between 10° and 35° C.

Lower or higher temp may cause issues.

I have seen servers with an inlet of 2° C already and very strange results - you may also see condensed water, which is very bad for a server.

[Mihai_S]

Hello,

Sorry to reopen this tocpic, but there is none other that I could find on google.

I have 3 HP servers (2x DL360p Gen8 and one DL380p Gen8) all in a rack in a room that is monitored with wireless temperature sensors (temp is between 10C - 16C).

Two of the servers have the fans between 9% - 20% exept one that is going to 88% - 94%.

The server in question DL 360p Gen8 has compatible HDD's WD10SPZX (consumer grade) no NAND error, not anything, the first week everything was silent and now it wont shutup like his brothers, even after ilo reset or full power unplug cycle.

No errors in any of the logs, even the one that has NAND error is silent.

Running latest iLO and bios version, any ideas? I'm going crazy with this issue in my homelab.

Kind regards,

Mihai S

[johncj79]

Mihai S

Did you ever find out the solution to this?

LIkewise if the issues is one of the hot swap hd's wouldnt the ilo log say so?

We have a Gen8 ml350p that has fans going 2 minutes full speed ,  30 seconds normal. 2 minutres full speed. cycle.

its in a school causing drama lol.  

Thank you guys!  you're all awesome!

[Mihai_S]

Hello,
In my case, the issue were the HDD's.
I found refurbished SAS drives to buy and the problem was solved.
Check every log and tab for what issue is in your case.

Kind regards,
Mihai