Excessive Fan Speed - Likely hard drive related

[RichardTR]

Running an ML30 gen9 with 4x 1tb drives in raid 5. 

Over the last day or so the server has been extremely loud (its in a small office environment, single server. Used as DC and for file shares). The site is approx 2hrs drive each way and i'd rather not visit if I can help it.. (fuel is pretty expensive right now haha..)

Spent a while investigating and also searching for people who experienced a similar issue, on the advice found I ran some updates. Interestingly I then updated some iLo bits and about an hour later the client called to say it had quitened down. Although this didnt last. There are no entries in event viewer etc which would identify the issue.

I checked Smart Storage Administrator and it was reporting 3x drives around 22 degrees and 1x drive at 61 degrees. Interestingly the 3x drives do fluctuate to 21, 23, 24 etc but the one drive at 61 stays at 61 (with a max of 82!) 

iLo is reporting temp of 46 degrees

SpeedFan reporting 46, 

'Get-PhysicalDisk | Get-StorageReliabilityCounter | Format-List Temperature' was reporting 40 degrees. I've ran this multiple times over two days and same result. Never moving from 40. Im assuming its averaging the four drives? The case sensors obviously think its 46 and the drive itself thinks its 61 so I have no idea how its arriving at this number.

The one hot drive doesnt seem overly hot to the touch when compared to the others. Resource monitor is showing fairly constant drive activity but it should be shared across the drives so for one to read as being 61 whilst the others surround it are in the low 20's just doesnt make sense.

Any help or ideas would be appreciated. 

[BPSingh]

Greetings!

-> Please check if the drives are HPE branded drives or 3rd party drives. We have seen this issue with 3rd party drives/add on cards. 

-> Verify system inlet temperature/environmental temperature. Page#29 https://www.hpe.com/psnow/doc/c04834991?jumpid=in_lit-psnow-red  

-> Verify that all air baffles and required blanks, such as drive blanks, processor heatsink blanks, power supply blanks, etc., are installed. to ensure proper air flow within server.

-> Ensure thermal configuration is set to optimal cooling in BIOS. F9->BIOS/Platform configuration (RBSU) ->Advance Options->Fan and Thermal Options->Thermal Configuration

-> Ensure server running on latest firmware revision. 

If you see experience fan issues, please log a support case so that the logs can be analysed for thermal events. 

[RichardTR]

Hi, 

- These are Seagate drives. They havent been a problem. This ML30 was running fine with no excessive fan usage for the past month (since building it)

- Inlet temp is around 20 degrees

- The only thing I havent done yet is attempted a firmware update (because i'm not on site so cant resolve if anything goes wrong). This doesnt explain why everything was working fine though. If the firmware was out of date then this issue would have occured since the beginning but thats not the case. 

As I say the drive itself isnt actually hot so I think persuing temperature related troubleshooting will not help. Measured with a laser thermometer and all drives are around 23 degrees and not hot to the touch. 

[BPSingh]

Greetings!

Thanks for the details. 

I would still advise you to test with HPE branded drives and check once. As mentioned earlier, please do check the thermal configuration as well. 

To understand more on the issue, you can log a support case and share the AHS logs so that the issue can be further isolated. 

[RichardTR]

Thanks for the help but Im not really willing to go and buy HP drives when I have fully functioning drives already. These seagate drives were fine for the last month or so, this wasnt an issue and had no fan problems. Theres zero valid reason why non HP drives would cause an issue. Especially considering the issue only appeared recently. Also keep in mind that the high temp reading is only occuring with one of these drives. If it was an issue with it being non HP branded drive then it should occur for all 4 drives, not just one?

I can see in the iLo that one of the fans is constantly at 66%. 

Is there any way for me to see the date I installed things from the hpe support pages? Over the past few weeks I installed a lot of updates. And as I'm not on site I have no idea at what specific point the fan speed issues began. It may have been an update which caused this issue (perhaps changed the way temps are reported etc).

Also I can't update the firmware on the server because HP blocks firmware updates for anyone who doesnt pay them an extortionate fee... Locking updates which protect against security vulnerabilities behind a paywall is quite disgusting practice to be honest. 

[BPSingh]

I can see in the iLo that one of the fans is constantly at 66%. -- Please ensure that there is no issue with cooling inside the server. 

Is there any way for me to see the date I installed things from the hpe support pages? -- Few firmware flash details including BIOS, iLO should be listed in the Integrated Management logs. 

As mentioned earlier, to identify any other underlying issue with the fan noise, please log a support case and share the AHS logs. 

[RichardTR]

Hi, as I mentioned in earlier comments; the drive is not hot and internal temperature is normal. 

I have tried to log a case but the rep wanted to forward me to the quotation team to pay for it... It seems wherever I ask for basic help HP tries to funnel me towards a sales team...  Thanks for your time but not being able to access basic support or critical updates without paying is disgusting practice and I will definitely think twice before buying HPE kit again..

Cant even view the AHS logs myself... HP's war against right to repair is a disgrace. 

[RichardTR]

I've come to the conclusion this was definitely caused by an update from HP. 

So HP provide free updates which break stuff to then force people to pay for the updates required to fix it? (with no guaruntee that update will even work?)

Extremely devious and underhanded tactics from HP. 

[RichardTR]

HP are literally now holding my server to ransom. The last thing I need to test is update BIOS/SPP update, and I cant do that without paying for an expensive support contract, but now ive been told that because the server has been out of warranty for a couple of months I cant even pay for the warranty to get access to the download I need.

This is crazy. Easily the most infuriating experience ive ever had with a vendor.