Hi, we had a brief blip in power the other night and the next day I noticed the server wasn't responding.

When I log into the admin console I see the following message next to the parity drive (also says "wrong"), and the array is not started:

"All existing data on this device will be OVERWRITTEN when array is Started".

Please advise best course of action to resolve this. I'm hoping parity does not need to be rebuilt.

Please post the diagnostics.

Diagnostics attached

nebula-diagnostics-20250624-1644.zip

Jun 22 21:28:33 Nebula kernel: ata3.00: HPA detected: current 39063648191, native 39063650304

Board enabled HPA, this is typically with some Gigabyte boards in particular, make sure o disable that in the BIOS, then remove the HPA, more info below:

Thank you, I will do this.

Would disabling and removing HPA be the only steps required to prevent this message?

If you disable it for the disk and the BIOS, so that it doesn't create it again, it should.

The hdparm guide in the thread you linked gives "The page you requested does not exist" so I searched and found some other examples in previous posts.

I ran hdparm -N /dev/sdb which gave:

/dev/sdb:

 max sectors   = 39063648191/39063650304, HPA is enabled

I then ran hdparm -N p39063650304 /dev/sdb which gave:

/dev/sdb:
 setting max visible sectors to 39063650304 (permanent)
 max sectors   = 39063650304/39063650304, HPA is disabled

I then ran the wrong command on the other drives. (I'm sleep deprived with a newborn baby). I was setting the HPA values on the other drives by mistake.

Here's the commands I ran and the hdparm checks on the drives after I realised the fuckup. The max sectors seem to look correct but I'm sure I've fucked up a setting somewhere, diagnostics are attached. What do I need to run to unfuck what I did?

root@Nebula:~# hdparm -N p39063650304 /dev/sdb

/dev/sdb:
 setting max visible sectors to 39063650304 (permanent)
 max sectors   = 39063650304/39063650304, HPA is disabled
root@Nebula:~# hdparm -N p39063650304 /dev/sdd

/dev/sdd:
 setting max visible sectors to 39063650304 (permanent)
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 40 01 21 04 00 00 a0 ff 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 11721045168/11721045168, HPA is disabled
root@Nebula:~# hdparm -N p39063650304 /dev/sde

/dev/sde:
 setting max visible sectors to 39063650304 (permanent)
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 50 01 21 04 00 00 a0 ff 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 7814037168/7814037168, HPA is disabled
root@Nebula:~# hdparm -N p39063650304 /dev/sdf

/dev/sdf:
 setting max visible sectors to 39063650304 (permanent)
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 40 00 21 04 00 00 80 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 40 01 21 04 00 00 a0 ff 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 40 00 21 04 00 00 80 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 11721045168/1(18446744072545694896?), HPA setting seems invalid (buggy kernel device driver?)
root@Nebula:~# hdparm -N p39063650304 /dev/sdc

/dev/sdc:
 setting max visible sectors to 39063650304 (permanent)
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 50 01 21 04 00 00 a0 ff 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 976773168/976773168, HPA is disabled
root@Nebula:~# hdparm -N p39063650304 /dev/sda

/dev/sda:
 setting max visible sectors to 39063650304 (permanent)
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 14 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 14 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 14 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 0/1, HPA is enabled
root@Nebula:~# hdparm -N /dev/sdb

/dev/sdb:
 max sectors   = 39063650304/39063650304, HPA is disabled
root@Nebula:~# hdparm -N /dev/sdd

/dev/sdd:
 max sectors   = 11721045168/11721045168, HPA is disabled
root@Nebula:~# hdparm -N /dev/sde

/dev/sde:
 max sectors   = 7814037168/7814037168, HPA is disabled
root@Nebula:~# hdparm -N /dev/sdf

/dev/sdf:
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 0a 04 51 40 00 21 04 00 00 80 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 11721045168/1(18446744072545694896?), HPA setting seems invalid (buggy kernel device driver?)
root@Nebula:~# hdparm -N /dev/sdc

/dev/sdc:
 max sectors   = 976773168/976773168, HPA is disabled
root@Nebula:~# hdparm -N /dev/sda

/dev/sda:
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 14 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
SG_IO: bad/missing sense data, sb[]:  70 00 05 00 00 00 00 14 00 00 00 00 20 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 max sectors   = 1159924272/1(120265081255028?), HPA setting seems invalid (buggy kernel device driver?)

The same message is present on the parity drive also after a restart.

nebula-diagnostics-20250629-0931.zip

You can ignore the missing/sense errors, that's typically when the controller doesn't support the hdparm command.

Parity still has HPA enabled:

Jun 29 09:16:28 Nebula kernel: ata1.00: HPA detected: current 39063648191, native 39063650304

Did you disable that function on the board BIOS? If not, it will create it again.

I didn’t this time, partly because it involves first getting up in the loft to dig out the wired keyboard and spare monitor, then going down into the crawl space under the house to get to the server.

But also because after reviewing those hdparm commands, I seem to remember I had this previously years ago and disabled the BIOS feature then.

Is it possible for this to re-enable itself? If so I will make the time to get down to the server and explore the BIOS again.

20 hours ago, gooner_47 said:

Is it possible for this to re-enable itself?

It shouldn't without it being enabled in the BIOS.

Interesting, thanks for confirming. I definitely remember checking for this before and I'm sure I disabled it, but it's been years so it's hazy. I will re-check the BIOS.

Bloody gigabyte board! Took me forever to find this setting - because the first time I got into the BIOS it wasn't even visible!

I had to go into "Dual BIOS/Q-Flash Utility", changed some settings in there, rebooted, and then the options available to me in the advanced bios screen changed and showed the "Backup BIOS Image to HDD" I wanted. You can see one screenshot of the advanced screen has fewer options, and then a later one more magically appear! Can't wait to build a new server and get rid of this hardware.

Anyway, error is now gone and the array is started again. Just wanted to say thanks for your help, and I've attached the screenshots of the BIOS settings in case it's useful to others in future.

IMG_7546.HEIC IMG_7554.HEIC IMG_7555.HEIC IMG_7557.HEIC IMG_7565.HEIC