Hi, the other day I got home and my server had lost power. Some problem had triggered the GFCI outlet I have it plugged in to. I believe that my UPS shut the system down gracefully, but when I fired the server up I noticed that my parity drive was offline. It didn't really bother me because I had a couple new drives sitting around for a capacity upgrade, so I replaced the parity, and then after a rebuild, the smallest drive in the array. I then formatted my old parity in my pc (the one that wasn't being detected), but when I tried to replace the last small drive in the unraid server, it still wouldn't show up in the main tab, or in system devices. It is still detected by the BIOS, my main PC, and worked previously as my parity drive, so I can't see why I can't get it to detect. I tried swapping SATA ports as I have 2 drives connected to the motherboard and 2 via a cheapo pcie adapter, but no difference.

Any help for a poor noobie?

Thanksgigamulch-diagnostics-20250308-1634.zip

Disk is failing to initialize:

 

Mar  8 16:10:36 Gigamulch kernel: ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Mar  8 16:10:36 Gigamulch kernel: ata2.00: failed to IDENTIFY (INIT_DEV_PARAMS failed, err_mask=0x80)
Mar  8 16:10:36 Gigamulch kernel: ata2: limiting SATA link speed to 3.0 Gbps
Mar  8 16:10:36 Gigamulch kernel: ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
Mar  8 16:10:36 Gigamulch kernel: ata2.00: failed to IDENTIFY (INIT_DEV_PARAMS failed, err_mask=0x80)

 

Sometimes a bad drive is still detected in the BIOS, but fail to initialize, though it's strange if you were able to format it in another PC, but if changing ports/cables doesn't help, it's likely a disk problem.

Hi Jorge,

Thanks for the reply. The drive ran fine in windows and had no errors in the SMART. It even performed well in a crystal disk check. How bizarre. I'll look for a replacement I suppose. 

Thank you!

Interestingly, I fiddled with the cable and rebooted the server again and now the drive is showing up, but is limited to 3 gbps and has errors in the short SMART test. The drive only has 42 days of total uptime so maybe it'll warranty. I have ordered a new one to replace it, which is kind of a bitter pill with refurb drive prices so high right now.

I chalked it up to drive failure, so I replaced it and rebuilt my array. Interestingly, I lost all the data on that drive even though it should have been protected by the parity. No problem, the data wasn't mission critical, so I set to work redownloading everything.

 

Now, I'm seeing problems occur again. I checked this morning and the new drive contents were emulated, and while it didn't have a red x the second array drive didn't seem to be working either. A quick reboot brought the second disk back up but the one I replaced is still emulated, although it is connected has temp, smart data, etc).

 

I only see one red error in the logs:

"Mar 16 08:00:26 Gigamulch root: ERROR: The filesystem has valuable metadata changes in a log which needs to Mar 16 08:00:26 Gigamulch root: be replayed. Mount the filesystem to replay the log, and unmount it before Mar 16 08:00:26 Gigamulch root: re-running xfs_admin. If you are unable to mount the filesystem, then use Mar 16 08:00:26 Gigamulch root: the xfs_repair -L option to destroy the log and attempt a repair. Mar 16 08:00:26 Gigamulch root: Note that destroying the log may cause corruption -- please attempt a mount Mar 16 08:00:26 Gigamulch root: of the filesystem before doing this."

but I can't really figure out what to do in xfs_admin.

 

What is it about my system that the drives hate? Granted my server is a laptop so I power the drives externally from a wall wart to sata power cable, but it's protected on the UPS so it shouldn't be cutting out. 

 

Attached are my new diagnostics. Thanks for the help

 

 

 

On 3/9/2025 at 12:56 AM, JorgeB said:

Disk is failing to initialize:

 

Mar  8 16:10:36 Gigamulch kernel: ata2: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Mar  8 16:10:36 Gigamulch kernel: ata2.00: failed to IDENTIFY (INIT_DEV_PARAMS failed, err_mask=0x80)
Mar  8 16:10:36 Gigamulch kernel: ata2: limiting SATA link speed to 3.0 Gbps
Mar  8 16:10:36 Gigamulch kernel: ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
Mar  8 16:10:36 Gigamulch kernel: ata2.00: failed to IDENTIFY (INIT_DEV_PARAMS failed, err_mask=0x80)

 

Sometimes a bad drive is still detected in the BIOS, but fail to initialize, though it's strange if you were able to format it in another PC, but if changing ports/cables doesn't help, it's likely a disk problem.

 

gigamulch-diagnostics-20250316-0808.zip

ATA errors with parity and disk2, check/replace cables for both, if issues persist, it could be the controller, since they are using the same one.

 

As for disk1, it also looks like a power/connection issue, and the emulated disk is amounting, so you can ignore the xfs error for now.

 

 

I swapped cables and still had random dropouts unfortunately. Thinking that it might be a hardware issue (3 drives hanging out of a laptop with an AliExpress a+e key sata card) I migrated my server to an optiplex which is thus far being reliable. Maybe laptop chipsets are just nerfed... Thanks for the help Jorge, I'll come begging for more if this configuration develops the same issues.