Problem preclearing drives

[Medwynd]

I am having new hardware slowly trickle in from black friday sales etc, so I have been getting drives cleared as they come in.  I just finished 2 WD Reds and wanted to start the preclear on the other two.

 

Bios recognizes them when the machine starts but unraid does not see them.  I am using the trial version of unraid for now until I get my server all put together.  I unhooked the sata and power cables from the two drives that I cleared and hooked up cables to the two I want to clear.

 

Running preclear_disk.sh -l shows no drives to preclear.  The webgui does not show the drives either.

Running dmesg|grep SATA|grep link shows 2 SATA links up at 1.5Gbps

 

I'd like to get these preclears running because they took almost 72hrs last time and I feel like I've lost 24 hours so far with no clearing running .

 

Is this some sort of weird issue related to the 3 drive license of the trial version of unraid?  Some other issue that I haven't discovered yet?  Any other things I could try to fix this or to tryto diagnose the issue?

 

Thanks.

[RobJ]

Please see Need help? Read me first!, and attach the diagnostics zip.

[Medwynd]

Diagnostics attached.

 

Looking through the syslog I noticed this but have no idea what it means or what would cause it. WCC4N4JPZZE6 and WCC4N4JPZL7Z are the correct new drives.

 

Dec  2 08:48:43 CrystalPalace2 kernel: ata3: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Dec  2 08:48:43 CrystalPalace2 kernel: ata3.00: ATA-9: WDC WD30EFRX-68EUZN0,      WD-WCC4N4JPZZE6, 82.00A82, max UDMA/133

Dec  2 08:48:43 CrystalPalace2 kernel: ata3.00: 5860533168 sectors, multi 16: LBA48 NCQ (depth 31/32)

Dec  2 08:48:43 CrystalPalace2 kernel: ata3.00: model number mismatch 'WDC WD30EFRX-68EUZN0' != 'C WD30EFRX-68EUZN0                    €'

Dec  2 08:48:43 CrystalPalace2 kernel: ata3.00: revalidation failed (errno=-19)

 

Dec  2 08:48:43 CrystalPalace2 kernel: ata4: SATA link up 3.0 Gbps (SStatus 123 SControl 300)

Dec  2 08:48:43 CrystalPalace2 kernel: ata4.00: ATA-9: WDC WD30EFRX-68EUZN0,      WD-WCC4N4JPZL7Z, 82.00A82, max UDMA/133

Dec  2 08:48:43 CrystalPalace2 kernel: ata4.00: 5860533168 sectors, multi 16: LBA48 NCQ (depth 31/32)

Dec  2 08:48:43 CrystalPalace2 kernel: ata4.00: model number mismatch 'WDC WD30EFRX-68EUZN0' != 'C WD30EFRX-68EUZN0                    €'

Dec  2 08:48:43 CrystalPalace2 kernel: ata4.00: revalidation failed (errno=-19)

crystalpalace2-diagnostics-20151202-0851.zip

[RobJ]

Rather bazaar, initial setup of the drives is corrupted.  Two standard drives on the first 2 SATA ports, that shouldn't happen.  It's an old board, BIOS from 2009.  To eliminate the drives, try one of them on any other machine, see if it will be recognized, and can be quick-formatted.  More likely, it's a problem with this motherboard.  Try checking for a BIOS update.  Try resetting the BIOS, and check the BIOS settings for anything strange.

[Medwynd]

I'm away from the machine until tonight but I'm pretty sure that is the latest BIOS for that motherboard.

 

As I noted, BIOS sees the drives correctly with the correct model numbers during POST.  It is just unraid that seems to be confused.  I'll try to hook them up to an external bay I have on another machine tonight and see if they can be recognized.

[RobJ]

As I noted, BIOS sees the drives correctly with the correct model numbers during POST.  It is just unraid that seems to be confused.

 

Hmmm...  It's a standard Linux kernel, so I suspect any Linux system would have the same issue.  The unRAID modules don't kick in until later, once the kernel has built a new OS for that machine, and setup the hardware and drivers.  It's the kernel that is finding corrupt identifications.

[Medwynd]

Upon closer examination, bios is reporting the wrong model numbers during POST.  I'm going to try these in another machine and see what happens.  Otherwise is this probably an RMA for these drives?

 

EDIT: This was a misreading on my part.  It is reporting the correct model number like I thought, but when I looked at it quickly when I got home, I was thinking I should be seeing the serial number.

So post shows the correct model number.

These are the correct model and serial numbers.

Dec  2 08:48:43 CrystalPalace2 kernel: ata3.00: ATA-9: WDC WD30EFRX-68EUZN0,      WD-WCC4N4JPZZE6, 82.00A82, max UDMA/133

Dec  2 08:48:43 CrystalPalace2 kernel: ata4.00: ATA-9: WDC WD30EFRX-68EUZN0,      WD-WCC4N4JPZL7Z, 82.00A82, max UDMA/133

 

Why these model numbers are not lining up is the question.

Dec  2 08:48:43 CrystalPalace2 kernel: ata3.00: model number mismatch 'WDC WD30EFRX-68EUZN0' != 'C WD30EFRX-68EUZN0                    €'

Dec  2 08:48:43 CrystalPalace2 kernel: ata4.00: model number mismatch 'WDC WD30EFRX-68EUZN0' != 'C WD30EFRX-68EUZN0                    €'

[Medwynd]

Here is what I have tried so far:

 

• Reset BIOS to default
  
• Moved drive to another sas port with the same result
  
• Added drive to windows 8 machine and did a quick format on it, drive was recognized and partioned. 
  
• Ran crystal disk info to get information on the drive (screenshot attached) and it all appears correct.
  
• Added drive back to unraid machine and get the same results as my original post.
  
• Ran fdisk - l and it only shows my usb stick which was used to boot into unraid
  

[Medwynd]

Syslog from the last boot is attached

syslog.txt

[Medwynd]

So I think there is something funky going on with my other SATA connections.  Once I hooked up the other two drives to exactly the cables that were used for the previous preclear the drives were recognized.  A problem to sort out for another day, on to preclearing!