[SOLVED] A little help with preclear, please...

[husky55]

I am using Unraid version beta 5.14 and preclear_disk.sh version 1.13. I just bought a Seagate USB3 3TB. After removing the drive from its enclosure, I proceeded to preclear it. What should be a couple minutes work has now turned into couple days of pulling my hair.

 

I used the command ./preclear_disk.sh -A /dev /sde    sde being the new Seagate drive in a USB2 dock. So I got an error, something about a SMART query failed. So I investigated and found that there were 3 partitions on the drive. Then I remember that 3TB drive needed to be formatted as GPT partition. Now I have a single partition of 3 TB. So I thought problem solved.

 

But no, preclear_disk.sh -l now showed the Seagate as sdf instead of sde as before. So I assumed that it was just the change in the GPT partition which caused that. But again preclear failed. Then I found out that the version I used is 1.13 but apparently there is a newer 1.6 version for beta 5.8 or later. Well I spent hours looking for 1.6 but apparently it was removed because I could not find it. Joe L did say to use 1.6 for later betas.

 

I check the Seagate drive with Seatools tests. It passed all the tests.

 

Any idea what I can do? I am afraid to just add the drive to the array and let Unraid do what it must do because I have lots of files that I would not want to lose. This is to replace a smaller drive which has my files on it.

 

here are some errors from my syslog:

 

un 19 07:50:34 Tower kernel: Buffer I/O error on device sde, logical block 195694143 (Errors)

Jun 19 07:50:34 Tower kernel: Buffer I/O error on device sdf, logical block 195694336 (Errors)

Jun 19 07:50:34 Tower kernel: Buffer I/O error on device sde, logical block 195694143 (Errors)

Jun 19 07:50:35 Tower unmenu[1089]: Precleared disk sde not found in drivedb[] (Drive related)

Jun 19 07:50:35 Tower unmenu[1089]: Precleared disk sdf not found in drivedb[] (Drive related)

[Joe L.]

You are confused somehow.  1.13 is the newest version of the preclear script.  preclear 1.6 was 7 versions earlier.

 

You must use 1.13 for the 3TB disk. (actually, for any disk greater than 2.2TB)

 

It will take 40 to 50 hours or more to clear the disk connected to an SATA port directly.  Over USB, it could take days.

 

The sdX devices are assigned by linux in the order in which the disks initialize and are found by the OS.  They can change from one boot to another, especially with a number of identical disks.  unRAID does not care what sdX device name has been assigned. 

 

Do not get hung up on a specific device name.  Instead, learn to confirm the identify of your disks by model/serial number.  Those will not change from one boot to another.

 

If you are seeing errors, it is as likely as anything the disk is reporting errors, regardless of the manufacturer diagnostics.

 

You may also want to make sure you have an updated versions of unMENU.  (press the Check for Updates button on the user-scripts page)

 

I would connect the disk directly to an SATA port.

use preclear_disk.sh version 13 to clear it, using the device it currently reports.    With a 3TB drive there is no need to specify -A, as that type of partitioning only applies to disks 2.2TB or smaller.  Specifying the "-A" option does no harm, but it is simply ignored as a GPT type of protective partition is put into place in the MBR for any disk over 2.2TB.

 

Joe L.

[husky55]

I am obviously confused as I have the latest preclear version already. However I just found out that even before any preclear activity is taken place, there is a SMART drive problem:

 

"Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)

 

A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options."

 

And then there is a bunch of errors:

 

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sdf, logical block 195694336

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sde, logical block 195694143

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sdf, logical block 195694336

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sde, logical block 195694143

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sdf, logical block 195694336

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sde, logical block 195694143

 

I will look for the newer UNMENU and will connect the Seagate drive to the SATA connection and see what happens.

[Joe L.]

I am obviously confused as I have the latest preclear version already. However I just found out that even before any preclear activity is taken place, there is a SMART drive problem:

 

"Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)

 

A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options."

 

And then there is a bunch of errors:

 

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sdf, logical block 195694336

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sde, logical block 195694143

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sdf, logical block 195694336

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sde, logical block 195694143

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sdf, logical block 195694336

Jun 19 09:12:59 Tower kernel: Buffer I/O error on device sde, logical block 195694143

 

I will look for the newer UNMENU and will connect the Seagate drive to the SATA connection and see what happens.

unMENU will update itself, if needed, if you use the "Check for Updates" button on its User-Scripts page.

 

As far as the errors go, there are two devices with errors.  I would check your power and SATA cabling.