New drive, errors in pre-clear but SMART is ok? [fixed, faulty SATA cable]

[Yarok]

Hello everybody,

 

I'm having a problem with a new drive i've been trying to install for a couple of days now. I'm not sure where to begin but here goes:

 

- When i try to preclear using the plugin, it either runs for anywhere from 90min to 11 hours, but at the end the drive remains 'not pre-cleared';

- At times, the preclear process stalls and stops completely, stuck at 0% after several hours;

- Testing the drive with the plugin gives the following report, please note 'failed tests' 1 and 3: 

 

Quote

 Pre-Clear unRAID Disk /dev/sdf
################################################################## 1.15
Model Family:     Western Digital Red (AF)
Device Model:     WDC WD20EFRX-68EUZN0
Serial Number:    WD-WCC4M0YP6XJH
LU WWN Device Id: 5 0014ee 2b9926fc4
Firmware Version: 82.00A82
User Capacity:    2,000,398,934,016 bytes [2.00 TB]
Disk /dev/sdf: 2000.4 GB, 2000398934016 bytes
1 heads, 63 sectors/track, 62016336 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disk identifier: 0x00000000
   Device Boot      Start         End      Blocks   Id  System
/dev/sdf1              64  3907029167  1953514552   83  Linux
########################################################################
failed test 1
failed test 3 00131 00000 00000 00000
========================================================================1.15
==
== Disk /dev/sdf is NOT precleared
== 64 3907029104 3907029104
============================================================================

 

- If i try and mount the drive, it remains red-crossed and lastly, the error count showed 14 errors (?) on the main dashboard;

- There are more than 6500 UDMA CRC errors but i'm not sure this is relevant or not.

 

I have a attached the latest SMART report, which passed.

 

My local reseller asks for the SMART report before validating the return of the drive, so i'm afraid that, is the drive is faulty but with a negative SMART, they'll refuse after sales service

 

Can you please help me troubleshooting this? Many thanks in advance for your support!

 

Cheers,

 

Patrick

deimos-smart-20180115-1222.zip

[JorgeB]

UDMA CRC errors are usually the result of a bad SATA cable, in this case a very bad one, replace it and try again.

[Yarok]

Thanks Johnnie, i did this and ran a new preclear, i'll keep you guys posted.

[Yarok]

well, this is looking promising already  Preclear is passed, now doing post-read!

[Yarok]

Okay so here we are a day later. A single cycle run of Pre-clear passed, this morning before leaving to work i ran another SMART test and, even though i replaced the sata cable i still get 6600 CRC errors - any thoughts about this? Is this result cumulative, meaning that i had 6600 errors with the old cable, 0 with the new, for a total of 6600 errors?

 

i'm planning on running a three cycle pre-clear, is there a point in doing so?

deimos-smart-20180116-1840.zip

[JorgeB]

Just now, Yarok said:

Is this result cumulative, meaning that i had 6600 errors with the old cable, 0 with the new, for a total of 6600 errors?

Yes, UDMA CRC errros don't reset, it only matters that it doesn't keep increasing.

[Yarok]

Good news, thanks again Johnnie!

[SSD]

Often the cable itself is not bad, but the cabling is not securely plugged, or connection knocked slightly askew.

 

This often happens when adding new or swapping disks, as seems to be the case here. It's just too easy to touch a neighbor's cable and create the opportunity.

 

Locking cables can help, but they are not that high quality and get knocked loose too. And not all ports support locking connections. Hot-swap cages are the best, as they allow drives to be plugged and unplugged without touching the cables that can compromise the sensitive connections. The worst is when you are doing a rebuild and the rebuild fails because another disk fails. You're then left with a nasty little recovery to negotiate. Dual parity can help, but still requires extra effort to recover. With the hot-swap bays it should never happen in the first place.

 

I personally would not have a server without the hot-swap cages. My favorate is the SuperMicro CSE-M35T-1B cages, which are often on eBay.

 

Cheers!

 

 

[Yarok]

Thanks for the tip, SSD! 

 

I could actually see some sort of oxydation on the faulty (locking) cable connectors. The replacing of this cable did the trick!