[SOLVED] Shall I RMA this drive? (WD20EARX)

[dheg]

Hi guys,

 

This is my first post here. I've been running unRaid without much trouble for about 3/4 months. Indeed the forum and a quick search on google got me out of a few problems, but now I'm completely puzzled.

 

I bought a Western Digital drive (WD20EARX) about 1.5 months ago. I run preclear (v1.11) for 5 times (this is the standard I use after long readings on the forums) and got a PASSED result. Please find below report:

 

http://pastebin.com/jm7HsqLp

 

Then about 2 weeks ago I started having problems on my XBMC box. Image froze and "bufffering" message appeared. I looked into the smart view of the mymain plugin and started seeing errors on this drive, though the overall status was passed, so I didn't pay much attention. Today I got up willing to find out about what was really going on with my drive and run a long smart report from the console, I have attached the report.

 

So now, I'm completely lost. From reading the SMART report I would say the drive is dying: 172 errors on a 700-hours old drive (I think these are too many). Besides all the errors on the log are UNC (I was told by a friend these are unrecoverable).  The Raw_Read_Error_Rate  is well over 95,000, but the overall result is PASSED   ???

 

Although I'm quite good at computers, I know very little about linux, and even less about hardware diagnostics. I though this would be the right time to get to know you guys  

 

BTW: I'm currently running another preclear on a new drive (WD20EARX also) I bought a couple of days ago. Results are good so far, but how much shall I trust the results?

smart.txt

[mbryanr]

While many attributes cannot be interpreted, and specific for each manufacturer - the two below can be an indicator of failure. While a drive may have >1000 spare sectors...this one is dying.

 

197 Current_Pending_Sector  0x0032   200   200   000    Old_age   Always       -       300

198 Offline_Uncorrectable   0x0030   200   200   000    Old_age   Offline      -      148

 

Also...Raw Read Errors are high as you noted.- due to the unreadable sectors

1 Raw_Read_Error_Rate     0x002f   174   174   051    Pre-fail  Always       -       95952

 

and the smart test aborts because sectors cannot be read from:

#11  Extended offline    Completed: read failure       90%       114         1572928

 

from above...looks like this drive was dying early.  I noticed that it threw up a single reallocated sector during preclear cycle 2.  Not a problem, just an indicator to monitor the drive closely which you did.  

 

[dheg]

Ok what I thought.

 

Why then the tests give me a PASSED result? Will Western Digital accept the drive if it passes the tests?

 

[mbryanr]

Because the smart test compares to the threshold value.  ie everything passes up to the sector that it can not be read.  WD will accept the RMA.  I believe wd has utilities to test the drive.  I know makes no sense. That is why it is important to trend your actual values over time and monitor the syslog. I send mine to a syslog server which sends emails based willon severity. I also have unraid status reports where you can see drive errors, note it can still report ok because a drive hasn't redballed

[lionelhutz]

Send it back.

 

Do not run a correcting parity check on the server until you have replaced and rebuilt the drive.

 

Peter

 

[dheg]

Thank you guys, I think I have it clear. I think I can read now a SMART report