54 errors, 1 par check error

[steini84]

Hey guys

 

I just added 2x 3tb drives to my array last month and now my first par check finished this morning.

 

I have 54 errors on disk 11 (sdj - one of the new drives) and got this from the par check "Last checked on Fri Mar 1 09:36:44 2013 UTC, finding 1 errors."

 

I have a couple of these in the log

Mar  1 06:07:05 SERVER kernel: handle_stripe read error: 3444003880/10, count: 1
Mar  1 06:07:05 SERVER kernel: md: disk11 read error

and then a chunk of these:

Mar  1 06:07:02 SERVER kernel: ata12.00: configured for UDMA/133
Mar  1 06:07:02 SERVER kernel: ata12: EH complete
Mar  1 06:07:05 SERVER kernel: ata12.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0
Mar  1 06:07:05 SERVER kernel: ata12.00: irq_stat 0x40000001
Mar  1 06:07:05 SERVER kernel: ata12.00: failed command: READ DMA EXT
Mar  1 06:07:05 SERVER kernel: ata12.00: cmd 25/00:e0:a0:50:47/00:03:cd:00:00/e0 tag 0 dma 507904 in
Mar  1 06:07:05 SERVER kernel:          res 51/40:00:e0:52:47/00:00:cd:00:00/00 Emask 0x9 (media error)
Mar  1 06:07:05 SERVER kernel: ata12.00: status: { DRDY ERR }
Mar  1 06:07:05 SERVER kernel: ata12.00: error: { UNC }
Mar  1 06:07:05 SERVER kernel: ata12.00: configured for UDMA/133
Mar  1 06:07:05 SERVER kernel: sd 12:0:0:0: [sdj] Unhandled sense code
Mar  1 06:07:05 SERVER kernel: sd 12:0:0:0: [sdj]  Result: hostbyte=0x00 driverbyte=0x08
Mar  1 06:07:05 SERVER kernel: sd 12:0:0:0: [sdj]  Sense Key : 0x3 [current] [descriptor]
Mar  1 06:07:05 SERVER kernel: Descriptor sense data with sense descriptors (in hex):
Mar  1 06:07:05 SERVER kernel:         72 03 11 04 00 00 00 0c 00 0a 80 00 00 00 00 00
Mar  1 06:07:05 SERVER kernel:         cd 47 52 e0
Mar  1 06:07:05 SERVER kernel: sd 12:0:0:0: [sdj]  ASC=0x11 ASCQ=0x4
Mar  1 06:07:05 SERVER kernel: sd 12:0:0:0: [sdj] CDB: cdb[0]=0x28: 28 00 cd 47 50 a0 00 03 e0 00
Mar  1 06:07:05 SERVER kernel: end_request: I/O error, dev sdj, sector 3444003552

 

A smart check of the drive can be found here: http://pastebin.com/epHRRsg9

 

What should my next steps be? I have started another par check and hopefully this was just a one off

 

It´s a new drive and even though it passed the smart check, I dont see anything too suspicious like bad blocks.

 

Looking forward to hearing what you guys think.

 

[Joe L.]

This looks odd, it appears as if the error was in accessing the last block on the disk:

40 51 00 ff ff ff 0f  Error: UNC at LBA = 0x0fffffff = 268435455

Yes, no other evidence appears in your SMART report of a UNC error.

However, the seek-time attribute is not looking too good. (getting closer to its affiliated failure threshold)

so perhaps something mechanical is not perfect.

 

About all you can do is another NOCORRECT check.  If the parity error is on the same place on the disk, then you'll probably need to perform a correcting parity sync.

 

Joe L.

 

[steini84]

Thanks for the super quick reply

 

It´s a brand new drive so i´m in denial that it´s broken already. But If that is the case i´ll just have to RMA it

 

Are you talking about this line in the seek time (I´m no so great at reading these.)?

  7 Seek_Error_Rate        0x000f  060  060  030    Pre-fail  Always      -      1083621

 

But after the next nocorrect check finishes should I compare the errors and if they match I can just do a correcting par check?

 

Thanks again for the quick reply

[steini84]

Strange did another nocorrect check and now it did not find any errors.

Strange

[dgaschk]

The ata errors occurred at 452 hours. Check for a loose cable or replace the SATA cable.

[steini84]

Will do. Thanks