Advice Needed - Noticed errors in syslog assoc. with parity Drive

[Queball]

I noticed several errors in my syslog (attached) that appear to be associated with my parity drive (smart report attached). I am using unRAID version 4.7.

 

Here are some examples from syslog:

Sep  4 05:57:34 Tower kernel: ata10.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0 (Errors)
Sep  4 05:57:34 Tower kernel: ata10.00: irq_stat 0x40000001 (Drive related)
Sep  4 05:57:34 Tower kernel: ata10.00: failed command: READ DMA EXT (Minor Issues)
Sep  4 05:57:34 Tower kernel: ata10.00: cmd 25/00:18:18:f8:f4/00:00:29:00:00/e0 tag 0 dma 12288 in (Drive related)
Sep  4 05:57:34 Tower kernel:          res 51/40:07:20:f8:f4/00:00:29:00:00/e0 Emask 0x9 (media error) (Errors)
Sep  4 05:57:34 Tower kernel: ata10.00: status: { DRDY ERR } (Drive related)
Sep  4 05:57:34 Tower kernel: ata10.00: error: { UNC } (Errors)
Sep  4 05:57:34 Tower kernel: ata10.00: configured for UDMA/133 (Drive related)
Sep  4 05:57:34 Tower kernel: ata10: EH complete (Drive related)

 

And:

Sep  4 20:39:12 Tower kernel: sd 4:0:0:0: [sdi] ASC=0x11 ASCQ=0x4 (Drive related)
Sep  4 20:39:12 Tower kernel: sd 4:0:0:0: [sdi] CDB: cdb[0]=0x28: 28 00 6f da 98 50 00 04 00 00 (Drive related)
Sep  4 20:39:12 Tower kernel: end_request: I/O error, dev sdi, sector 1876596816 (Errors)
Sep  4 20:39:12 Tower kernel: md: disk0 read error (Errors)
Sep  4 20:39:12 Tower kernel: handle_stripe read error: 1876596752/0, count: 1 (Errors)
Sep  4 20:39:12 Tower kernel: md: disk0 read error (Errors)
Sep  4 20:39:12 Tower kernel: handle_stripe read error: 1876596760/0, count: 1 (Errors)
Sep  4 20:39:12 Tower kernel: md: disk0 read error (Errors)
Sep  4 20:39:12 Tower kernel: handle_stripe read error: 1876596768/0, count: 1 (Errors)
Sep  4 20:39:12 Tower kernel: md: disk0 read error (Errors)
Sep  4 20:39:12 Tower kernel: handle_stripe read error: 1876596776/0, count: 1 (Errors)
Sep  4 20:39:12 Tower kernel: ata10: EH complete (Drive related)

 

Do I need to replace the parity drive? I am thinking that I should replace the parity drive with a new drive and let parity rebuild. After which I plan to test the current (potential bad) parity drive with a few pre-clear passes to see if the pending sectors are resolved. Based on this testing I may add back to the array.

 

Any thoughts on this plan or recommendations for better/alternate ways forward? Also interested in any advice that will allow me to upgrade to version 5 during this process/parity drive replacement. Current hardware is:

 

MB - Biostar A760G M2+.

1 x SATA2 Serial ATA II PCI-Express RAID Controller Card (Silicon Image SIL3132)

1 x Supermicro AOC-SASLP-MV8 8-Port SAS/SATA Add-on Card

3 x Norco SS-500

Azza Helios 910

Corsair TX650

 

I run the following add-ons:

SabNZBD

Couchpotato

Sickbeard

Unmenu

Crashplan

 

Thanks in advance for any help.

syslog-2012-09-05.zip

Smart_Report.txt

[JonathanM]

Your current roadmap of building parity to a fresh drive sounds good. I'm assuming the drive you plan to put in place has been burned in with preclear.. As far as simultaneously upgrading to 5.x, I'd strongly recommend NOT doing it until the array is stable and working well for a period of time. Too many changes at once causes troubleshooting headaches.

[Queball]

Your current roadmap of building parity to a fresh drive sounds good. I'm assuming the drive you plan to put in place has been burned in with preclear.. As far as simultaneously upgrading to 5.x, I'd strongly recommend NOT doing it until the array is stable and working well for a period of time. Too many changes at once causes troubleshooting headaches.

 

Thanks for the advice. I'll wait until this issue is resolved to upgrade to 5.xx. I'll be certain to use an already pre-cleared drive for the parity replacement.