Lakz

How bad is it?

Recommended Posts

I bought a 4TB Seagate over a year ago. I use it as a parity disk in a software RAID environment. I just looked at the SMART attributes, and this is what I got :

smart.gif

Should I be worried?

Should I replace it right away?

Would it be covered by Seagate's warranty? (expiring in June 2015)

Thanks guys.

Share this post


Link to post
Share on other sites

Warranties are worthless to me. I don't want my data in others hands.

I'd stop using it immediately.

Buy an identical drive, or larger.

Don't even mount the file system. Fire up Linux and use gddrescue, copy it at the block level. Good luck.

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

And a follow-up up, issue a dco factory reset with hdparm, it'll wipe reallocation that the firmware has already done.

Refill with zeros via dd from /dev/zero, and see if they reappear.

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

Share this post


Link to post
Share on other sites

Nothing appears to be near any failure thresholds, although SMART data typically only predicts about 50% of drive faiures... and the drive does show a few reallocated sectors, but that's not a sign of immentient failure.

(would I be entirely happy? Not really... but I wouldn't be all that worried-- keep an eye on the count, if it's increasing then yes, I would worry...)

Pull out the drive, run a full (long) test in Seagate's SeaTools, and see what it says.

Share this post


Link to post
Share on other sites

I tried running Seagate tools for my 18 3tb 'cudas..... I couldn't justify buying a M$ license to run free software.

So I stick with hdparm and smartmontools.

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

Share this post


Link to post
Share on other sites

Warranties are worthless to me. I don't want my data in others hands.

You can wipe the drive, or use encryption.

Edited by eshap

Share this post


Link to post
Share on other sites

 

You can wipe the drive, or use encryption.

Wipe a drive that a firmware spits back because of bad sectors? Then For a $90 drive?

No.

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

Share this post


Link to post
Share on other sites

Haha, then deal with this one top of a 20+tb raid 6 MD array that needs to be shrank as well ?

No.

I'll spend $100 bill, save 3 weeks of resize calcs, and not worry of goofing somewhere ad loosing it all.

Way easier to grow then shrink. Growing encryption is 3x, shrinking is 12x.....

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

https://help.ubuntu.com/community/ResizeEncryptedPartitions

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

Share this post


Link to post
Share on other sites

Thank you for responding.

I did run Seatools via the provided image, but it had troubles finding the drive (probably because it's a RDM). I ultimately pulled the drive out of the server and ran Seatools on a Windows machine I have.

SMART - Pass
Short DST - Pass
Long Generic - Pass
I'm not too worried about having the drive dying on me. If it ever happens, I could easily rebuild the data from the other drives in the array (FlexRAID).
What absolutely do not want, is encountering any sort of data corruption. I'm no expert, but wouldn't a 0xC6 error (uncorrectable sector) implies that the data written in that sector is definitively lost?

Share this post


Link to post
Share on other sites

Actually I hope you are running at least raid5, you'll need parity to rebuild from besides a straight mirror.

You really should be running raid6, hardware or software via mdadm , or go zfs and raidz3.....

I got worried after 12tb and added a 2nd parity drive to my array, and issued a grow to level raid6. Didn't expand storage, but rewrote a 2nd parity bit...

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

Share this post


Link to post
Share on other sites

Not sure on flex, but mdadm does a monthly check. It'll catch issues if reads aren't matching parity.

Hope for the best. If it's just media that's replaceable, giggle, fiddle, and have fun.

Sent from my rooted HTC Supersonic using Tapatalk 2 Pro

Share this post


Link to post
Share on other sites

FlexRAID does verify the data and is supposed to catch bit rot (I have it scheduled twice a week). Luckily, it didn't pick up any error so far. However, the Reallocated Sectors Count is up by nearly 1000 since my first post. Seek Error Rate is rising as well. I think it's time for a new drive.

Thanks a lot!

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now