Failing Drive? (Again?)

[SmallwoodDR82]

Thanks in advance!

 

I had a drive fail on me about 2 months ago, maybe less.  It was still under warranty so I had it replaced through Seagate. (Not trying to start a pissing match between the best Manu on this thread  )

 

I wasn't going to wait for the drive to get here so in the mean time I swapped with my hot swappable and all went fine.

 

I received the drive from Seagate and use it to swap out another older drive in my array, again all went fine until now.  I'm getting odd things from it since a Parity Check last night.

 

See the image

 

 

I've rebooted the server a few times and even moved it to another bay on my case.  I can navigate the drive fine but the array is acting like it's invalid.

 

All self tests seem to pass.

 

Any ideas?

 

Syslogs after reboot attached.

 

Thank you!

syslog.zip

[trurl]

Don't see anything in the syslog but I think the grey triangle means the disk is disabled and also spun down. See here.

[SmallwoodDR82]

Thanks for the reply!

 

I went ahead and follow that proceed and it seems to be rebuilding now.

 

A few questions

 

1.) What caused it to go "disabled"?

2.) Why is it rebuilding all the data on the drive if I could see the drive no problem?

3.) Is this drive failing or was this just a "choke" on the unRAID software side?

 

 

[trurl]

1) unRAID disables a disk when a write to it fails.

 

2) When a disk is disabled, unRAID simulates it by calculating its data from all the other disks plus parity. This calculation is also how it can rebuild it.

 

3) Click on the disk in Main and it will take you to a page where you can access the drive Health. Select Disk attributes and post the results.

[SmallwoodDR82]

[dirtysanchez]

The drive is failing (reallocated sectors) and should be replaced.

[Frank1940]

SmallwoodDR82 ----

 

You should probably look up S.M.A.R.T. on Wikipedia.  There is a table of all the SMART attributes in the article.  Here is a link to the appropriate page:

 

    https://en.wikipedia.org/wiki/S.M.A.R.T.

 

Frank

 

EDIT: This replacement drive is probably a 'reconditioned' drive.  So it has probably already been returned as defective...

[SSD]

I had it replaced through Seagate. (Not trying to start a pissing match between the best Manu on this thread  )

 

Drive manufacturer and model are the most important consideration for drive longevity. For me personally, Seagate 2T drives were the worst of the worst. I didn't buy any more Seagtes for a long time, but did buy 2 of their 4T drives that got (relatively) good marks from BackBlaze, and so far they have worked well.

 

Not bashing all Seagate drives, but no amount of lipstick is going to make the Seagate 2T pretty!

 

Here are some excerpts from Newegg owner comments on this Seagate 2T model ...

 

Cons: Mine has a Date of Manufacture of 05/2013, and like so many others mine has completely died.

 

Cons: Died at 1 year 5 months of light use

 

Cons: All I have to say is I bought 12 (TWELVE) of these - OVER HALF were DOA. Seagate replaced them fast - but when the ones that weren't DOA started to die after a SHORT while (AND THEY WILL), they replaced those with REFURBISHED DRIVES - ALL of which have been DOA (I JUST tried to use the last one today - opened the packaging and - DEAD. Certified Refurbished is TOTAL cowpie (the censors here...)!). The ONE good thing I have to say is when I spoke to a CSR, he broke the rules and replaced two of them with brand new 3TB drives - but couldn't do it for all of my RMA's (the he had to send me what have turned out to be USELESS refurbished drives) - so I am forced to buy MORE drives (BUT NEVER AGAIN FROM SEAGATE) to replace the ones that I should have had great life with from Seagate. It really is a shame as I USED to be a Seagate evangelist - but in their vigor to earn more money at the expense of quality - their drives have gone to the dogs. Paperweights - that's all they are good for.

 

Other Thoughts: I would NOT (a VIGOROUS NOT) recommend these drives to ANYONE. And Seagate will NOT realize that replacing a DEAD drive with ANOTHER dead drive is NOT customer service.

[dgaschk]

Amazon reviews are anecdotes and I think the ratings a skewed to the low end because people with bad experiences are more likely to post. I've 2T Seagates that are showing no degradation after 4 yrs 24/7. Unless I get real data showing that a particular model should be avoided I always get what's on sale. This saves money and yes occasionally disks fail. I've no emotional baggage due to a particular brand losing my important data. I've seen fairly even failure rates across brands and models.

[SmallwoodDR82]

Thanks for the feedback!  I'll get to work on replacing the drive!

[JonathanM]

Amazon reviews are anecdotes and I think the ratings a skewed to the low end because people with bad experiences are more likely to post.

This.

 

Unless you are buying by the 100's like backblaze, you can't definitively link your single drive failure rate with a particular model. It's much more likely that the particular drive was the victim of shipping or handling or power or other after the factory incident than a factory defect. Even with backblazes worst drives that they purchased in quantity way more than half of a statistically significant number survived.

 

I see poor packaging and ignorant handling by the user as way more significant than factory defects for the small number of drives that most individuals go through.

 

Consider that most people don't think twice about stacking 2 drives on a desk while working on a machine, but the instantaneous G force of accidently setting one drive on top of another with an audible CLACK well exceeds the design limits of many desktop drives.

[SSD]

Amazon reviews are anecdotes and I think the ratings a skewed to the low end because people with bad experiences are more likely to post.

 

Amazon drive feedback is lousy because they combine feedback of the same model regardless of size. There are definitely differences in the same model in different capacities.

 

I've 2T Seagates that are showing no degradation after 4 yrs 24/7. Unless I get real data showing that a particular model should be avoided I always get what's on sale. This saves money and yes occasionally disks fail. I've no emotional baggage due to a particular brand losing my important data. I've seen fairly even failure rates across brands and models.

 

As you say, personal anecdotal stories are not that helpful. Read about any drive and you see issues with DOAs and failures. But the BackBlaze reports consistently give HGST drives the highest reliability ratings. And the range of failure rates - from < 1/2% to well over 20% per year certainly provides evidence that different drive manufacturers and models have very different reliability characteristics.

 

The fact that my personal anecdotal experiences align closely with what I see reported tends to substantiate it to me. But understand you and many others may be part of the 80% per year that didn't fail, and your anecdotal experience might not align.

 

I will say that I went out on a limb and bought 4 white labeled 5T drives that had all the tell take signs of being WD drives. Every one of them developed reallocated sectors during preclear - 3 on the first cycle, and the last one on the 2nd. I got a full refund. There disks were packaged perfectly. What does this have to do with Seagate? Nothing except to say that not all drives are not built to the same quality standards. So buying the cheapest is not always the best idea.

 

Just my $0.02.