Computers & Electronics

Experienced my first HDD failure in 15 years...hear the HDD scream in agony!

  • Last Updated:
  • Jun 13th, 2016 3:09 am
Tags:
None
[OP]
Deal Addict
User avatar
Sep 23, 2014
1928 posts
561 upvotes
Toronto, ON

Experienced my first HDD failure in 15 years...hear the HDD scream in agony!

Well I guess HDD Satan decided to visit me for the first time in my life and one of my seagate HDD failed a horrible death - it is the notorious Seagate 3TB Barracuda, one of the worst HDDs are rated by Rackblaze - https://www.backblaze.com/blog/3tb-hard-drive-failure/

The good thing with FreeNAS is (HDD used in NAS box) the system has started warning me via email on SMART error beginning with this:

First Warning - Bad sectors are appearing (I thought no big deal, after all the drive is almost 4 years old
Device: /dev/da0 [SAT], 8 Offline uncorrectable sectors
Device: /dev/da0 [SAT], 8 Currently unreadable (pending) sectors


Second Warning:
Device: /dev/da0 [SAT], 112 Offline uncorrectable sectors
Device: /dev/da0 [SAT], 112 Currently unreadable (pending) sectors

Third Warning:
The volume HTPC (ZFS) state is ONLINE: One or more devices has experienced an unrecoverable error. An attempt was made to correct the error. Applications are unaffected.

Fourth Warning
Device: /dev/da0 [SAT], ATA error count increased from 0 to 43 The volume HTPC (ZFS) state is ONLINE: One or more devices has experienced an unrecoverable error. An attempt was made to correct the error. Applications are unaffected.

Fourth Warning
Device: /dev/da0 [SAT], unable to open device
Device: /dev/da0 [SAT], ATA error count increased from 0 to 43 The volume HTPC (ZFS) state is UNAVAIL: One or more devices could not be opened. There are insufficient replicas for the pool to continue functioning.
Device: /dev/da0 [SAT], 8 Currently unreadable (pending) sectors (changed -65368)
Device: /dev/da0 [SAT], 8 Offline uncorrectable sectors (changed -65368) The volume HTPC (ZFS) state is ONLINE: One or more devices has experienced an error resulting in data corruption. Applications may be affected.

Too bad all the critical error happened around around 2 am so all the data (mostly useless) are kaput...but the warning system is very elegant and don't we wish Windows has the same features :D
Mining & Workstation EPYC 7702 64 Core & Dual 3950x Rigs | ASUS STRIX RTX 3090 OC | 3X NEC PA301W | 64GB DDR4 ECC |5 TB Samsung NVMe Storage
12 replies
Deal Addict
Aug 25, 2006
3675 posts
967 upvotes
Firstly, my condolences! But the best part as you mentioned is that the system tried to warn you. I had that happen to me on a WHS 2011 machine and it was cool indeed. The warning system I had was actually ahead of the curve and warned me when I logged in to the server. I did not set up the extra portion of the software where it would migrate data off the disk going bad. Yes, if Windows had these features, it would be fantastic. I think there are 3rd party installs that can do this, I should really look into this as well.

The drive I had die was a WD Green drive. It threw and error, I could fix it but it would return. So I migrated all data off and replaced it with another drive in the WHS server. All is well again.

Good luck!
Deal Fanatic
Sep 4, 2009
6575 posts
717 upvotes
it does say uncorrectable sectors, unreadable sectors that should be concerning because it means you lost data. It should never get to this point to begin with. I would have started migrating the data once this happens.
[OP]
Deal Addict
User avatar
Sep 23, 2014
1928 posts
561 upvotes
Toronto, ON
rageking wrote: it does say uncorrectable sectors, unreadable sectors that should be concerning because it means you lost data. It should never get to this point to begin with. I would have started migrating the data once this happens.
I know but the HDD is storing garbage so I wanted to see how long it would've lasted...and it wasn't that long haha..less than a week.
Mining & Workstation EPYC 7702 64 Core & Dual 3950x Rigs | ASUS STRIX RTX 3090 OC | 3X NEC PA301W | 64GB DDR4 ECC |5 TB Samsung NVMe Storage
Deal Expert
Oct 6, 2005
16599 posts
2315 upvotes
I rotate my NAS hard drives every 1.5 to 2 years - keeps them fresh, just outside of when the drives will start to fail at a higher frequency.
[OP]
Deal Addict
User avatar
Sep 23, 2014
1928 posts
561 upvotes
Toronto, ON
coolspot wrote: I rotate my NAS hard drives every 1.5 to 2 years - keeps them fresh, just outside of when the drives will start to fail at a higher frequency.
Wow..isn't that a bit extreme? Unless you hammer then everyday with high workload, most quality enterprise grade HDDs should last for years! My first NAS was an Ubuntu based system from 2004 running a few U320 SCSI 10K RPM HDDs and had zero failures up to 2014 when they finally retired :D MY current NAS HDDs are Hitachi Ultrastars and lets see how long they last!
Mining & Workstation EPYC 7702 64 Core & Dual 3950x Rigs | ASUS STRIX RTX 3090 OC | 3X NEC PA301W | 64GB DDR4 ECC |5 TB Samsung NVMe Storage
Deal Fanatic
User avatar
Mar 20, 2009
8862 posts
2680 upvotes
Vancouver
bowmah wrote: The drive I had die was a WD Green drive.
Those WD Green drives turned out to be a disaster. We had several, and every one of them has failed now (and not a single other drive model failed over the same time period!). WD has dropped the "Green" line entirely now to try to get rid of the bad reputation, and if they had any integrity they should have offered a replacement program like IBM did for their Deathstar drives.
Deal Addict
Feb 9, 2008
2363 posts
173 upvotes
I've used a ton of WD 'Green' drives and only had one with problems (it still worked, but started to get some bad blocks). Still have one in my Windows PC from 2013, but it's almost full and I'll have to upgrade before long.
Deal Expert
Mar 23, 2004
28941 posts
9460 upvotes
JamesA1 wrote: Those WD Green drives turned out to be a disaster.
Yeah that's why WD simply renamed them "Blue", bringing back the Blue line from the brink of extinction, lol. So now they still have the exact same drives, they just renamed them because of the bad rep :rolleyes: And the real Blue drives are still there (barely) too, but it's just rather weird. Instead of Green and Blue, now there's two totally different Blue drives, all from the flak they got from the Greens. I guess they are fooling someone out there.
Deal Expert
Oct 6, 2005
16599 posts
2315 upvotes
traderjay wrote: Wow..isn't that a bit extreme? Unless you hammer then everyday with high workload, most quality enterprise grade HDDs should last for years! My first NAS was an Ubuntu based system from 2004 running a few U320 SCSI 10K RPM HDDs and had zero failures up to 2014 when they finally retired :D MY current NAS HDDs are Hitachi Ultrastars and lets see how long they last!
It's cheaper to replace the drives then to buy enterprise grade drives; plus I upgrade the capacity as well (i.e. 2 TB -> 3TB --> 4TB --> Saving up for 6TB drives).

I don't replace all the drives at once, but do it in staggered pairs - this ensure that I won't be killed by a bad batch failure (i.e. all drives failing at the same time).

Due to the number of drives in my system, each drive ends up with 4 years of life, before cycling out to my bulk storage (non-critical) NAS.

JamesA1 wrote: Those WD Green drives turned out to be a disaster.
My Green Drives have been OK, but I had to disable the Drive Parking timers to ensure that there are no excessive parks.
Deal Expert
Aug 2, 2004
33740 posts
7374 upvotes
East Gwillimbury
If you only had one drive fail and you used ZFS1, all you need to do is offline the drive, your data should be intact. It will run in a degraded state but you shouldn't lose any files. No matter how useless they are
[OP]
Deal Addict
User avatar
Sep 23, 2014
1928 posts
561 upvotes
Toronto, ON
Gee wrote: If you only had one drive fail and you used ZFS1, all you need to do is offline the drive, your data should be intact. It will run in a degraded state but you shouldn't lose any files. No matter how useless they are
The drive is not part of an array - just a single drive in its own ZPOOL as a network junk drive...FreeNAS can't even detect the drive anymore lol
Mining & Workstation EPYC 7702 64 Core & Dual 3950x Rigs | ASUS STRIX RTX 3090 OC | 3X NEC PA301W | 64GB DDR4 ECC |5 TB Samsung NVMe Storage
Deal Expert
Aug 2, 2004
33740 posts
7374 upvotes
East Gwillimbury
I had 24 of those 3 Terabyte Seagate drives. One died under warranty, another 18 died after 1000 days. Just after the warranty expired. They all died within 6 weeks of each other. Fortunately I was able to recover 99.9% of my data. I think I lost a total of a dozen files that can easily be replaced.

If it wasn't for ZFS and the warnings, it would have been a lot worse

Top