Failure is NOT an Option!

[jakkuh_t]

 

[TetraSky]

Yep, I won't pretend to know what it means but my external drive still works even after 10 years. Good video though, to remind people to check the health of their drive.

Spoiler

 

[da na]

IDK man, my 10 year old HGST Ultrastars that came with my workstation in 2012 and have 8 years of flying hours are holding up well... as in, 0 bad sectors, 0 read errors, 0 ECC recovered well.

[Ultraforce]

I'd be curious about USB drives since it feels like a lot of my USB drives fail and am just not sure why.

[Mike Quint]

Had HHD that run years and some failed...but what about tools to check ssd's...and yes they do fail. It's a matter of when...even backblaze is watching this...

[dogwitch]

1 hour ago, Ultraforce said:

I'd be curious about USB drives since it feels like a lot of my USB drives fail and am just not sure why.

some are do to heat thru.

[da na]

1 hour ago, Ultraforce said:

I'd be curious about USB drives since it feels like a lot of my USB drives fail and am just not sure why.

Had a few fail over the years. Cheaper drives use worse NAND which = shorter lifespan. 

[darkImpetus]

Kind of tangent, but related to this. I was wondering if anyone had any experience with Recertified drives. I came across this seller and the seller looks legitimate, but for a home NAS with light to medium use I was wondering what the opinions were or if anyone had any experience.

 

https://serverpartdeals.com/collections/manufacturer-recertified-drives?gclid=Cj0KCQjwlK-WBhDjARIsAO2sErSRD9gLYfwLyJiwSsvZE_KA438Nqm-JWqGc1UmMzH4vsTEMlhtDBV4aAmDlEALw_wcB

[Rysters Tech]

1 hour ago, darkImpetus said:

Kind of tangent, but related to this. I was wondering if anyone had any experience with Recertified drives. I came across this seller and the seller looks legitimate, but for a home NAS with light to medium use I was wondering what the opinions were or if anyone had any experience.

 

https://serverpartdeals.com/collections/manufacturer-recertified-drives?gclid=Cj0KCQjwlK-WBhDjARIsAO2sErSRD9gLYfwLyJiwSsvZE_KA438Nqm-JWqGc1UmMzH4vsTEMlhtDBV4aAmDlEALw_wcB

usually recertified drives were already sent back by someone beforehand, but this is sometimes because the customer wasnt technically informed and thought the product was bad when it actually worked fine, other times it was a dead drive thats had its smart values messed with and bad sectors masked (at least with seagate), it can be hit or miss, i have a recertified seagate medalist drive from the mid 90's that still works, a whopping 2557mb of space, but i also had 2 recertified sshd's from seagate that both failed, granted the other 2 drives in that group of 4 were brand new and they too failed. 

[Rysters Tech]

3 hours ago, Mel0n. said:

IDK man, my 10 year old HGST Ultrastars that came with my workstation in 2012 and have 8 years of flying hours are holding up well... as in, 0 bad sectors, 0 read errors, 0 ECC recovered well.

HGST ultra stars are universally known to last forever, but unfortunately that product line doesn't really exist anymore.

[Nystemy]

Personally use a lot of second hand drives these days.
They are a dime a dosen and with an 8 bay storage server it is easy to just have 2 drives or redundancy + hot spare. And Proxmox that I use as the hypervisor on the server gives SMART data access as well.

 

Though, haven't had many drive failures over the years, but still a few and always nice knowing that the data is in more than just 1 place. _{(The server has RAID, and I have a second machine that I am working on making into a backup. Though no off site for now, other than a bit of cloud.)}

[da na]

30 minutes ago, SiliconMagician said:

HGST ultra stars are universally known to last forever, but unfortunately that product line doesn't really exist anymore.

WD Golds now use the ever reliable Ultrastar chassis. From what I've seen the design is widely unchanged.

 

[kokosnh]

Good luck for novice reading S.M.A.R.T. parameters.    
I'm amazed, the didn't go for the sector dalay scan, only error delay scan. Sector dalay scan in MHDD or Victoria it on of the most trueful test, that do give you actual state of the platter.

[PRSXFENG]

4 hours ago, Mel0n. said:

WD Golds now use the ever reliable Ultrastar chassis. From what I've seen the design is widely unchanged.

 

Isn't it that both HGST Ultrastar and WD Gold are merged into 1 and became WD Ultrastar?

[da na]

8 hours ago, PRSXFENG said:

Isn't it that both HGST Ultrastar and WD Gold are merged into 1 and became WD Ultrastar?

Yeah for their super high end datacenter drives.

[KageUrufu]

For linux users, you can use https://github.com/AnalogJ/scrutiny to get SMART monitoring with a nice web interface, smart attribute history, and push notifications

[ze77y]

My 16TB Seagate Exos somehow managed to corrupt the files I just copied onto it with Total Commander. Most of the new files are 0 bytes and unreadable -- and even some folders appear as 0-byte files. I didn't disconnect the drive during copying or after or do anything else wrong or different than usual. The only thing is that I was away for 3 months and the drive was just sitting in a 5-bay JBOD enclosure.

 

The error appeared while I was syncing it with the backup drive which is another 16TB Exos in the same enclosure (Sabrent DS-SC5B). I am using FreeFileSync for this and haven't had any problems before. Now it actually pointed out that an older directory on the source drive that's been there since the day one was now corrupt.

Quote

A corruption was found in a file system index structure.  The file reference number is 0x10000000004e7.  The name of the file is "\~ COMMISSIONS\2015 -- Voices from the New Jungle\2015.08.1x -- Trip 2\video.gh4".  The corrupted index attribute is ":$I30:$INDEX_ALLOCATION".

 

I retried a few times and by then Windows was already offering me Chkdisk -- I quit FreeFileSync and let it do its thing and while the drive was declared error free, I had now those zero sized files and I am not sure if that's chkdisk doing or were there like that as soon as I copied them. Meanwhile, the directory above can be opened and seems fine.

 

There was no way to restore the files with any tools I tried (DMDE, DiskDigger, Recoverit) but SMART reports no errors, so I am thinking maybe just reformat it and give another go instead of sending it off to Seagate for replacement? I am in Europe and while I can ship it to Netherlands, they still redirect it to the US and it takes weeks and weeks to receive the replacement, all while being left without a proper backup. I only had the drive for less than a year and haven't used it that much although it's now 80% full of video footage.

 

What could have caused this and what are the odds of it happening again with SMART being positive? Do I do a full surface scan -- would that be enough?

 

Also, I was thinking maybe it's the JBOD enclosure I am using (Sabrent DS-SC5B) as I vaguely remember that there was another issue with a disk early on but it was an empty disk and I just reformatted it.

 

There's one other error I could find in Event Viewer at the time which is related to the disks, which is:

Quote

 

Disk 6 has the same disk identifiers as one or more disks connected to the system. Go to Microsoft's support website (http://support.microsoft.com) and search for KB2983588 to resolve the issue.

 

Disk 7 has the same disk identifiers as one or more disks connected to the system. Go to Microsoft's support website (http://support.microsoft.com) and search for KB2983588 to resolve the issue.

 

Upon quick research it seems like this doesn't affect functionality -- and it all worked fine before.

 

Thoughts?