• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

26TB drive for $250

I shucked a 28TB one recently, got an assumedly binned EXOS
nice! most 26TB seem to be Barracuda's now. Likely why they are on sale. I think 28TB might be worth buying to check if its exos, don't even have to shuck it can run it through crystal disk first. I am thinking thats because haven't seen any 28TB barracudas and the agressive sale on 26TB.
 
Aww man, I bit on the 24 tb drives for the same price in May. Those were bare drives, but supposed to only run them 12 hours a day (I'm running them 24 hours a day though)
 
Do you guys even test these drives before you use them? IIRC 14GB drives took more than 24 hours to do a full write then read sector by sector test, these would take 2+ and that doesn't even take into consideration copying the data over or rebuilding the array if in a raid. I am hesitating to start using larger drives into my system only because of how long it would take to just test them to be sure they aren't defective on arrival.
 
Do you guys even test these drives before you use them?

No

rocky-iv-dolph-lundgren.gif


Edit: I also power my drives on and off everyday along with the server too ☠️
 
Do you guys even test these drives before you use them? IIRC 14GB drives took more than 24 hours to do a full write then read sector by sector test, these would take 2+ and that doesn't even take into consideration copying the data over or rebuilding the array if in a raid. I am hesitating to start using larger drives into my system only because of how long it would take to just test them to be sure they aren't defective on arrival.
I don't do a full sector by sector test, but I always do a smart conveyance test (fast), and then a smart long test, which takes forever and has the worst progress bar system ever ... let's see this takes 39 hours, so we should only update progress by 10%, makes sense.

I don't really need that much space, so I run zfs mirrors, and scrub weekly. I'll do a scrub right after adding to the mirror, and if that's fine, I'm not going to worry. I have backups to other systems on the data I'd like to be sure to keep. I got 2x24tb because my 2x10tb had hit 5 years power on time, and 24 tb was the best $/TB. I've got my tv recordings on the 2x10tb now, and will probably maybe do dvd rips to that mirror as well, when I eventually get around to it. If I'm just doing an upgrade with no sense of urgency, I try to wait and buy the second drive a few weeks later in a different order, hoping for a different production batch, and so that if they have a firmware bug leading to death at a specific number of hours, I'll have time to notice and replace the first failure.... I hope.
 
I don't do a full sector by sector test, but I always do a smart conveyance test (fast), and then a smart long test, which takes forever and has the worst progress bar system ever ... let's see this takes 39 hours, so we should only update progress by 10%, makes sense.
Same here. I have a 8-bay esata tower I got here many years back and have a dedicated system where I boot up a parted magic live cd and run the tests because they can run in parallel. So when I did the tests on my 16TB drives back when I got them, it only took the time of testing a single drive to test all of them. I've caught some doa drives over the years this way and am able to go through the painful warranty process before actually needing the drive, saving quite a bit of stress.
 
Showing $329 here...
And what else could this be besides EXOS?
Four of these at original price would be nice for UNVR duties.
But if they are some SMR crap, hell no!
 
Showing $329 here...
And what else could this be besides EXOS?
Four of these at original price would be nice for UNVR duties.
But if they are some SMR crap, hell no!
They are likely exos but binned as barracudas due to being external and cooking in that box if used too much. I think they should be mostly fine if they are run at good temps. Read from few posts on reddit that the most of 24+ TBs expansion drives are exos that didn't pass the strict standards for data centers so they bin them as barracudas, + they are in external box that run hot if on all the time lowering lifespan, hence lower ratings on them.
 
They are likely exos but binned as barracudas due to being external and cooking in that box if used too much. I think they should be mostly fine if they are run at good temps. Read from few posts on reddit that the most of 24+ TBs expansion drives are exos that didn't pass the strict standards for data centers so they bin them as barracudas, + they are in external box that run hot if on all the time lowering lifespan, hence lower ratings on them.
Makes perfect sense. Temp isn't an issue in NVR but multiple drives in RAID can be. Paying twice for 24TB WD purples stings. Losing recordings is worse with some sites.
 
They are likely exos but binned as barracudas due to being external and cooking in that box if used too much. I think they should be mostly fine if they are run at good temps. Read from few posts on reddit that the most of 24+ TBs expansion drives are exos that didn't pass the strict standards for data centers so they bin them as barracudas, + they are in external box that run hot if on all the time lowering lifespan, hence lower ratings on them.

I would not use them in the external enclosure they ship in. The first 26TB I did a full smart test on and it alerted on temp after a few hours. Now I just do a quick test to look at the smart stats and then shuck them.
 
I would not use them in the external enclosure they ship in. The first 26TB I did a full smart test on and it alerted on temp after a few hours. Now I just do a quick test to look at the smart stats and then shuck them.
Yea I have 5 wd red pros in my nas I am about to throw this as 6th drive. I will lose 4TB since they are all 24TB. but my nas actually kees drive surprisingly cool, running in 30s. I would keep it in external enclousre connect it via usb but heard terrible things about temps in that enclousre.
 
This is why I only use an external enclosure with a fan on it. I learned how hot these got back in the IDE days when they were all the rage.
 
Not quite as good a deal but BestBuy has these for $300 at the moment.
 
Last edited:
My 7 should be here on Thursday, apparently. I'll get them shucked and added to my pool. As more sales come up, I'm going to slowly swap out my 8TB drives with these. After next week though I'll have 324TB (usable) in the array.
I don't have a huge amount of pressure to replace the disks, but some of the oldest ones are really getting up there in age, some with >75k power on hours. Still passing smart, but, might as well slow roll the replacing so I don't have to spend a lot at once or in an emergency failure scenario.
 
Shows $329 for me and doesn't come with the backup software.

Not quite as good a deal but BestBuy has these for $300 at the moment.

Well it's actually cheaper now AND it comes with the backup software. I don't know if it's any good but better than nothing. I've been using Paragon forever but it's kinda slow and a chore to use
 
This is why I only use an external enclosure with a fan on it. I learned how hot these got back in the IDE days when they were all the rage.
That's a good point. I kinda want to downgrade my tower "server" with 4 drive raid down to a mini pc and 1 external usb drive like this. It wouldn't be on 24/7, but could be on for 8+ hours at a time... and if these things don't have any cooling that sucks.

I have a WD external 8tb that I use for backups. It doesn't have a fan either, and it doesn't complain when I'm doing backups all day... but that's once a month at best
 
Shows $329 for me and doesn't come with the backup software.



Well it's actually cheaper now AND it comes with the backup software. I don't know if it's any good but better than nothing. I've been using Paragon forever but it's kinda slow and a chore to use
It shows $329 because you're looking at the wrong size. The 26TB are out of stock now, so you missed out entirely.
 
Shucked and tossed into my zfs pool.
Code:
NAME        SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOT
data        360T   167T   193T        -         -     4%    46%  1.00x    ONLINE  -

360T should last me awhile, but I do have a couple 8TB drives that I may want to replace sooner, so that entire vdev will get expanded with 20TB+ drives. We'll see what other sales I can find between now and the end of the year.
 
lol no. What a colossal waste of time and does not at all prove longevity of a storage device. Ship it and let it fail if/when it fails. Use it properly with redundancy and a real backup and a drive dying is entirely a non-event.
yes, always have backup whatever it might be. so many options. the worst outcome is time wasted coping your data onto a new drive from backup.

always amazed me when i was younger. HDD fails and friend/family would want me to somehow magically get the data off the drive? didn't want to pay for recovery service. I would always ask if they had a backup. when they predictably said no, i'd tell them that it couldn't have been important if they didn't bother with a backup.
 
Shucked and tossed into my zfs pool.
Code:
NAME        SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOT
data        360T   167T   193T        -         -     4%    46%  1.00x    ONLINE  -

360T should last me awhile, but I do have a couple 8TB drives that I may want to replace sooner, so that entire vdev will get expanded with 20TB+ drives. We'll see what other sales I can find between now and the end of the year.
What do you store that requires 360tb?
 
That's a good point. I kinda want to downgrade my tower "server" with 4 drive raid down to a mini pc and 1 external usb drive like this. It wouldn't be on 24/7, but could be on for 8+ hours at a time... and if these things don't have any cooling that sucks.

I have a WD external 8tb that I use for backups. It doesn't have a fan either, and it doesn't complain when I'm doing backups all day... but that's once a month at best
This is what cheap 2-bay NAS units are great for--stick in one nice big drive and set the fans to 100% and no worries.
 
28TB is now $299.99 this morning at Seagate's site. Really debating do I pull the trigger on 8 of these drives for my NAS or not. lol

I wish the 10% coupon welcome discount still worked. Would be a nice chunk off. Probably just wait until Black Friday I suppose.
 
Last edited:
28TB is now $299.99 this morning at Seagate's site. Really debating do I pull the trigger on 8 of these drives for my NAS or not. lol
I wouldn't. These are going to be desktop series drives at this point (Barracuda).
 
I wouldn't. These are going to be desktop series drives at this point (Barracuda).

They're all just re-labeled Exos anyway, it's not like they're cranking out different HAMR drives for these sizes. At this point the Barracuda branding for Seagate is just like a white label WD.
 
Yeah, so far, they're just fine in my ZFS storage. I have a lot of cooling there though, and they're 7200 rpm drives so more power and heat than I'd like (all my other drives in there are 5400 rpm). Of course it's only been a little over a week, but, any large issues would've cropped up by now. Happy enough with them, especially at the price point.
One thing to note, when you shuck, there's a sticker over the SATA/power plug, and when you remove it, it tears off. I tried a couple of times to see if I could get the enclosure's plug out without ripping it, and there's just no way. So, you have zero warranty with these when shucked, even if you keep the enclosure.
 
They're all just re-labeled Exos anyway, it's not like they're cranking out different HAMR drives for these sizes. At this point the Barracuda branding for Seagate is just like a white label WD.
I wouldn't want to bet on this for a large array.
 
I wouldn't want to bet on this for a large array.

I have an 6 disk ZFS RAID Z1 with them now for going on 6 months as a backup, we'll see how long they last. Prod is an 8 disk RAID 6 array with all WD white labels and have been good enough for my needs.
 
I wouldn't want to bet on this for a large array.
I'm curious on why? I've been running desktop, and cheap, disks in my ZFS pool for over a decade now. Outside of normal failures from things being used beyond their design, they've been fine - and ZFS buys you a lot of leeway on a failure, too. I run raidz2 on each vdev, similar to RAID6, so I could lose two disks in each vdev, meaning if the "right" disks failed, I could lose 10 disks total without any data loss.
 
Out of curiosity what is the rebuild time on something that large should a drive die?
that's a valid question, I am also curious

on my raid 5 - 4 x 6tb with a perc h700 raid card it takes about 4 hours to rebuild 1 drive. That is with no in progress usage, if the raid is online and actively being used that can make a big impact on rebuild time.
 
Out of curiosity what is the rebuild time on something that large should a drive die?
On mine? It depends on which vdev has a failure. My tiny vdevs have 8TB drives, a resilver takes just about 12 hours on it. The 16TB based vdevs take somewhere around 18-20 hours. I haven't yet had to resilver with the 26TB drives, but, I'm expecting those to be somewhere around 30hr mark, give or take.
I could actually make improvements to my system to make that faster, but, I have never lost more than 1 disk at a time, so it's never been a priority for me to see how fast I could make it go. I don't let it slam the disks or backplane during resilvering, cos I still need the array to be as performant as possible, so resilvering basically gets only headroom I/O, as one example.
 
On mine? It depends on which vdev has a failure. My tiny vdevs have 8TB drives, a resilver takes just about 12 hours on it. The 16TB based vdevs take somewhere around 18-20 hours. I haven't yet had to resilver with the 26TB drives, but, I'm expecting those to be somewhere around 30hr mark, give or take.
I could actually make improvements to my system to make that faster, but, I have never lost more than 1 disk at a time, so it's never been a priority for me to see how fast I could make it go. I don't let it slam the disks or backplane during resilvering, cos I still need the array to be as performant as possible, so resilvering basically gets only headroom I/O, as one example.
what raid are you using? 6/10 ? 5 would be sketchy with that kind of rebuild time.
 
Back
Top