[Deadish - up to $80] [Warm] Refurbished HGST Ultrastar HE10 10TB SATA HUH721010ALE601 - eBay $70

(tho I wouldnt say "no" vibration when fired up wink wink)

Perhaps a bit of YMMV. I actually had trouble on the first one detecting that it spun up when I was checking if I needed to use the included adapters (I did not). This was also just running on the bench, don't know yet how it would sound in the case. Also don't know how loud it will be when accessed, but at idle they are very quiet.
 
Also said 14TB drives have 512mb cache vs 256 with the 10TB drives. Kinda wish I didnt jump the guns on the 10TB model. Oh well. Considering I got the coupon $63 for an enterprise 10TB drive with a warranty, dont feel like I got burned or anything.
For almost all use cases that extra cache will be no real benefit unless you have some large enterprise setup with many many of these.
 
My 10TB has checked out so far. Did full format. Now running full surface scan. 48 power ons running 5 years.

I also hear the “chunk” noise every 5 seconds or so. Not sure what that’s about.
 
For almost all use cases that extra cache will be no real benefit unless you have some large enterprise setup with many many of these.
Actually, it's for certain uses cases that the larger cache can't help (large contiguous files)--for most uses cases where the file size is < cache size, there's a performance gain for sure.
 
Actually, it's for certain uses cases that the larger cache can't help (large contiguous files)--for most uses cases where the file size is < cache size, there's a performance gain for sure.
This is very true. It does have its use cases but generally it’s an enterprise feature and specifically excels for certain data sets. I’m sure some home users may benefit also in some use cases.
 
My 10TB has checked out so far. Did full format. Now running full surface scan. 48 power ons running 5 years.

I also hear the “chunk” noise every 5 seconds or so. Not sure what that’s about.
The "chunk" noise, especially if it's repeating, is generally a physical issue with the drive. I would return it.
 
Mechanical drives are weird tamales! The various noises at random, etc. I remember Seagate's STIR (Seek To Improve Reliability) on the X15 ones. IBM's 10K UltraStar models made a "cat meowing" sound, perfectly normal and many of those sub 100G UWSCSI 10K spinners are still working today (over 22 years in service).

RE: parked/power down. Yes the drive definitely can take much more movement/impact as opposed to running. The ratings in g force are in mfr specs. Bubble wrapped drives is NOT acceptable form and in fact manufactures state this clearly in RMA instructions and will REJECT and void all warranties when returned like that. Proper packaging holds the drive firmly with lots of space in between the outer carton walls. This ensures survival under typical shipping/handling routines.
 
This is very true. It does have its use cases but generally it’s an enterprise feature and specifically excels for certain data sets. I’m sure some home users may benefit also in some use cases.
Cache always helps--especially with writing. I still remember back in the day creating 16MB smartdrv caches in dos and the awesome improvements that came from it. :) Today's cache hits are more subtle, but they are there for sure. Case in point how I use them is when I have to replicate data from one source to two destinations, I will use the command prompt and 'for' and 'start robocopy' so that the copy reads from the source and copies to both destinations simultaneously. If done right, all the reads from the source are cached so they can be served up to both drives at full speed. This even works 3 way too, or even more in theory.
 
The "chunk" noise, especially if it's repeating, is generally a physical issue with the drive. I would return it.
Not necessarily as some drives do this as a periodic self-test when idle. A standard seek on an enterprise drive generally sounds the same as what would be trouble on a consumer drive.
 
Mechanical drives are weird tamales! The various noises at random, etc. I remember Seagate's STIR (Seek To Improve Reliability) on the X15 ones. IBM's 10K UltraStar models made a "cat meowing" sound, perfectly normal and many of those sub 100G UWSCSI 10K spinners are still working today (over 22 years in service).

RE: parked/power down. Yes the drive definitely can take much more movement/impact as opposed to running. The ratings in g force are in mfr specs. Bubble wrapped drives is NOT acceptable form and in fact manufactures state this clearly in RMA instructions and will REJECT and void all warranties when returned like that. Proper packaging holds the drive firmly with lots of space in between the outer carton walls. This ensures survival under typical shipping/handling routines.
I miss the reliability of the old UWSCSI drives--when they said 1M MTBF that wasn't an estimate, it was for real! Very curious about the 22yr in service drive model numbers and their poh and load/unload start/stop stats if you can share.

And very good point about the bubble wrap! I forgot about that in the warranty documentation by both WD and Seagate. They are very, very particular about this point too. Bubble wrap generally can't keep a drive from hitting the box walls unless it's a whole wad with a drive in the center (which these do sound like they were shipped that way). Still, the ultimate if you're going to ship a drive with bubble wrap is to double box it.
 
Not necessarily as some drives do this as a periodic self-test when idle. A standard seek on an enterprise drive generally sounds the same as what would be trouble on a consumer drive.
Yes, drives make noises, but he said it's doing it every five seconds and doesn't stop, but we need more information. It sounds like a physical drive issue. It could be something else, but that doesn't sound good.
 
Yes, drives make noises, but he said it's doing it every five seconds and doesn't stop, but we need more information. It sounds like a physical drive issue. It could be something else, but that doesn't sound good.
I would run a smart short test on it. Usually ferrets out issues like this.
 
I miss the reliability of the old UWSCSI drives--when they said 1M MTBF that wasn't an estimate, it was for real! Very curious about the 22yr in service drive model numbers and their poh and load/unload start/stop stats if you can share.
Those were IBM 10k U320 (IC35L036UWDY) drives hosted on a Mylex ExtremeRAID 2000 HBA with 128MB cache with battery backup. Entered service in late 2002. Running Windows 2000 Advanced server for an accounting system. System was never turned off and diskeeper was scheduled to keep it defragged which particularly during cache dumps sounded like BBs being poured into a frying pan! Did not have SMART specifics but never heard a peep from the controller and the two hotspares in the pool were never called into action for rebuilds. The drives weren't in hotswap bays either just wired with shielded round 320 cabling with good terminators at the end.

And when I say never turned off, it wasn't! Facility used Ferrups 3.1kVA UPS that was well maintained and there were dual diesel backup gennies so power was never, ever interrupted over two decades. Lots of equipment that was never power cycled/restarted for upgrades showed years of uptime. Like the AT&T Merlin phone system that ran (since the late 80s!) until it started dying.
 
I got my one 16TB drive yesterday and currently testing. Very quiet, runs cool and it was at 66% when I last checked. So far so good.
 
Those were IBM 10k U320 (IC35L036UWDY) drives hosted on a Mylex ExtremeRAID 2000 HBA with 128MB cache with battery backup. Entered service in late 2002. Running Windows 2000 Advanced server for an accounting system. System was never turned off and diskeeper was scheduled to keep it defragged which particularly during cache dumps sounded like BBs being poured into a frying pan! Did not have SMART specifics but never heard a peep from the controller and the two hotspares in the pool were never called into action for rebuilds. The drives weren't in hotswap bays either just wired with shielded round 320 cabling with good terminators at the end.

And when I say never turned off, it wasn't! Facility used Ferrups 3.1kVA UPS that was well maintained and there were dual diesel backup gennies so power was never, ever interrupted over two decades. Lots of equipment that was never power cycled/restarted for upgrades showed years of uptime. Like the AT&T Merlin phone system that ran (since the late 80s!) until it started dying.
Thank you! I guess I have the baby brother of those drives, the Seagate 2nd gen 9GB Cheetahs connected to a Mylex DAC960SUI SCSI-SCSI raid controller. Massive overkill for a desktop, but we wanted the ultimate. :D Still have it, but the Supermicro motherboard in that system needs fixing before it will boot again. It was booting but then I think the varta battery must have corroded something too far and killed the board for now. I had this on an APC 1500VA sine-wave ups and it would use up to half the load when powering the system on--back in the 1990s. Those Ferrups were the Best (pun intended)--if APC didn't come around, I best Best would still be around--well, I guess they still are since the Ferrups model name is being used by Eaton. And I guess the 3.1kva ups was good because a well used 4.3kva one is still a chunk of money today! https://www.ebay.com/itm/266834947336

Yes, systems without updates can last literally forever now. The hard part is somehow severing the cord from the Internet to allow them to 'just work'. I block all my NAS units from the outside since they're just LAN units and even old ones can work in safety for the rest of their days this way.

And those AT&T Merlins were legendary. I heard about them because I would work on our Mitel SX-50 which was a staple in the hotel industry for the better part of 2 decades. But it was also a beast, on 24x7 through a lot of stuff and pretty much bulletproof unless something went wrong in programming or the hardware.
 
As an eBay Associate, HardForum may earn from qualifying purchases.
The original Cheetah 9GB drives were crazy loud but crazy reliable too! At the time Hitachi came out with a 12K RPM Pegasus drive and it was pretty much a beta product and VERY rare today. Never had my hands on one as the Cheetah 36LP was around the corner. Quantum/Maxtor had their very nice 15K too. But the monster was the Fujitsu MAU series. So much quieter seek times. Still hard to tame the hum of a 15K spindle motor, particularly when you have dozens of them within the same chassis but no (sane) person had that many on a desktop. Now we have what amounts to a size of a stick of gum that can read and write close to 15GB a SECOND!

The idea around the ferroresonant transformer was to provide zero transfer to battery without the need to run the inverter full time like conventional double conversion units. Noisy AF (hummed like crazy) and put out lots of heat (high core loss) but talk about great quality power. Still have some SOLA conditioner units that work today but with inverters getting damn near close to 100% efficiency, double conversion is leading the way. Still though, the FerrUPS could take lightning like a bullet. 911 call centers and casinos need that reliability for sure!
 
RE: bubble wrap - the drives were in static bags and the bubble wrap was like double wrapped in a very tight fit in the box. I think it was more than adequate. Like someone mentioned far better than say newegg.

RE: clunking noise - mixed feelings about this - hearing it makes one feel uneasy but I have heard it with other previous new drives Ive worked with before. Dont rememer the models, becomes a blur when you are a datahoarder and work for an IT company. :ROFLMAO:
Cache always helps--especially with writing.
Der, I thought it helped with erasing or something. 😜
 
The original Cheetah 9GB drives were crazy loud but crazy reliable too! At the time Hitachi came out with a 12K RPM Pegasus drive and it was pretty much a beta product and VERY rare today. Never had my hands on one as the Cheetah 36LP was around the corner. Quantum/Maxtor had their very nice 15K too. But the monster was the Fujitsu MAU series. So much quieter seek times. Still hard to tame the hum of a 15K spindle motor, particularly when you have dozens of them within the same chassis but no (sane) person had that many on a desktop. Now we have what amounts to a size of a stick of gum that can read and write close to 15GB a SECOND!

The idea around the ferroresonant transformer was to provide zero transfer to battery without the need to run the inverter full time like conventional double conversion units. Noisy AF (hummed like crazy) and put out lots of heat (high core loss) but talk about great quality power. Still have some SOLA conditioner units that work today but with inverters getting damn near close to 100% efficiency, double conversion is leading the way. Still though, the FerrUPS could take lightning like a bullet. 911 call centers and casinos need that reliability for sure!
Very cool to know about the drive that came after the Cheetah. After seeing 22MB/sec in DOS on our setup, I didn't care about what was going on in storage for a while. :D

Wow, interesting. I remember SOLA! Another great name in power conditioning that was far outside our price range. That's awesome they could take a lightning hit! We needed that for our Mitel SX-50--I think we spent 2x as much as a new one over the decades of repairing that thing from storm damage.
 
RE: bubble wrap - the drives were in static bags and the bubble wrap was like double wrapped in a very tight fit in the box. I think it was more than adequate. Like someone mentioned far better than say newegg.

RE: clunking noise - mixed feelings about this - hearing it makes one feel uneasy but I have heard it with other previous new drives Ive worked with before. Dont rememer the models, becomes a blur when you are a datahoarder and work for an IT company. :ROFLMAO:

Der, I thought it helped with erasing or something. 😜
Sounds like they did bubble wrap right--if you can keep the drive from moving, you're in good shape--but once it gets momentum it will pop those bubbles. The worst is those single big air cushions--drives pop those in a second.

Yep, understand the blur for sure. :D

Only if you're doing it wrong! :ROFLMAO: :ROFLMAO: :ROFLMAO:
 
WD Ultrastar DC H530 14tb Drives for $99 on newegg by GoHardDrive. Refurbished of course...Same model that GoHardDrive is selling on ebay, along with the 5 year seller provided warranty. $7.14 a gigabyte....Not too shabby.

https://www.newegg.com/hgst-western...-01N48?Item=9SIA5ADK031508&Tpk=9SIA5ADK031508
Was seriously considering some of these but they are also dead/out of stock.

Oh well, always something bigger and cheaper gonna come along.

Edit: I take that back they are back in stock, were out of stock earlier today.
 
I picked up 2 more of these 14tb drives and just loaded them up to test. Looks like 4+ years of power on time with 10-15 power cycles each, no errors or issues reported on SMART or tests. Good deal for just under a $100.
 
I will also say that when I get a drive ghd, I will occasionally get a drive where when I do test it, it simply will not spin up at all. At one point, it took me three RMA's before I got a working drive. But once I did, I did not have any other issues.

They are very good about swapping them out. Just feel that they might not test very well before shipping out to the consumer.
Devil's advocate, but why pay labor for inhouse testing, when for the less than minimum wage (the price of shipping labels back and forth) you'll do the testing for them AND thank them for the privilege with praise about how good they are with returns? Just sayin.

They made a spreadsheet decision that testing nothing and paying return shipping for the exceptions is cheaper than time, labor, space, and inventory flow slowdown testing everything inhouse first. That's why the nervous smile behind their "easy returns", because their house of cards depends on that one thing.
 
Last edited:
Devil's advocate, but why pay labor for inhouse testing, when for the less than minimum wage (the price of shipping labels back and forth) you'll do the testing for them AND thank them for the privilege with praise about how good they are with returns? Just sayin.

They made a spreadsheet decision that testing nothing and paying return shipping for the exceptions is cheaper than time, labor, space, and inventory flow slowdown testing everything inhouse first. That's why the nervous smile behind their "easy returns", because their house of cards depends on that one thing.
That's one way to do it, but then you'd have to really trust who you're buying from and would need to do some spot checks to insure that your bet will pay off.
 
Devil's advocate, but why pay labor for inhouse testing, when for the less than minimum wage (the price of shipping labels back and forth) you'll do the testing for them AND thank them for the privilege with praise about how good they are with returns? Just sayin.

They made a spreadsheet decision that testing nothing and paying return shipping for the exceptions is cheaper than time, labor, space, and inventory flow slowdown testing everything inhouse first. That's why the nervous smile behind their "easy returns", because their house of cards depends on that one thing.

Well, they need to wipe the drives. Chances are that wiping and QA are one step. If the drive can be wiped it is declared OK.
 
Bummer. I checked SMART on all four drives when I got them, but did no testing. First drive passed full erase. Put second drive in and the machine didn't boot, hit reset and it booted find. SMART was still okay. Ran the WD quick test and the drive failed and now clicks. Could not do a clean shutdown.

Other two drives did pass quick test and are now being full erased. I guess I'll find out how good that RMA is.
 
Historically these have hit $70 semi frequently. If you want one for that price I’d just wait, it will happen again.
 
Devil's advocate, but why pay labor for inhouse testing, when for the less than minimum wage (the price of shipping labels back and forth) you'll do the testing for them AND thank them for the privilege with praise about how good they are with returns? Just sayin.

They made a spreadsheet decision that testing nothing and paying return shipping for the exceptions is cheaper than time, labor, space, and inventory flow slowdown testing everything inhouse first. That's why the nervous smile behind their "easy returns", because their house of cards depends on that one thing.
I mostly agree with this. Plugging in a drive and starting a test takes only minutes, but it is an inconvenience and takes resources and quite a wait (not like you have to "do" anything while waiting)

These drives cost less than 1/3rd of new equivalents. I think for a lot of us it’s kind of like the lottery ticket affliction, plus maybe frugalness. If you are willing to take the time to test and/or go through the stupid RMA process, you can get a great price per TB.

If you order say 10 of the 14tb and 1 is bad, that’s 140tb of very fast dynamic storage for $1000 + a bit of inconvenience. That’s a winning ticket. If 1/2 of them were bad I’d be pissed, a losing ticket.
 
  • Like
Reactions: DPI
like this
Bummer. I checked SMART on all four drives when I got them, but did no testing. First drive passed full erase. Put second drive in and the machine didn't boot, hit reset and it booted find. SMART was still okay. Ran the WD quick test and the drive failed and now clicks. Could not do a clean shutdown.

Other two drives did pass quick test and are now being full erased. I guess I'll find out how good that RMA is.
So, I got one bum one out of 4, 25% failure rate. We'll see how good the RMA is.
 
So, I got one bum one out of 4, 25% failure rate. We'll see how good the RMA is.
Followed their instructions to send an email and had a prepaid return label in 2 hours on a Saturday. Doesn't look like they cross ship which kind of makes sense as they have no way to charge a deposit.
 
So both drives are looking good.

The way I figure it I got 2x used 10tb drives for $63 each (or ~$70 if you add tax, bascially the coupon cancelling it) + spending about an hour of my time + increase to about a total of 1 week of shipping when including the RMA.

Meh, could be better, could be worse.

But I got something else: Knowledge.

Durring this week I have done some extensive researching on the refubishing of HDs, statistics at places like backblaze, insigiht / thought about the market for HD's and datacenters, lurked into other forums, knew of the 3.3v power pin but not really what it was for or the intricacies, read about/reminded my self about things like FDB (fluid dynamic bearrings), SMR (shingled magnetic recording), SAS, etc etc etc. Plus countless ads and pop ups lol. (yes I have a handful of blockers)

That in it self might be worth more than the drives.
 
Back
Top