Best Buy WD 10TB w Flash Drive $169.99, now $159.99

[svtkobra7]

Nice. On the bad-blocks tests, can you do them on more than one drive simultaneously on the same machine? Would I need a USB3.0 hub / multiple host ports, or could you even plug these in to a USB2.0 port if you are only running the bad block tests? Actually going to need a power strip that can do that many wall-worts now that I think about it....

• You absolutely can run bb in parallel. I have a script that I personally like, but know the same individual that created that video also created a burn in script that you may want to take a look at. His solution may be cleaner, but with the script I use, I open 12 tmux sessions [1] to ensure burn in continues in the event of connectivity loss, and check in from time to time.
  
• Some suggest that its a good idea to test via USB prior to shucking in the event you run into a bad drive, so you don't have to deal with sticking them back in the enclosure for return / RMA; however, with 7 - 12 drives in our respective cases, that becomes a complete PITA.
  
• I personally think it more appropriate to shuck them, drop them into caddies, load them into the server, and enjoy the efficiency of that approach. That being said, I have complete confidence that I could get the HDD back in the enclosure if need be. You become a veritable shucking pro after a few .
  
  • O/c not everything you read on the internet is not true, but there have been recent reports here (It's A Digital Disease!) of RMAs without the enclosure and no issue encounter.
• And hearing power strip makes me cringe, please do yourself a favor and protect your investment properly via UPS!

Do you do something separate for burn-in or are you counting these tests as the burn-in?

• I tend to be a bit sarcastic and it can be tough to tell when I'm joking, but due to the density, SMART + 0xaa (10101010) + 0x55 (01010101) + 0xff (11111111) + 0x00 (00000000) + SMART = a full week, literally.
• I think most would agree that is a sufficient burn in period.

I would expect at least one of them to be bad, you think?

• Disagree - in fact - statistically speaking, I'd be surprised.
• I did a dozen in November as well - not a blip to speak of.

I have some time because I am still building the server these are going to go in. So even if it takes a month to do this it will be OK.

• You have more willpower than I - thats like staring at presents under the Christmas tree for a month!

(I hope that is some help)

 

[jcl333]

• Some suggest that its a good idea to test via USB prior to shucking in the event you run into a bad drive, so you don't have to deal with sticking them back in the enclosure for return / RMA; however, with 7 - 12 drives in our respective cases, that becomes a complete PITA.

Yes, I was planning on going the USB route, until I am more confident of putting them back in the enclosures at least.

• And hearing power strip makes me cringe, please do yourself a favor and protect your investment properly via UPS!

No no, this would just be for running the tests. When I put them into service I have an APC double-conversion UPS, network managed, the whole deal. This gives me both backup and power filtering. My house has Tesla Power Walls for backup power as well. I have also been known to throw a few Isobar's on there to cut down on noise.

For my use case, I could very likely get away with USB, but I put them in a proper server for better cooling, I would not trust those USB enclosures under any kind of stress.

My theory is that WD has some kind of test they perform on REDs, and the ones that fail become USB enclosures. So they don't mind because it is better than throwing them away or recycling them.

• Disagree - in fact - statistically speaking, I'd be surprised.
• I did a dozen in November as well - not a blip to speak of.

• You have more willpower than I - thats like staring at presents under the Christmas tree for a month!

(I hope that is some help)

OK, well don't jinx me. I agree with the timing thing, but I am already used to things happening slowly like this on my QNAP, so I just set my expectations accordingly.

-JCL

 

[Marsh]

Yes, I was planning on going the USB route, until I am more confident of putting them back in the enclosures at least.

You really should check out the 10TB drive before removing it from the USB external case.
At least , check the drive ID, smart info, run a short or long smart test before shucking the drive.

The reason is :
If you follow Reddit, there were many reports that some ( thief / jerk ) replace the 8/10TB drive with an old drive,
put the USB case back together , return the fake 10TB drive to BB for full refund.

 

[svtkobra7]

Yes, I was planning on going the USB route, until I am more confident of putting them back in the enclosures at least.

• I think you will find you get the hang of it pretty quickly - I'm not the most delicate / patient person on the planet and I didn't break a single retention clip on the first batch. I have faith in you!

No no, this would just be for running the tests. When I put them into service I have an APC double-conversion UPS, network managed, the whole deal. This gives me both backup and power filtering. My house has Tesla Power Walls for backup power as well. I have also been known to throw a few Isobar's on there to cut down on noise.

• Good deal ... I may have misread your post.

For my use case, I could very likely get away with USB, but I put them in a proper server for better cooling, I would not trust those USB enclosures under any kind of stress.

• 100% agree; however, they do run pretty cool ...
• My baseline = 7200 non-He drives though.

My theory is that WD has some kind of test they perform on REDs, and the ones that fail become USB enclosures. So they don't mind because it is better than throwing them away or recycling them.

• I strongly suspect that is the case as well, but the question I have is how would you "bin" hard drives like a manufacturer might with CPUs. I don't have the technical knowledge to answer that myself.

OK, well don't jinx me. I agree with the timing thing, but I am already used to things happening slowly like this on my QNAP, so I just set my expectations accordingly.

• OK - consider yourself un-jinxed.
• One thing to potentially consider ... I would think that if you find yourself with a bad drive, but still within the exchange / return period, you just head back to BB and get a new HDD, right? But if outside of that window, you have to deal with an RMA / likely receive a refurb as a replacement (not sure if that bothers you) and that window is 15 - 45 days depending on whether you are subject to standard or BB Elite Member policy.

 

[svtkobra7]

You really should check out the 10TB drive before removing it from the USB external case.
At least , check the drive ID, smart info, run a short or long smart test before shucking the drive.

The reason is :
If you follow Reddit, there were many reports that some ( thief / jerk ) replace the 8/10TB drive with an old drive,
put the USB case back together , return the fake 10TB drive to BB for full refund.

BAH - those hooligans are putting non-He drives back in there, so you just weigh the drives before you unbox and any "heavier" drive = air and not a white label.

(obv kidding )

I don't disagree with your advice one bit and its always smart to "err" conservative (if you will), but wouldn't the likelihood of that be excruciatingly low given he used ship to store? I would think the entire "chain of custody" = BB in that scenario, with no opportunity for a hooligan to do ever put their hands on it. I'm assuming that BB doesn't take returns and send them back upstream their supply chain and rather returns stay in store. I'm asking because I don't actually know, but I'm curious.

 

[jcl333]

BAH - those hooligans are putting non-He drives back in there, so you just weigh the drives before you unbox and any "heavier" drive = air and not a white label.

(obv kidding )

I don't disagree with your advice one bit and its always smart to "err" conservative (if you will), but wouldn't the likelihood of that be excruciatingly low given he used ship to store? I would think the entire "chain of custody" = BB in that scenario, with no opportunity for a hooligan to do ever put their hands on it. I'm assuming that BB doesn't take returns and send them back upstream their supply chain and rather returns stay in store. I'm asking because I don't actually know, but I'm curious.

True, mine came from the warehouse basically, so not an issue unless the employees do it.

Pretty sure returns go back to the manufacturer, they don't do anything with them.

-JCL

 

[Marsh]

I don't know but why risk it.
Chances may be less than 0.001 percent , but if it happened to you, then it is 100%

I buy 4 to 8 at a time, I unbox 4 drives, plug in 4 USB and power cables.

Fire up a Window machine, check smart data, then run hdsentinel pro.
Takes 5 min to start the 28 hours long testing.

 

[jcl333]

• I strongly suspect that is the case as well, but the question I have is how would you "bin" hard drives like a manufacturer might with CPUs. I don't have the technical knowledge to answer that myself.

Well, I don't know. But, hard disks are complicated enough that it would not surprise me if there was a 50' long machine in the factory that made this determination, and out come reds on one side and whites on the other.

• OK - consider yourself un-jinxed.
• One thing to potentially consider ... I would think that if you find yourself with a bad drive, but still within the exchange / return period, you just head back to BB and get a new HDD, right? But if outside of that window, you have to deal with an RMA / likely receive a refurb as a replacement (not sure if that bothers you) and that window is 15 - 45 days depending on whether you are subject to standard or BB Elite Member policy.

Well, here is a thought, considering these drives were $30 off (and I saved another $10.63 in tax), you could make an argument that it might be worth considering paying the $20 for 2-years geek squad protection. This overlaps the 2-years you get from the mfg., so you would never have to worry about sending them back, just swap it at best buy. I am going to think about this.... $140 for that on all 7 drives, might be slightly cheaper for "Elite"... Still less than the $284 I saved with the sale ($984 compared to list price).

But, at the very least yeah, get the testing done within the normal return period. Sounds like the mortality rate beyond that will be low.

Also, depending what is (was) on the drive, unless it was part of a RAID array and unintelligible, I am not going to send a drive back. I am going to smash it with a hammer and just buy another one. You never what someone might do with a presumably faulty hard drive that still has perfectly readable data on it...

-JCL

 

[BLinux]

Nice. On the bad-blocks tests, can you do them on more than one drive simultaneously on the same machine? Would I need a USB3.0 hub / multiple host ports, or could you even plug these in to a USB2.0 port if you are only running the bad block tests? Actually going to need a power strip that can do that many wall-worts now that I think about it....

Here's my video on how to use my "bht" script (mentioned here https://forums.servethehome.com/ind...ng-script-for-linux-on-sth.21511/#post-200137) to burn in several hard drives at a time:

Does require you boot Linux and install a few dependencies (see README on github).

 

[BLinux]

@jcl333 I do recommend going direct connection (SAS/SATA) versus USB3.0. I use to use USB3.0 hub, but it's hard to get max bandwidth during the burn-in, which prolongs the already very lengthy process, sometimes by 2~3x. The concerns about getting something you might want to return and accidentally breaking the casing is legitimate, but here are some mitigating factors:

1) If you follow my video that @svtkobra7 posted, the chances are really, really slim that you break something.

2) If you do break something, usually it's the tabs on the outer casing of the USB drive. This casing is exactly the same as all the others. The serial number is on the inner casing, and that's the part that matters. So, if you do break the outer casing, and it turns out the drive is bad, just use the outer casing from another drive to put it back together. As long as the HDD's serial number and the printed serial number of the inner casing match, I can't imagine there will be a problem returning/exchanging it. If you're a BestBuy "Elite Pro" member, and you're likely to become one if you're buying a lot of these HDDs, they are usually pretty quick to make the customer happy without asking a lot of questions. Be sure to sign up for the Best Buy rewards if you're spending thousands on these HDDs...

 

[svtkobra7]

Well, I don't know. But, hard disks are complicated enough that it would not surprise me if there was a 50' long machine in the factory that made this determination, and out come reds on one side and whites on the other.

• LOL

Well, here is a thought, considering these drives were $30 off (and I saved another $10.63 in tax), you could make an argument that it might be worth considering paying the $20 for 2-years geek squad protection. This overlaps the 2-years you get from the mfg., so you would never have to worry about sending them back, just swap it at best buy. I am going to think about this.... $140 for that on all 7 drives, might be slightly cheaper for "Elite"... Still less than the $284 I saved with the sale ($984 compared to list price).

• Or you could keep your savings to fund the path to 1 PB and pay with an AMEX CC, which gets you
  
• an extra year warranty through an AMEX subsidiary (and of course they don't have a supply of hard drives, so they just pay you the cost of the drive should one fail).
• for $0. Just throwing ideas out there.

Also, depending what is (was) on the drive, unless it was part of a RAID array and unintelligible, I am not going to send a drive back. I am going to smash it with a hammer and just buy another one. You never what someone might do with a presumably faulty hard drive that still has perfectly readable data on it...

• Encrypt your array? This becomes a moot concern (although others created).

Pros and cons. Pros and cons.

 

[ArthurA]

Nice. On the bad-blocks tests, can you do them on more than one drive simultaneously on the same machine? Would I need a USB3.0 hub / multiple host ports, or could you even plug these in to a USB2.0 port if you are only running the bad block tests? Actually going to need a power strip that can do that many wall-worts now that I think about it....
-JCL

If you search techbargains for jackyled there's currently (I just checked it's still valid) a promo code for an 18 outlet tower power strip that's $25.36, typically $42.99. I have three of this style and each socket can handle a large wall wart.

 

[svtkobra7]

@jcl333 I do recommend going direct connection (SAS/SATA) versus USB3.0. I use to use USB3.0 hub, but it's hard to get max bandwidth during the burn-in, which prolongs the already very lengthy process, sometimes by 2~3x.

• Completely agree.
• For me personally, the likelihood that I knock a HDD in an enclosure off the table, the dog decides he likes the taste of USB cables, or the fiance decides to use an AC adapter powering the enclosure is orders of magnitude greater than finding a non WDC WD100EMAZ-00WJTA0 inside. Not knocking the advice - and if 1 /2 drives I might feel differently if not for that long week of waiting ...
  
• I will sleep well at night knowing that my hard earned EasyStores are in a secure environment during such a crucial developmental period in their lives. I pity the fool that dare disturb them and encounter the wrath of my non-lethal threat mitigation schema (OK, fine its just a lock).
  
• Is SMART Extended slower via USB too? @Marsh noted a 28 hour test, but I just looked at logs from the last burn in and they are ~19 hours for SMART Extended.

 

[Marsh]

Is SMART Extended slower via USB too? @Marsh noted a 28 hour test,

hdsentinel pro have various surface tests,
such as write test, read test , write then read test,
Read + WRITE + read test (refresh data area) ,
Reinitialize disk surface (test for weak sector ).

For 8TB disk , the surface test ranging from 12 hours to 50 hours , I usually pick the write then read surface test.
I run 4 disks concurrently , if I have more table space, I would run 8 disks.
Then I finished the stress testing with long smart test, another 18 hours.

 

[BLinux]

• Is SMART Extended slower via USB too? @Marsh noted a 28 hour test, but I just looked at logs from the last burn in and they are ~19 hours for SMART Extended.

The SMART extended test is mostly redundant and less effective than a full run of badblocks. The "extended" or "long" test is basically the short test, with a full 1-pass surface "read/verify" scan. Since badblocks does a "write/read/verify" scan, and 4x times over with different patterns, badblocks is more extensive than what the "extended" test will do. This is based on information I gathered from T10, so I consider it reliable and not "some nameless tech forum guru" on the internet. I think what makes the most sense is a SMART "short" test, to get the testing that are not part of the surface "read/verify" scan + badblocks. the "short" test is identical to the "long" test except that the "read/verify" scan is time limited, and with today's large drives, usually means only a small portion of the beginning of the disc is tested.

 

[jcl333]

Hello all,

Thanks for all the replies and information. Rather than reply to each one, I will keep you all in the loop when the drives arrive. Looks like 5 are ready to pick up and the other 2 next week.

On the USB vs. direct attach issue. It's funny, USB3 should have plenty of bandwidth on even one bus to handle multiple drives before it became saturated, and it is the first USB standard that is full duplex. Of course, it depends on how many buses you have, how good the chip is, and in this case how good the SATA to USB3 converter in the external drive enclosure is.

These drives appear to clock in at about 200MB/sec, so under ideal conditions it would take around three to saturate a USB3 channel/bus, I am sure someone has benched these somewhere.

-JCL

 

[Marsh]

USB or direct attach for normal usage....

It is not about USB bandwidth.
The disk drive inside the external case would hit over 50 degree C.
It is not a viable option for long term drive health.

I would remove the disk drive from usb case, install the drives into a proper chassis with fans.

 

[Northern]

Agreed, these enclosed drives get way to out for a complete media burn in. Any hard drive in an enclosure that I "burn in" I always have a fan blowing on it to keep them cool... errr make them cooler.. errr make them not so hot...

 

[Terry Kennedy]

Well, I don't know. But, hard disks are complicated enough that it would not surprise me if there was a 50' long machine in the factory that made this determination, and out come reds on one side and whites on the other.

QA is a bit different with Helium drives than it is with traditional drives. Once the HDA is sealed (welded), it isn't getting opened again - it just isn't cost-effective. I also believe that there are only a small number of HDA variants in a given series - pretty much "number of platters" and "spindle speed". I suspect that all 7200 RPM 10TB drives get the same HDA, regardless of SAS / SATA / whatever. The logic boards are different and then the firmware that is loaded on those boards determines what exact sub-model the drive comes off the line as. That lets the manufacturer make the "what will it be when it grows up" decision much later in the assembly process than if HDAs were model-specific.

And, given the above, a HDA either meets the spec or it doesn't - SAS isn't "harder" on an HDA than SATA, nor is sector size. The HDAs are likely tested at several points before they are sealed, with one or more of those tests being identification of defects (likely in "bytes from index" form) so that the defect list can be written to the completed drive later on. Any HDA that didn't meet the spec before it was sealed probably goes to a separate rework line - sending them back to some stage of the regular assembly line would disrupt the normal flow of that line.

Even before Helium drives came out, manufacturers often had rooms full of trays of HDAs waiting for rework. These could be problems found during manufacturing or drives returned from the field that were tested to the point where they knew the problem was in the HDA. Those trays would often sit around for a very long time while the manufacturer decided if it was more cost-effective to rework them or if they should be scrapped as a write-off. One reason for this is the risk to the rework line if a head-crashed drive gets opened in the clean room and contaminates the room. With pre-Helium drives there was usually an inspection port where the HDA could be visually inspected before fully opening it, but with Helium drives any inspection port would leak, so the HDAs are permanently sealed.

 

[Joel]

• orders of magnitude greater than finding a non WDC WD100EMAZ-00WJTA0 inside.

Generally speaking, as long as the capacity is correct you're virtually guaranteed to get the correct drive. If someone's shady enough to swap the drive out, they won't put another 10TB drive inside, they'll put a 500GB WD Blue or whatever they had laying around. And btw, there have been some reports of pre-shucked drives with shrink wrap on them too, indistinguishable in the store from one that's not "pre-bought."

Also, I'm pretty sure the drive model number and serial number is available on the SMART output.