That's why I buy used drives at 33% of new prices. My RMA process is to grab one of the extra drives I bought and replace the dead drive.Anyone who has ever dealt with Toshiba's RMA dept.
That's why I buy used drives at 33% of new prices. My RMA process is to grab one of the extra drives I bought and replace the dead drive.Anyone who has ever dealt with Toshiba's RMA dept.
Smart man. DIY replacements has been my strat since at least 2007 (for home storage anyway, not in a business), as once you are dealing in enough quantity then there's a crossover point where you're overpaying for warranty if cheaper variants are available.That's why I buy used drives at 33% of new prices. My RMA process is to grab one of the extra drives I bought and replace the dead drive.
Yup and then you will have people like some of our clients who will complain about a NVMe drive being 5 years old but having 99% life on it.Smart man. DIY replacements has been my strat since at least 2007 (for home storage anyway, not in a business), as once you are dealing in enough quantity then there's a crossover point where you're overpaying for warranty if cheaper variants are available.
Person A buys 20 x WD Reds for $300/ea (5yr warranty)
Person B buys 20 x WD Externals for $200/ea and voids 2yr warranty shucking
Person B could buy 30 drives for what Person A paid for 20, and have 10 sitting on the shelf as cold spares. Or buy 25 drives, keep 5 on the shelf for replacements, and save $1000. This isn't shining a light on anything previously unknown, but it's one way to go.
Some will argue "but a drive with a 5 year warranty must be inherently better quality since the manufacturer has high confidence in it." And that's not necessarily true, for reasons too boring to get into.
I think almost every business vastly over-buys on NVMe endurance, which is why we get to buy those 5-year-old drives on eBay.Yup and then you will have people like some of our clients who will complain about a NVMe drive being 5 years old but having 99% life on it.
the pain/cost of having to replace worn drives would far outweigh the cost imho, also if you end up using for more than 5 years, its sooo nice. we still got SM863a's in prod! just recycled as boot drives for new systems, and misc lab stuff etc.I think almost every business vastly over-buys on NVMe endurance, which is why we get to buy those 5-year-old drives on eBay.
This is especially true in a distributed storage system with an NVMe in each node, and the NVMe is only used for ingest from clients, and not for the data protection scheme (i.e., copying the data to multiple nodes). With 10 nodes each having just a 6.4TB NVMe with 3DWPD, that's 192TB per day new/changed data written to the storage (assuming the writes are balanced across all the nodes). If each of those nodes have 480TB of raw storage, you will completely fill all the storage in less than a month, and the NVMe will have 99.986% of life remaining. The only way to wear out a 3DWPD NVMe is to have at least 1000x as much storage as the size of the NVMe.
One "ingest" node having the only NVMe, and lots of storage behind it (racks of JBOD, etc.), then you could conceivably wear out a 3DWPD before the warranty expires, but it's still going to require 2-3PB of storage per TB of NVMe.
Most companies buy a pre-built box from Dell, HP, Cisco, etc., and when it becomes too underpowered, dump the whole thing and get a new box.the pain/cost of having to replace worn drives would far outweigh the cost imho, also if you end up using for more than 5 years, its sooo nice. we still got SM863a's in prod! just recycled as boot drives for new systems, and misc lab stuff etc.
Restocking Fee
Returns of product, other than with defect, are subject to a 20% restocking fee. This will be deducted from your refund. We also do not refund the original shipping and handling that you paid on the order.
Manufactured in week 16 of year 2019
Specified cycle count over device lifetime: 50000
Accumulated start-stop cycles: 20
Specified load-unload count over device lifetime: 600000
Accumulated load-unload cycles: 20
Elements in grown defect list: 0
Error counter log:
Errors Corrected by Total Correction Gigabytes Total
ECC rereads/ errors algorithm processed uncorrected
fast | delayed rewrites corrected invocations [10^9 bytes] errors
read: 0 0 0 0 0 8198295.010 0
write: 0 3 0 0 0 37171.280 0
verify: 0 1322 0 0 0 3206192.625 0
Non-medium error count: 0
SMART Self-test log
Num Test Status segment LifeTime LBA_first_err [SK ASC ASQ]
Description number (hours)
# 1 Background short Completed - 38876 - [- - -]
Eeeek, that smart data is concerningUpdate:
Grade B drives.
These all seem to be from about april of 2019, Average Power hours varies between 35k and 38k so they've had a life.
My concern from the cut labels only grows, the case damage is deep within "someone keyed my car!" territory.
Compared to the 12TB drives that have a 5yr ebay refurb promise for $80, I think I've got to vote that this is a pretty bad deal by comparison, even at a $65 bulk price.
A warning if people order via the website instead of ebay:
Code:Manufactured in week 16 of year 2019 Specified cycle count over device lifetime: 50000 Accumulated start-stop cycles: 20 Specified load-unload count over device lifetime: 600000 Accumulated load-unload cycles: 20 Elements in grown defect list: 0 Error counter log: Errors Corrected by Total Correction Gigabytes Total ECC rereads/ errors algorithm processed uncorrected fast | delayed rewrites corrected invocations [10^9 bytes] errors read: 0 0 0 0 0 8198295.010 0 write: 0 3 0 0 0 37171.280 0 verify: 0 1322 0 0 0 3206192.625 0 Non-medium error count: 0 SMART Self-test log Num Test Status segment LifeTime LBA_first_err [SK ASC ASQ] Description number (hours) # 1 Background short Completed - 38876 - [- - -]
the damage is only cosmetic ?like... how bad are these labels cut?
I was wondering the same, being cut seemed a little sketchy looking.like... how bad are these labels cut?
On the drives I received, the 2mm-thick case is scratched about 0.1mm deep, which is purely cosmetic.the damage is only cosmetic ?
No, it's not. Bytes written and read on spinning rust are not indicators of drive health. And, 1 ECC correction in every 4 billion verify operations is nothing. For total operations, ECC kicks in once every 16 billion operations. That's exactly what ECC is supposed to do.Eeeek, that smart data is concerning
Maybe. I'm guessing that it's a one-off. There are many mistakes in the item description of that so I even question if they were new, as listed.This listing is already sold out, but at $64 shipped, maybe it's a sign that prices are dropping:
Toshiba MG09SCA14TE Hard Drive 14TB SAS 12Gb/s 3.5in 7200 RPM 512MB Standard... | eBay
<p dir="ltr" style="margin-top:0; margin-bottom:0;">Increase your storage capacity with this Toshiba 14TB SAS hard drive. The 3.5-inch form factor makes it easy to install in your computer or server, and the 7200 RPM rotation speed ensures fast and efficient performance. With 512 MB cache, this...www.ebay.com
God, you're going to cost me $500... I am so damn tempted to get 8 of these and some cabling (already got a LSI 3008 in the server, just would need to flash it to IT mode)......There are more available for $66 shipped:
Toshiba MG09SCA14TE Hard Drive 14TB SAS 12Gb/s 3.5in 7200 RPM 512MB Standard... | eBay
<p dir="ltr" style="margin-top:0; margin-bottom:0;">Increase your storage capacity with this Toshiba 14TB SAS hard drive. The 3.5-inch form factor makes it easy to install in your computer or server, and the 7200 RPM rotation speed ensures fast and efficient performance. With 512 MB cache, this...www.ebay.com
Lucky for you they are already gone lol.God, you're going to cost me $500... I am so damn tempted to get 8 of these and some cabling (already got a LSI 3008 in the server, just would need to flash it to IT mode)......