I think that three pre-clear cycles is excessive for consumer grade disks. OK, some datacentre disks are rated at 500 or so TB/year and a typical workload for NAS disks is 180 TB/year but look at the spec sheet for Seagate desktop disks and you'll see that the workload rate limit (as defined here) is a mere 55 TB/year. Three pre-clear cycles (read, write, read, write, read, write, read) shifts seven times the disk's capacity or, in the case of the latest and greatest ST8000DM0002, 56 TB. That means that it's annual workload is exceeded before it has even been put to "profitable" use. It is, quite literally, being pre-cleared to death.

That is all smoke and mirrors.  Have you ever had them refuse a drive or anyone you know have one refused?  They may have this policy but they don't want the bad press even more so it is highly unlikely they would ever decline a return because of this.  The only ones that have to worry are those people that have the drives in continuous operation reading and writing to the drive after the preclear cycle.  Anyone that has the drive spun down by unRAID 75% of the time or more has nothing to worry about because the 1 week of continuous use is counter balanced by the following 3 weeks of almost NO use.  I would not let this determine whether you trust the disk in your array or not.  You should use personal experience.  If you have never had a failure past the 1st preclear cycle then by all means only do 1 cycle.  I have had drives die in cycles 2 and 3 and I don't want to trust a 1 pass precleared drive in my array.

If anyone has had an experience to the contrary, I would appreciate if that those persons post about the problem and the manufacturer involved

Not hard disks specifically, but I know people who have failed to have printers and paper shredders repaired under warranty because the specified duty cycle had been grossly exceeded.

If anyone has had an experience to the contrary, I would appreciate if that those persons post about the problem and the manufacturer involved

Not hard disks specifically, but I know people who have failed to have printers and paper shredders repaired under warranty because the specified duty cycle had been grossly exceeded.

Printers are not the same situation.  I've seen it too on a printer but I've never had an RMA refused or heard of one on a HDD.  The only exception to that is when I tried to RMA a drive from a prebuilt laptop and didn't return the WHOLE laptop - so I pitched the drive - the laptop itself was fine.

I have had drives pass cycle 1 and 2 but die during cycle 3.

 

Seagate Desktop (DM) disks, by any chance?

 

Nope. Mostly Western Digitals. This was before I moved onto HGSTs.

Anyone that has the drive spun down by unRAID 75% of the time or more has nothing to worry about because the 1 week of continuous use is counter balanced by the following 3 weeks of almost NO use.

Not so, Bob. Those ST4000DM000 disks in your Number 5 server are using up 48 of their 55 TB annual workload limit if all they do is a monthly parity check. I didn't make the numbers up. Seagate published them in its spec sheets.

 

I have had drives pass cycle 1 and 2 but die during cycle 3.

 

Seagate Desktop (DM) disks, by any chance?

 

Nope. Mostly Western Digitals. This was before I moved onto HGSTs.

 

Neither of whom publish the figures, unfortunately.

Anyone that has the drive spun down by unRAID 75% of the time or more has nothing to worry about because the 1 week of continuous use is counter balanced by the following 3 weeks of almost NO use.

Not so, Bob. Those ST4000DM000 disks in your Number 5 server are using up 48 of their 55 TB annual workload limit if all they do is a monthly parity check. I didn't make the numbers up. Seagate published them in its spec sheets.

But as I said that is nothing to worry about.  If they go bad I will RMA them and get them replaced if they are still in warranty.  Never had them refuse a drive that is still in the warranty period and I don't expect to in the future.

Anyone that has the drive spun down by unRAID 75% of the time or more has nothing to worry about because the 1 week of continuous use is counter balanced by the following 3 weeks of almost NO use.

Not so, Bob. Those ST4000DM000 disks in your Number 5 server are using up 48 of their 55 TB annual workload limit if all they do is a monthly parity check. I didn't make the numbers up. Seagate published them in its spec sheets.

But as I said that is nothing to worry about.  If they go bad I will RMA them and get them replaced if they are still in warranty.  Never had them refuse a drive that is still in the warranty period and I don't expect to in the future.

 

Not a worry, no. But it's certainly a nuisance when something fails - something you could well do without. I'm simply advocating striking a sensible balance between testing and wearing out and choosing the appropriate disks for the job. For Seagate's NAS disks (with an annual workload limit specified as 180 TB) it simply wouldn't be an issue, even for very large (8 TB) ones.

 

Edit: The thing that sparked this off for me was trying to decide which of the several classes of 8 TB disk that have recently become available would be appropriate for me to use as a parity disk alongside 8 TB Archive data disks. I've never felt comfortable with the idea of using an SMR disk for parity, and now I have a real choice. But narrowing it down to just the two consumer models I see that the ST8000VN NAS model is quite a lot more expensive than the ST8000DM Desktop model. However, after looking at the specs and seeing the annual workload limit of each, I've come to the conclusion that the latter (at 55 TB/year) would be an extremely poor choice for unRAID, while the former (at 180 TB/year) would be much more appropriate.

 

Not a worry, no. But it's certainly a nuisance when something fails - something you could well do without. I'm simply advocating striking a sensible balance between testing and wearing out and choosing the appropriate disks for the job. For Seagate's NAS disks (with an annual workload limit specified as 180 TB) it simply wouldn't be an issue, even for very large (8 TB) ones.

In that I would agree but I don't think testing thoroughly before being put into an array to be a problem.  I don't buy standard disks any more - at least not WD or Seagate anyway.  My HGST's (cool spin 5400 type) have only had a single failure and I just got that one RMA'd and the replacement back in my array.  I've been putting WD Red NAS and HGST NAS drives in my arrays now.  I am slowly migrating off of my standard Seagate's they will be off-line backups.

I've had several DOAs on the second or third cycle but made it through the 1st just fine.  I've had very few drives die after the 3rd cycle after they are in the array.  At least none that have done so within a short time of being put into an array.  Most of my failures after 3 cycles have been after 1-3 years of use.  I would do 6 cycles if I though it would make a big difference but 3 cycles appears to catch 95% of the bad drives.

 

The preclear script apparently stress a lot more the RW heads than the other tests. The post-read rans in a loop that read the first sector, two random middle sectors and the last sector of the disk and then it reads the disk content to verify if it's being zeroed. It hammers the hard drive a lot more than the badblocks program, for an example.

 

I think if a disk passes 3 cycles, it's very improbable it will prematurity fail.