My home NAS has been running on WD Green drives that are from the 2010 era with >110K power on hours and 40K start stop cycles, 300K ramp load/unloads. Since this is a home system, I have always had aggressive spin down enabled on the drives to save power and it has never proven to be less reliable.

With these stats on the drives, I'm preparing for the inevitable by scooping up spare drives whenever I see a good deal. I recently picked up some Seagate Barracuda LP 2TB drives (ST32000542AS) for a really good deal and brand new still sealed. Despite being 15 years old, they started just fine and a full check disk tested fine. However, I just discovered they use CSS (contact start stop) which only has a rating of 50,000 that is much less than anything modern using ramp load/unload.

So I'm wondering, are CSS drives inherently less reliable since the heads are scraping on the platters on start up/shut down? If it was a desktop PC where I'd only turn it on/off once a day I may put these drives into that service, but my home NAS may be spinning up the drives several times a day. Would these old CSS drives not be considered suitable here? Any other issues with keeping 15 year old drives as spare? Or maybe I should sell these off and get something more modern, but likely with higher hours and an unknown usage history.

Technically, yes, hard drives that park their heads on the platter are considered less reliable because the heads will contact the surface and travel a greater distance in the event of impacts or drops, compared to hard drives with a ramp, which is closer to the most frequently accessed area. In fact, this 50,000 head parks is a rather conservative figure, just like the 500,000 head parks for hard drives with ramps. I have drives with 1.5 million head parks, and they have had no problems.

The 110,000 hours of operation doesn't mean anything, since this counter on WD hard drives counts the time the drive is stopped but powered. The actual lifespan of your drives is likely much lower.

Also, considering that these WD Green hard drives from the early 2010s were never considered good due to their aggressive head parking mechanism, you're not losing anything by replacing one drive of questionable quality with another of the same quality. At least you bought brand new drives.

Since you're using a NAS, I hope you're also using RAID or a similar redundant system to prevent data loss. In this case, you don't need to worry about drive quality at all; just make sure the array is in good condition, and if any drive starts to fail, simply replace it.

I'm running Openmediavault, so using SnapRAID + MergerFS. I have it configured so that SnapRAID does a sync to the parity drive only once a day. I don't really feel the need for RAID as it's overkill and a waste of power keeping the drives spinning all the time. The array also does a scrub on every sync, so in theory every 30 days the entire array should be checked for bitrot.

Yea I remember the fiasco with the WD Green head parks. I personally was too noob and lazy at the time to change the timer, so my drives are stock. Strangely, my WD10EADS drive only has 80K cycles whereas the WD10EARX has 300K...both with similar power on hours. So somewhere back then, WD must have adjusted the firmware.

I would personally like to use these newer Seagate drives for data, but if frequent spin up/downs might be an issue then maybe it's better to keep them as parity drives that only spin up/down once a day. I would imagine for a 15 year old drive, there's other considerations like lubrication hardening or the magnetism in the platters weakening, so would I even get another 10 years out of these drives if I were to put them in service now who knows.

I have some WD Green 2TB disks but stopped using as I have an even larger number disks in the 4TB and up capacity. My 18TB WD enterprise disks can gobble up a lot of games or media as needed.

_________________
USB 3.5" enclosures are $19 USD postage paid from China, I own several

The hard drive, or rather the servo information and data in the system area, became demagnetized before the hard drive left the factory. What the disk is now reading is residual magnetism, which will last it for decades. And besides, these things are written in a slightly different way than user data is recorded, at a lower data density, so if something is demagnetized there, then under equal conditions the user data will be difficult to read, but the disk itself will work fine. The magnetic layer itself does not suffer from time; it will be magnetized in hundreds of years in the same way as now. So don't worry about this.

You also shouldn’t worry about the lubrication in the bearings; there is almost no lubrication in modern discs, and what is there does not dry out, but runs out. And this mainly affects incredibly worn-out hard drives from the early 2000s. It happened that the lubricating layer came off from the surface of the platters, classic Seagates suffered from this and the problem was characterized by the fact that the reading speed decreased, and writing occurred at a speed of hundreds of kilobytes per second. Data from them was usually successfully copied to a new hard drive, and the old one was recycled.

Personally, I would not be happy with the fact that these are not so fast drives, and they have a fairly small capacity for today's demands. And the fact that they were made 15 years ago would not bother me at all. This is not a pizza that is best eaten hot. But since the price was low, this is completely justified.