Don't take this personally but this advice is fairly useless.
For every person like you, who has "found Hitachi to be the most reliable" and "had universally bad luck with Seagate", there's someone else who has the exact opposite advice.
Our opinions are shaped by our own experiences, of course, and (most of the tine) it's hard to change our minds or convince us otherwise.
The (unfortunate) truth is that there's no one brand of HDDs that really is "the best" -- especially for all use cases! If there were, well, we'd all be using them by now and all of the other manufacturers would have already went out of business.
Well, in any case now I bought 8 Toshiba N300 8TB (4 HDDs for a server and another 4 for a NAS, both used by a RaidZ1) and so far they worked (and they're not loud and don't vibrate like hell when the disk is spinning, unlike the WD Gold).
It's the first time that I buy Toshiba HDDs - I'm quite sure that if they would have been Seagate then at least 1 would have already failed.
I've had good luck with the few Toshiba drives I've had. That's what my storage system is running right now because it was the only thing I could find in 1TB laptop form factor when I bought them 6-7 years ago.
Thinking about it now, it might make sense.
On the other hand, nowadays there isn't a lot of choice (only 3 manufacturers? WD, Toshiba and Seagate? If yes, then it's quite sad...) and honestly for me it hasn't been easy to choose even just a single model (reading all user reviews, comparing prices, searching specs for e.g. operating temperature, mtbf, finding infos about SMR, etc...) => I felt exhausted after deciding for Toshiba/N300 and I didn't have any other good 2nd candidate... .
Let's see in another ~4 years, maybe I'll adopt this strategy. At least I ordered 2 groups of 4 HDDs from 2 different shops, so I should have at least a minimal amount of "mix" in it. And the ones used in the NAS host the backup of the ones used by the server, so I just hope that I don't have a 2+2 HDD failure in the two raidZ1 :P
Thanks for the enlightenment :)
But: in such a case, having a normal raid5/raidZ1 or even a raid6/raidZ2, etc..., you might still get scrd by just having 2/3/4 such HP-drives in such a raid (that might be composed of 8/9/10 drives), right?
That's true. Ah well, you know what people say: RAID is for high-availability, not a replacement for proper backups.
25,000+ drives according to 2020 Q1 report in case anyone is wondering
Don't regret it! It allowed you to find a drive failure before a second drive failed, and allowed you to recover your data while it was still recoverable! This is a good thing!
Archival data is a tricky problem. If you don't regularly read it, you may find that when you do need to read it, that the sector has suffered some bit rot. What's one of the most likely cases where you have to read archived data? When a hard drive fails and RAID is doing a rebuild...
So, by all means, run a ClamAV scan! Or better yet, run a ZFS scrub monthly or so. I've been a huge fan of ZFS for my archive data, mostly historic photos and Google takeout data now, because it can detect silent corruption and can do scrubs to verify and repair any hard drives sectors that have problems.