It’s SMR, it’ll be slow as hell and rearrange itself sometimes for days if you don’t let it sleep. But there is no choice for 2.5" drives, at least if you want a somehow large size (and even for many worthless smaller sizes nowadays).

That would work. Actually if it’s a constellation that supports TRIM (OS-Filesystem-whatever it sees on the USB - see this to get an idea how complex things can get) reading the saved backup might be equivalent to reading the whole SSD. Even if you used only 64 GBs of 1TB if the rest is TRIMed nothing (more) would be “really” read even if you do a full badblocks (or dd to /dev/null or any other full read test). Sure, it’ll take a while to feed 900+ GBs of zeroes (or whatever the TRIMed sectors return) over USB but not much will be really read from the SSD.

Give it a full read, which you should anyway to check your backups?

The point is you could run out of space anywhere, and if you’re suggesting 50GBs is a lot and you’ll never run out of space I think you’re preaching to the totally wrong choir.

That sounds like an application issue, you can run out of space locally as well (especially with the Macs that are shockingly stingy with the storage in the basic configurations and cost tons to buy the devices with more, and the SSDs were soldered even when they were more like PCs than phones). Backups wouldn’t help to recover some work that was never saved.

These drives have been on the market since early 2021, sure they’re still kind of new but with 5 years warranty (and they decided since recently to actually offer warranty to consumers in Europe too) shouldn’t be that much of a risk.

It’s probably the cheapest 20TB Seagate, with what it could be replaced, with something more expensive (not that there would be much difference)?

He he, one of the drives that started SMRgate. At least it should have 3 years of warranty, but you might have trouble to get it in your region.

NAS drives are the “shut up about your disk being slow, your gigabit is even slower” category (back when they were introduced most NASes couldn’t even fill up the gigabit, if they had it at all). That is if anyone asks how they’re different from the “DAS” and “Server” category. That somehow the marketing was so successful that now they’re considered superior to the others is another story.

As usual the answer is of course rclone. It even has --drive-skip-gdocs switch. Just rclone move --drive-skip-gdocs gdrive_remote:whatever_directory_if_any_at_all some_local_place

Are we counting here only the submarined ones? Because I don’t see the Archive ones where they are starting with “key features” “Host aware, optimized for SMR performance and capable of ZAC command support”. Sure, one might argue these were folded into the (OG of submarining) Barracuda Compute line, but still they’re different drives.

people told me it wouldn’t

It is beyond belief how misinformation gets propagated over Internet, at least here it’s something more subtle, but it’s really weird how people constantly come asking about 3.5" if they are shuckable at all (as in if the internal drive is “normal”, not that they would want the data, they’d like to just shuck the drive from the start). What’s more people even manage to gaslit themselves that somehow in the past they’ve had personally some 3.5" external with the USB directly on the PCB!

Any medium can fail, and if you need something and get it easily any other way you should have backups, not micromanaging which one fails more, CMR or SMR.

HOWEVER, when discussing cost effective and hoarding the best price/TB is by far for larger disks and larger disks aren’t SMR, problem solved.

Will rot away on anything, use or no use, at least in this universe. It’s called entropy. Have multiple copies, check them periodically and refresh as they die.

This is why you ALWAYS need INDEPENDENT backups. You can think all day long about detecting bitrot, and how well you’re protected against X drive failures but then something comes from the side and messes up your data in a different way than you’ve foreseen.