Number 3 there is just… wrong. It shows selecting striped, bit striped is raid0, not 1, so if you selected that, that’s why. If you selected mirrored though you should indeed only see 20tb if each drive is 20. There should be no issue with having had storage spaces previously no.

Ok so, storage spaces isn’t the same as raid. A mirrored storage spaces pool, is not raid1. It’s very similar to it in that it’s a mirrored set, but it’s not the same thing. In a raid, because see, in storage spaces, you can have a mirrored set with 3 drives, and you’ll actually be able to about one and a half times one drive of data in that pool. This is because in storage spaces, it’s the DATA that is being duplicated, not the drives. So don’t confuse the concepts.

Now, as for why it shows that size, well, because you configured it to. Storage spaces completely decouple the pool size from actual currently usable space. You can create a pool of only 1 8gb drive, and yet say it’s a 1PB pool and it’ll happily do it for you. You’ll still only be able to actually store 8gb ofc, but the pool will report rhe 1PB of maximum space.

Generally the enclosures are just that, enclosures that offer the connection. There are exceptions though where enclosure does something more. Some enclosures do encryption and some just use the same controller for single drive and multidrive and your one drive is actually set up as a 1 drive raid array in which case you may have data slightly shifted to accomodate the headers for that. You can then still recreate everything, but it’s a pain.

But as I said, generally they’re just providing the drive as is in which case there won’t be any issue.

It really depends. For like a desktop, I’d avoid unless it was really cheap as I’d basically nullifies the value of all non standard parts and I’d include things like cpu if the motherboard is nonstandard. So value basically becomes only like drives and such.

For a server though, non standard is the norm and hete vendors even do stuff like vendorlocking instead which then IMO is a way bigger issue, especially since knowing beforehand if it does or not isn’t something anyone actually tells you before testing.

You could have tried on an older android device first. But for 8 gigs, I wouldn’t even have bothered. I’d just throw them in a bucket and be done with it.

I don’t do backups of a lot. The important stuff that has like, an actual backup, has 2 copies off site. One at my work, one at my wife’s work. Just a regular LTO tape. The vast majority of what I store, is not something I consider important enough to backup as it’s stuff like, dvds and blurays I’ve ripped and still have and thus could get out of storage and rip again should that need arise. I do however donoff site replication. 3 replica ceph pool where one copy has to be off site. There’s again one node at my work and one at my wife’s work, plus ofc the local cluster which for drive constraints in the work nodes, always has two of the replicas. Might add a second node at my work but while we both get to place the servers there, I still pay for power, and wife for power and connection. With power prices these days, it gets kind of expensive.

Don’t need to. I use ceph with replication, so the drive can die after 5 mins and it wouldn’t matter at all to the array. I just put a sticker on it that tells me when it’s bought so I can RMA it if it dies quickly. Beyond that, I can basically lose two thirds of my drives and still be perfectly fine, or I could lose data from 3 drives failing if I’m unlucky and it’s from three different servers, one of which is off site. And that loss has to be within like a couple of hours between them or it’ll have time to create new replicas of it, and it will only be a small subset of the data that has priority once the second drive fails as it will only need to create new replicas of the data that only exist on those two drives which isn’t going to be more than a couple of hundred megs really.

Basically, I love Ceph if that doesn’t show ;P