RAID helps protect you from certain failures, and also allows you to add more storage inexpensively. Typically buying multiple small drives is cheaper than one larger drive, but you do have to buy at least one extra drive for failure.

You do need to plan for backups. RAID is not a backup, it doesn’t protect you from fires, computer failure, or accidentally deleting things. So a backup is necessary. Look into 3-2-1 backup solutions. At least 3 copies of your data, 2 different hardware types (like SSD and HDD), with at least 1 offsite copy.

People often keep their NAS at home, with a cloud storage provider as an offsite backup, and an external drive for local backup.

If you have the money, and a family or friend with space, you can buy multiple complete NAS solutions and back them up to each other.

You’ll have to find some kind of balance, ad it js a game of chance. You are always limited by number of slots in your server and current largest drive size. Then you are trying to balance price, speed, and durability.

For exampl, let’s say maximum amount of drives is 10, and maximum manufactured size is 50 TB. You probably don’t need 500 TB of storage that is in no way durable (if a drive dies, all the data on it is lost) and on a single server.

Death of drives is almost certain, two drives dying at the same time is quite low, so something like RAID 4, 5, 6, or 10 is a great start. Depends on how much storage you want, and then partition it accordingly. If you want 20 TB, you can do 4x 8TB in RAID 5, which yields 3x 8TB (=24) of effective storage.

Adding new drives is easy, and you are are always wasting just one drive. Then it depends if you want to sacrifice more space for more durability and switch to RAID 6 later on.

If you want even more storage, you can buy a micro server like ODROID H4+ and use it as network connected storage.

I’m confused that you’re talking of buying 20tb SSDs - you must be very rich. Spinny drives are more usually used in homelab archive RAIDs since they are more cost effective at large size and RAID offsets some of the slowness associated with them. I’m going to assume you meant HDDs not SSDs, but the advice applies to both if I’m wrong about that.

Yes, you will want to RAID them. That gives some protection against individual drive failure, and yes, absolutely that is a concern. Whilst the chance of drives failing these days is less than it was, they still do fail without warning, even when relatively new, and because of the bigger sizes, the consequences are greater.

The alternative to RAID is JBOD (Just a Bunch Of Drives) which means lots of individual drives being presented, each with their true size, in multiple shares. Most folk don’t want that.

What RAID level you choose depends on:

1. How many drives you fit. 4+ is good, and “more smaller” is better than “fewer larger” for safety, although the compromise is an extra 10watts or so of power per drive.
2. Current best practice; Don’t use RAIDs 0 or 5 on large arrays. (0 means exponential increase of data loss. 5 is strongly discouraged due to rebuild times of large disks) 6 is good if you have enough disks. 1+0 (mirrored and striped) is reasonable, and the choice I made for mine.
3. The hardware you’re using. Whether a linux PC or a bespoke NAS tool. Whilst the RAID levels are similar, the tools used vary a lot.

Notes:

• Also, be realistic about the space you need. Don’t over-size. Plan for 3-5 years growth, by then you’ll be wanting to change because of speed changes or drive failure.
• Some raid types slow down writing of data, some speed it up. Most are much faster at READing data.
• https://en.wikipedia.org/wiki/Standard_RAID_levels gives some explanation of the types.
• Google for “RAID CALCULATOR” for lots of free websites that allow you to see what space different sized drives give you with different RAID levels.
• Do not omit a strong backup strategy. RAID only protects against some types of hardware failure. A lightning strike, fire, rogue bios or software update, the host dying with an incompatible raid system. Buy disks for backups that aren’t in your RAID. (Good branded USB 3 disk and caddies are sensible). Automate backups if you can. Backup only what’s not easily replaceable.
• I wrote some thoughts on backups here.

Couple things to warn you about as I understand them (take with a grain of salt, this is what I remember from reading about this stuff quite some time ago in school).

RAID is great, but if you are performing mirroring (RAID 1) in your RAID levels, it can be dangerous to do it on multiple SSD’s of the same size. This is because SSD’s longevity is based on how many writes they have in their life time - it is a set amount. This is also why larger SSD’s will typically last longer than smaller SSD’s.

The problem therein is that if you are mirroring data between 2 SSD’s that are of the same size, when one hits that lifetime write limit, it is likely the other will do the same in very quick succession, or at the same time, meaning that mirroring was not the best idea because both disks can fail at the same time. This is not true for something like a RAID 5, and should only be something to be concerned about if, again, you are using a RAID 1, 51, or something like that etc.

If for some reason you do want to do mirroring, then look at using spinning disks instead as their mean time between failure is wider in these cases, meaning you are less likely to lose 2 disks at the same time when mirroring.

Spinning disks are also a good bet if you don’t really need the speed of the SSD’s. They are cheaper for more storage and these days you can typically find some surplus seller who can sell you a lot of them for cheaper than what you would pay new or used for new spinning disks or used SSD’s, but you will lose a lot of read/write speed. Again, which one to go with depends on what kind of data you store and how you access it.

Also make sure you understand the difference between software RAID and hardware RAID, and the benefits and downsides of using each. Hardware RAID has to be supported by your hardware, but it is much faster when it is writing any additional data to other disks (like when it clones bits onto a second drive in a RAID 1 mirror, or when it writes parity to the disks in a RAID 5). Software RAID you can set up on anything even if the hardware does not support it, so it allows more chances to play around with what you want to invest in for hardware (you can use an old PC tower as a RAID NAS even if it doesn’t support hardware RAID in it’s BIOS for example, which means you don’t have to pay a lot for an actual NAS).

As others have said as well, you also want to buy more disks than you actually need (I like to have at least 2 unused disks on hand) for cases where you get a disk failure in your RAID array. Furthermore, like others have said, this does not protect against floods, fires, a meteor falling on your house etc - and so it is good to look into a way to make off-site backups on a rotating schedule.

For this, look up what a 3-2-1 backup is and how that works if you really want to make sure your data is safe.

Drives fails so always be ready for that.

How many drives you buy depends on your goals. I wanted a small nas so I’m using raid1 and just have 2 drives.

If your nas has more drive bays and supports it you can start with 2 drives and simple add a drive as needed. You wouldn’t be using raid at that point as man nas have more flexible options. My Synology calls it SHR.

This assumes you need simple file storage and not running multiple VMs with databases.