ZFS is 20 years ahead of any other filesystem around. I could list at least 20 reasons why it's better but heres the most imports:
1
ZFS is a pooled filesystem. You have what is called a "zpool" made up of "vdevs" A vdev is a virtual device, it can be a single hard drive, a mirrored group of drives (2 way, 3 way, n-way mirrors), or a raidz (1,2,3) group of drives (raidz is similar to raid5 in the way that it is "single parity" but that's where the similarities end..., raidz2 is double parity (like raid 6))
2
Each pool can have multiple filesystems. Filesystems share the space of the pool. It works similar to the way that memory works....with memory, if you have 4 gb of ram, and an application needs some ram, it pulls it from the pool, when it doesn't need it anymore, it goes back to the pool. )
3
It is a copy-on-write transactional filesystem. Basically this means that when it writes a file, then later that file changes, it just keeps track of what has changed. This is one of the coolest things because it makes snapshots possible
4
The Snapshots in ZFS are near instant, and inintally take up no space. You can take snapshots frequently. A snapshot is a "moment-in-time" read only copy of a filesystem. Because of how ZFS works, like i said before, these snapshots initially take up no space, and only grow in size based on how different the filessytems are. This is one of the most important things about ZFS which owns.
5
Clones. A clone is like a snapshot, except you can write to it. Basically, when you clone a filesystem, you end up with a second filesystem which shared a point in history with the first, but can diverge. Because of this, it takes up no extra space, and only grows in size based on new transactions.
6
Compression. ZFS can do "block level" on the fly compression. You can set compression at the filesystem level making everything in a given filesystem compressed. This actually makes the filesystem faster in many situations because disk i/o is often the bottleneck, which CPU cyles are easy to come by.
7
Software raid. It is software raid but it is as fast, if not faster than hardware raid in many situations. It uses your memory and cpu as the "hardware" and can take full advantage of the disk i/o available.
8
End to end Checksumming. Perhaps the best feature of all is the end to end checksumming. This basically ensures data integrity. No other filesystem around right now has this feature. Basically, it works like this: Each block written to a ZFS filesystem is hashed, then the next block is hashed against that block, making the entire filesystem structure a self validating merkle tree. (each block checks against the last) Because of this, ZFS can detect, and FIX data corruption that other filesystems simply miss. This is probably the number one best feature in my opinion because, what most people don't understand is: Hard drives RARELY fail "hard". Most of them fail over time, where some blocks just "go bad" while others work file...thsi type of failure happens over time and other filesystems just "lose" the data. With ZFS, if you are using some sort of redundancy (raidz or mirrors) then not only can it find these errors, but it will fix them. Just the act of READING data checks it in this way, and if it fails the checksum, it will fix it. This "SELF-HEALING" is absolutely amazing and will come in more handy than you can image
9
Raidz eliminates the "raid5 write hole"
10
Transactions are atomic. Because of this, if you have a sudden power failure, you don't have to worry about your data being corrupted. There is no "fsck" with ZFS.
11
It is SERIOUSLY fast when set up correctly.
12
128 bit. This means it is future proof.....all other filesystems are 64 bit.
I could keep going....but you get the point.
