ZFS is a next generation file system: it improves upon the reliability, flexibility, performance and scalability of traditional file systems. ZFS better protects data by hashing stored data and making updates to the underlying drives atomic. ZFS is flexible: it is a copy-on-write file system, which enables inexpensive snapshotting of data sets. This, in turn, enables quickly rolling back a data set to a given snapshot or even forking a data set. ZFS also supports hierarchical storage management: it can use an SSD to transparently cache reads. Finally, ZFS is scalable: some existing installations have over a PB
of storage.

In this talk, I will present ZFS's features and provide practical advice on how to deploy ZFS.