Software Defined Storage.
Scale-out, Distributed, Parallel, Geo-redundant, Highly Available File System.
All data is distributed among multiple chunk servers.
Disk and server failures are handled transparently, without any downtime or loss of data. If storage requirements grow, it is straightforward to scale an existing LizardFS installation by simply adding new servers – at any time and without any downtime. The system will automatically move data to the newly added servers, as it continuously manages the balancing of disk usage across all connected nodes. Removing a server is just as easy as adding a new one.
Chunk servers are build on standard commodity platforms and layered on top of any POSIX filesystem like ZFS so all the performance tuning options can be used.
LizardFS is 100% Hardware agnostic. You can run it on any hardware you want as long as it can be managed by a Linux or Unix system.
Performance depends on how you build your chunk servers so if you want extra high performance, you just improve your chunk servers or use a small group of chunk servers as your high performance tier.
Official part of Debian, Ubuntu, Fedora.