
It's kind of like a distributed RAID system. Note that there's (currently) no way to access it directly from user space; you need to use an API.
The storage nodes are just HTTP servers. You can define the appropriate level of redundancy for different classes of files.
Sounds quite interesting.