CephFS is a POSIX-compatible filesystem that sits on top of RADOS pools. Being POSIX-compliment means that it should be able to function as a drop-in replacement for any other Linux filesystem and still function as expected. There is both a kernel and userspace client to mount the filesystem on to a running Linux system. The kernel client, although normally faster, tends to lag behind the userspace client in terms of supported features and will often require you to be running the latest kernel to take advantage of certain features and bug fixes. A CephFS filesystem can also be exported via NFS or Samba to non-Linux-based clients, both software have direct support for talking to CephFS. This subject will be covered in more detail in the next chapter.

CephFS stores each file as one or more RADOS objects. If an object is larger than 4 MB, it will be striped across multiple objects. This striping behavior can be controlled by the use of XATTRs, which can be associated with both files and directories, and can control the object size, stripe width, and stripe count. The default striping policy effectively concatenates multiple 4 MB objects together, but by modifying the stripe count and width, a RAID 0 style striping can be achieved.