Ceph

From Wikipedia, the free encyclopedia

Ceph is also the abbreviation for the orchid genus Cephalanthera.

Ceph is a free software distributed file system created by researchers at the Storage Systems Research Center^[1] in the Jack Baskin School of Engineering at the University of California, Santa Cruz. The main developer is Sage Weil (developer of the Webring concept and founder of Dreamhost), who implemented much of the Ceph code for his doctoral dissertation based both on earlier SSRC research^[1] and his own research. Ceph's main goals are to be POSIX-compatible, and completely distributed without single points of failure. As with most distributed file systems, it supports transparent replication and fault tolerance.^[2] As of November 2007, Ceph is considered to be in an "early alpha stage" and is not yet ready for production.^[3]

The client computers rely on FUSE to mount the file system, although a simple Linux kernel client is also in development. The server runs as a regular Unix daemon.

1 Design
2 References
3 See also
4 External links

[edit] Design

Ceph employs three distinct kinds of daemons:

Cluster monitors, which keep track of active and failed cluster nodes.
Metadata servers (MDS) which store the metadata of inodes and directories.
Object storage devices (OSDs) which actually stores the content of files. OSDs usually store data on raw block devices, using a custom storage format called EBOFS. Alternatively, a local disk file system can be used with some caveats.^[2]

All of these are fully distributed, and may run on the same set of servers. Clients directly interact with all of them.^[3]

Unlike most distributed file systems, Ceph does striping of individual files across multiple nodes to achieve higher throughput, similarly to how RAID0 stripes partitions across multiple hard drives. An planned extension to this feature is adaptive load balancing, whereby frequently accessed objects are replicated over more nodes.^[3]