Reliable, Scalable, And High ­Performance Distributed Storage: Distributed Metadata Management

Ceph’s MDS diverges from conventional metadata storage techniques, and in doing so facilitates adaptive file system and workload partitioning among servers, improved metadata availability, and failure recovery. Specifically, file system metadata updates are initially written to large, lazily-trimmed per-MDS journals that absorb temporary and repetitive updates. File (inode) metadata is then embedded in the file system namespace and stored inside per-directory objects for efficient read access and metadata prefetching. Notably, MDS defines the namespace hierarchy in terms of directory fragments, facilitating fine-grained load distribution even for large or busy directories, and implements a traffic control mechanism for dispersing load generated by flash crowds—sudden concurrent access by thousands of client nodes—across multiple nodes in the MDS cluster.

Publication Date
19 February 2014

metadata server metadata Ceph


Click here to download link on

Click to share this page via your favorite social network.

Learn more about defensive publications with our examples and frequently asked questions

What we are trying to do?

We are attempting to mobilize the creativity and innovative capacities of the Linux and broader open source community to codify the universe of preexisting inventions in defensive publications that upon publication in the IP.COM database will immediately serve as effective prior art that prevents anyone from having a patent issued that claims inventions that have already been document in a defensive publication. In addition to creating a vehicle to utilize this highly effective form of IP rights management for known inventions, it is hoped that the community will use defensive publications as a means of codifying future inventions should the inventors prefer not to make their invention the subject of a patent disclosure and application.