In computing , Ceph (pronounced / s ɛ f / gold / k ɛ f / ), a free-software storage platform , implements object storage distributed was single computer cluster , and Provides interfaces for object-, block- and file – Level storage. Ceph distributed completely Call AIMS Primarily for operation without a single point of failure , scalable to the exabyte level, and freely available.

Ceph replicates data and Makes It fault-tolerant , [7] using commodity hardware and Requiring no specific hardware support. As a result of its design, the system is both self-healing and self-managing , aiming to minimize administration time and other costs.

On April 21, 2016, the Ceph development team released “Jewel”, the first Ceph release in which CephFS is considered stable. The CephFS repair and disaster recovery tools are feature-complete ( snapshots , multiple active metadata servers and some other functionality is disabled by default). [8]


A high-level overview of the Ceph’s internal organization [9] : 4

Ceph employ four distinct kinds of daemons : [9]

  • Cluster monitors ( ceph-mon ) that keep track of active and failed cluster nodes
  • Metadata servers ( ceph-mds ) that store the metadata of inodes and directories
  • Object storage devices ( ceph-osd ) that actually store the content of files in a XFS file system. [10]
  • Representative state transfer (RESTful) gateways ( ceph-rgw ) that exposes the object storage layer as an interface with Amazon S3 or OpenStack Swift APIs

All of these are fully distributed, and may be performed on the same set of servers. Clients directly interact with all of them. [11]

Ceph does striping of individual files across multiple nodes to achieve achievement Higher throughput, similarly to how RAID0 stripes across multiple partitions hard drives . Adaptive load balancing is supported whereby frequently accessed objects are replicated over nodes. Citation needed ] As of December 2014 , XFS is the recommended underlying filesystem type for production environments, while Btrfs is recommended for non-production environments. Ext4 filesystems are not recommended because of limitations on the maximum RADOS objects length. [12]

Object storage

An architecture diagram showing the relations between components of the Ceph storage platform

Ceph implements distributed object storage . Ceph’s software libraries customer Provide applications live with access to the reliable autonomic distributed store object (RADOS) object-based storage system, and aussi Provide a foundation for Reviews some of Ceph’s features, Including RADOS Block Device (RBD) RADOS Gateway , and the Ceph File System .

The librados software libraries provide access in C , C ++ , Java , PHP , and Python . The RADOS Gateway aussi exposed the object store as a RESTful interface qui can have this native Both Amazon S3 and OpenStack Swift API.

Block storage

Ceph as a thin-provisioned block device . When an application writes data to Ceph using a block device, the data across the cluster. Ceph’s RADOS Block Device (RBD) also integrates with Kernel-based Virtual Machines (KVMs).

Ceph RBD interfaces with the same Ceph object storage system that provides the librados interface and the CephFS file system, and it stores block device images as objects. Since RBD is built on librados, RBD inherits librados’ abilities, including read-only snapshots and revert to snapshot. By striping images across the cluster, Ceph improves read access performance for large block device images.

The Apache CloudStack , OpenStack , OpenNebula , Ganeti , and Proxmox Virtual Environment .

File system

Ceph’s file system (CephFS) runs on top of the same object storage system and provides storage and block device interfaces. The Ceph metadata server cluster provides a service that maps the directories and file names of the file system to objects stored within RADOS clusters. The metadata server cluster can expand or contract, and it can rebalance the file system dynamically to distribute data evenly among cluster hosts. This ensures high performance and prevents heavy loads on specific hosts within the cluster.

Customers mount the POSIX -compatible file system using a Linux kernel client. On March 19, 2010, Linus Torvalds merged the Ceph client into Linux kernel version 2.6.34 [13] which was released on May 16, 2010. An older FUSE -based client is also available. The servers run as regular Unix daemons .


Ceph Was INITIALLY created by Sage Weil (developer of the Webring concept and co-founder of DreamHost ) For His doctoral dissertation , [14] qui Was Advised by Professor Scott A. Brandt in the Jack Baskin School of Engineering at the University of California, Santa Cruz and sponsored by the United States Department of Energy (DOE) and Oak Ridge National Laboratory (ORNL), Lawrence Livermore National Laboratory (LLNL), Los Alamos National Laboratory (LANL), Intel Corporation, Microsoft Corporation, SAP Laboratories and others. [15]

After his graduation in Fall 2007, Weil continued to work on a full-time Ceph and expanded to include Yehuda Sadeh Weinraub and Gregory Farnum. In 2012, Weil created Inktank Storage for professional services and support for Ceph. [16] [17]

In April 2014, Red Hat purchased Inktank, bringing the majority of Ceph development in-house. [18]

In October 2015, the Ceph Community Advisory Board was formed to assist the community in driving the open source software-defined storage technology. The charter advisory board includes the Ceph project, including individuals from Canonical , CERN , Cisco , Fujitsu , Intel , Red Hat , SanDisk , and SUSE . [19]

  • Argonaut – on July 3, 2012, the Ceph development team released Argonaut, the first major “stable” release of Ceph. This release will provide stability and performance updates only, and new features will be scheduled for future releases. [20]
  • Bobtail (v0.56) – on January 1, 2013, the Ceph development team released Bobtail, the second major stable release of Ceph. Argonaut stable series (v0.48.x) Argonaut stable series (v0.48.x). [21]
  • Cuttlefish (v0.61) – on May 7, 2013, the Ceph development team released Cuttlefish, the third major stable release of Ceph. This is the first step in the implementation of the ‘ceph-deploy’ deployment tool in the previous ‘mkcephfs’ method of deployment. [22]
  • Dumpling (v0.67) – on August 14, 2013, the Ceph development team released Dumpling, the fourth major stable release of Ceph. This release included a first pass at global namespace and media area, a REST API for monitoring and management functions, Improved Support for Red Hat Enterprise Linux derivatives (RHEL) -based platforms. [23]
  • Emperor (v0.72) – on November 9, 2013, the Ceph development team released Emperor, the fifth major stable release of Ceph. This release brings several new features, including multi-datacenter replication for the radosgw, improved usability, and a lot of incremental performance and internal refactoring work upfront features in Firefly. [24]
  • Firefly (v0.80) – on May 7, 2014, the Ceph development team released Firefly, the sixth major stable release of Ceph. This release brings several new features, including erasure coding, cache tiering, primary affinity, key / value OSD backend (experimental), standalone radosgw (experimental). [25]
  • Giant (v0.87) – on October 29, 2014, the Ceph development team released Giant, the seventh major stable release of Ceph. [26]
  • Hammer (v0.94) – on April 7, 2015, the Ceph development team released Hammer, the eighth major stable release of Ceph. It is expected to form the basis of the next long-term stable series. It is intended to supersede v0.80.x Firefly. [27]
  • Infernalis (v9.2.0) – on November 6, 2015, the Ceph development team released Infernalis, the ninth major stable release of Ceph. It will be the foundation for the next stable series. Hammer, and the upgrade process is non-trivial. [28]
  • Jewel (v10.2.0) – on April 21, 2016, the Ceph development team released Jeff, the first Ceph release in which CephFS is considered stable. The CephFS repair and disaster recovery tools are feature-complete (bidirectional failover, active / active configurations), some functionalities are disabled by default. This release includes new experimental RADOS backend named BlueStore which is planned to be the backend storage in the upcoming releases. [8]
  • Kraken (v11.2.0) – on January 20, 2017, the Ceph development team released Kraken. The new BlueStore storage format, introduced in Jewel, has now a stable on-disk format and is part of the test suite. Despite still marked as experimental, BlueStore is near-production ready, and should be marked as such in the next release, Luminous. [29]


The name “Ceph” is a common nickname given to pet octopuses and derived from cephalopods , a class of molluscs , and Ultimately from Ancient Greek κεφαλή ( ke-PHA lee ), meaning “head” and πόδι ( po -dhi ), meaning “leg”. The name (emphasized by the logo) suggests the highly parallel behavior of an octopus and was chosen to connect the file system with UCSC’s mascot, a banana slug called “Sammy”. [30] Banana slugs are gastropods , which are also a class of molluscs.

See also

  • Free software portal
  • Linux portal
  • Distributed file system
  • Distributed parallel fault-tolerant file systems
  • BeeGFS
  • Gfarm file system
  • GlusterFS
  • IBM General Parallel File System (GPFS)
  • Chandelier
  • LizardFS
  • Moose File System
  • OrangeFS
  • Parallel Virtual File System
  • Quantcast File System
  • RozoFS
  • XtreemFS
  • ZFS


  1. Jump up^ “Ceph Community Forms Advisory Board” . 2015-10-28 . Retrieved 2016-01-20 .
  2. Jump up^ “v11.1.1 Kraken released” .
  3. Jump up^ . Missing or empty( help ) |title=
  4. Jump up^ “GitHub Repository” .
  5. Jump up^ “FreeBSD Quarterly Status Report” .
  6. Jump up^ “LGPL2.1 license file in the Ceph sources” . 2014-10-24 . Retrieved 2014-10-24 .
  7. Jump up^ Jeremy Andrews (2007-11-15). “Ceph Distributed Network File System” . KernelTrap .
  8. ^ Jump up to:b Sage Weil (2016-04-21). “V10.2.0 Infernalis Released” . Ceph Blog.
  9. ^ Jump up to:b Mr. Tim Jones (2010-06-04). “Ceph: A Linux petabyte-scale distributed file system” (PDF) . IBM . Retrieved 2014-12-03 .
  10. Jump up^ “Hard Disk and File System Recommendations” . Retrieved 2017-03-17 .
  11. Jump up^ Jake Edge (2007-11-14). “The Ceph filesystem” . .
  12. Jump up^ “Hard Disk and File System Recommendations” . . Retrieved 2017-06-26 .
  13. Jump up^ Sage Weil (2010-02-19). “Client merged for 2.6.34” .
  14. Jump up^ Sage Weil (2007-12-01). “Ceph: Reliable, Scalable, and High-Performance Distributed Storage” (PDF) . University of California, Santa Cruz .
  15. Jump up^
  16. Jump up^ Bryan Bogensberger (2012-05-03). “And It All Comes Together” . Inktank Blog.
  17. Jump up^ Joseph F. Kovar (July 10, 2012). “The 10 Coolest Storage Startups Of 2012 (So Far)” . CRN . Retrieved July 19, 2013 .
  18. Jump up^ Red Hat Inc (2014-04-30). “Red Hat to Acquire Inktank, Provider of Ceph” . Red Hat . Retrieved 2014-08-19 .
  19. Jump up^ “Ceph Community Forms Advisory Board” . 2015-10-28 . Retrieved 2016-01-20 .
  20. Jump up^ Sage Weil (2012-07-03). “V0.48” Argonaut “Released” . Ceph Blog.
  21. Jump up^ Sage Weil (2013-01-01). “V0.56 Released” . Ceph Blog.
  22. Jump up^ Sage Weil (2013-05-17). “V0.61” Cuttlefish “Released” . Ceph Blog.
  23. Jump up^ Sage Weil (2013-08-14). “V0.67 Dumpling Released” . Ceph Blog.
  24. Jump up^ Sage Weil (2013-11-09). “V0.72 Emperor Released” . Ceph Blog.
  25. Jump up^ Sage Weil (2014-05-07). “V0.80 Firefly Released” . Ceph Blog.
  26. Jump up^ Sage Weil (2014-10-29). “V0.87 Giant Released” . Ceph Blog.
  27. Jump up^ Sage Weil (2015-04-07). “V0.94 Hammer Released” . Ceph Blog.
  28. Jump up^ Sage Weil (2015-11-06). “V9.2.0 Infernalis Released” . Ceph Blog.
  29. Jump up^ Abhishek L (2017-01-20). “V11.2.0 Kraken Released” . Ceph Blog.
  30. Jump up^ “How the Banana Slug became UCSC’s official mascot” . Retrieved September 22, 2009 .

Leave a Reply

Your email address will not be published. Required fields are marked *