What is Ceph?

What it is

Ceph is free software developed by a broad community. Instead of buying one large storage appliance, a business builds a cluster from several standard servers and adds more later when it needs room.

By default, Ceph keeps three copies of each piece of data on different machines. When a drive or a whole server fails, the cluster keeps answering requests and rebuilds the missing copies in the background. Ceph offers three kinds of storage from the same pool: block (virtual disks), file (a shared file system called CephFS) and object (an S3-compatible interface).

What businesses use it for

• Virtual machine disks: block storage lets several virtualization hosts share the same disks, so a virtual machine can restart elsewhere if its host fails.
• Shared files: CephFS gives many computers one common file system without a single file server holding it all.
• Object storage: applications and backup software can write to an S3-compatible endpoint, which is the same style of storage that large public clouds offer.
• Growth in small steps: adding servers and drives increases capacity, and old hardware can be retired without moving everything at once.

How KawaConnect uses Ceph

We build Ceph clusters for shared block, file and object storage, and our own infrastructure runs on Proxmox and Ceph. In the Proxmox clusters we build, Ceph can hold the virtual machine disks.

Our encrypted S3-compatible object storage, which holds offsite backups, also runs on Ceph. You can read how that fits into a recovery plan on our backup and disaster recovery page, or see the Ceph storage service if you want a cluster of your own.

Learn more

• Ceph on WikipediaA neutral overview of how Ceph works, its history, and the parts that make up a cluster.
• Ceph project websiteThe home of the open-source project, with release announcements, the project blog, and community links.
• Ceph documentationThe official manual for installing, operating, and troubleshooting a cluster. This is what engineers work from.