Scalable distributed storage system for object, block, and file data
- Stars16.7k
- Forks6.4k
- Open Issues1.1k
Other
- C++
- Python
- TypeScript
About Ceph
Ceph is a distributed storage system that exposes object, block, and file storage from a single cluster. One pool of commodity servers serves all three access patterns, replacing separate systems for each.
Data is spread across nodes with replication or erasure coding for redundancy and high availability. Object storage is reachable through an S3-compatible API, block storage through iSCSI and NVMe over Fabrics, and file storage through POSIX, NFS, SMB, and FUSE. A CephFS layer and an HDFS interface are also available.
Ceph runs on Linux and integrates with Kubernetes. Most of the code is dual-licensed under LGPL 2.1 or 3.0, with some files under BSD-style or GPL licenses.
Key features
- Object, block, and file storage from one cluster
- Replication and erasure coding for redundancy
- S3-compatible object storage API
- Block storage via iSCSI and NVMe over Fabrics
- File access via POSIX, NFS, SMB, and FUSE
Details
- Platforms
- Linux
- Storage
- Object · block · file
- Access
- S3 · NFS · SMB · iSCSI · NVMe-oF
- Redundancy
- Replication · erasure coding
- Orchestration
- Kubernetes
- License
- LGPL 2.1 / 3.0