Ceph Cluster
Ceph is a distributed storage system that unifies object, block, and file storage on one self-healing cluster of commodity hardware, using the CRUSH algorithm to place and replicate data without a central metadata bottleneck. It began as a doctoral research project before growing into an independent open-source project.
Baseline topologies
| Topology | Minimum nodes | Roles |
|---|---|---|
| converged-small | 3 | ceph-mon-mgr-osd |
| dedicated-roles | 6 | ceph-mon-mgr, ceph-osd |
What drives the cost
- Raw storage capacity (TB)
- Total raw storage capacity across all OSDs before replication or erasure coding overhead. Primary driver for OSD count and disk provisioning.
- Replication factor
- Number of data replicas (for replicated pools) or equivalent redundancy level. Multiplies usable storage into raw storage requirements.
- Target OSD count
- Target number of Object Storage Daemons. Each OSD typically maps to one physical disk. More OSDs improve throughput parallelism and recovery speed.
Where it runs
bare_metal, dedicated, cloud_vm
Size Ceph Cluster and get a cost estimate