Grow a cluster from one node to thousands using the powerful CRUSH algorithm.
Cope with outages automatically; grow & shrink capacity at will.
Make forklift upgrades and data migration obsolete.
Hot or phased software upgrades
Upgrade clusters in phases with no or minimal downtime.
APIs and Extensibiliy
S3 and Swift API support.
Full integration with OpenStack block storage (Cinder/Glance).
Native language bindings
Native bindings in a variety of languages provide the richest application experience.
Manage all cluster and object storage functions with a RESTful API.
At-rest encryption of storage disks.
Access control Lists
Exert granular control over object storage user and bucket-level permissions.
Set pool or object storage limits.
Reliability and Availability
Striping, erasure coding, or replication across nodes
Enjoy data durability, high availability, and high performance.
Use the CRUSH algorithm to allow every client to calculate where data is located without the need for lookup tables and speed performance.
Self-healing prevents failures from impacting integrity or availability.
Dynamic block resizing
Expand or shrink Ceph Block Devices with no or minimal downtime.
Improved parallelism for data I/O
Leverage a “client/cluster” model, not a “client/server” one.
In-memory client-side caching
For both kernel and hypervisor access to Ceph block device.
Promote hot data to SSDs with expiration policies.
Enhance the write performance of data.
Provision virtual machine (VM) images quickly (block only).
Customizable stripe sizes
Achieve optimal performance whether storing multi-gigabyte video files or small pictures.
Multi-Site & Disaster Recovery
Zones and region support
Deploy topologies similar to S3 and others with a global namespace (object only).
Data center synchronization
Back-up full or partial sets of data between regions (object only).
Serve local copies of data to local users (object only).
Export snapshots to geographically dispersed data centers
Institute disaster recovery (block only).
Export incremental snapshots
Minimize network bandwidth by only sending changes (block only).
Enjoy the value of a cost-effective data-durability option.
Allows over-provisioning (block only).
Tailor the price/performance mix to the workload.
Avoid having to replace older nodes as newer ones are added.
Development libraries for direct application integration
Allow such advanced functions as triggers, message passing, and in-situ data transformations (object only).
High-performance native protocol
Remove the overhead of the RESTful interface for performance-sensitive software that needs low latency and high throughput I/O (object only).
Get quick status on your cluster and take action with an integrated Ceph management platform for Inktank Ceph Enterprise.
Per-disk/pool performance stats
Measure IOPS over time and spot bottlenecks quickly and easily.
Survey cluster-wide component status with this speedy tool.
Monitor disk usage and adjust cluster, pool, and OSD settings.