Caringo has announced that it has been awarded a patent for erasure coding and replication in storage clusters. The patent covers the “combining of replication and erasure coding managed by the storage cluster in order to take advantage of the benefits of each without needing control computers or control databases.” Caringo’s patents are current deployed in its unified storage software, Caringo Swarm (formally called CAStor). Swarm is a mature object storage system for private and public cloud usage and a target archive system.
Caringo has announced that it has been awarded a patent for erasure coding and replication in storage clusters. The patent covers the “combining of replication and erasure coding managed by the storage cluster in order to take advantage of the benefits of each without needing control computers or control databases.” Caringo’s patents are current deployed in its unified storage software, Caringo Swarm (formally called CAStor). Swarm is a mature object storage system for private and public cloud usage and a target archive system.
The patent includes methods of:
- Specifying replication or erasure coding by “instructions from a client, an inherent property of the object, the metadata of the object, a setting of the cluster, or by other means.”
- Replacing content associated with a unique identifier
- Using the manifest or replicated objects in the case of a failed disk scenario
- Relocating an object within a storage cluster without the need for an extra control computer or control database
- Using metadata to specify the transformation from replication to erasure coding automatically based on trigger conditions
- Moving an object from one cluster to another and automatically converting to the storage format used by the second cluster, dictated by default cluster settings, by user metadata of the object, or by instructions from the program initiating the move.
The new patented approach will enable users to take advantage of the footprint efficiency erasure coding and low processing utilization and rapid access of replication without the introduction of complexity or single point of failure associated with control computers and control databases. The new approach gives customers:
- The flexibility to set data protection based on data value, down to object level, on the same cluster without special purpose nodes
- The ability to dynamically transfer replicated objects to erasure coding in a single cluster
- Move from replication to erasure coding, erasure coding to replication and even change the erasure coding formula, all in the same cluster, all automated by Lifepoints
- Set defaults by domain or down to the object level and override the defaults at any point
