HPE StoreOnce
You can use HPE StoreOnce storage appliances as backup repositories. Depending on the storage configuration and type of the backup target, HPE StoreOnce can work in the following ways:
Source-Side Data Deduplication
- You have a Catalyst license installed on HPE StoreOnce.
- You use a Catalyst store as a backup repository.
- The Catalyst store is configured to work in the Low Bandwidth mode (Primary and Secondary Transfer Policy).
- You add the HPE StoreOnce Catalyst as a deduplicating storage appliance, not as a shared folder to the backup infrastructure.
HPE StoreOnce deduplicates data on the source side, before writing it to target:
- During the backup job session, HPE StoreOnce analyzes data incoming to the HPE StoreOnce appliance in chunks and computes a hash value for every data chunk. Hash values are stored in an index on disk.
- The HPE StoreOnce Catalyst agent calculates hash values for data chunks in a new data flow and sends these hash values to target.
- HPE StoreOnce identifies which data blocks are already saved on disk and communicates this information to the HPE StoreOnce Catalyst agent. The HPE StoreOnce Catalyst agent sends only unique data blocks to target.
Target-Side Data Deduplication
- The Catalyst store works in the High Bandwidth mode (Primary or Secondary Transfer Policy is set to High Bandwidth).
- The Catalyst license is installed on the HPE StoreOnce (required).
- The Catalyst store is added as a backup repository to the backup infrastructure.
- The Catalyst license is not required.
- The CIFS store is added as a backup repository to the backup infrastructure.
For more information about work with CIFS stores, see Shared Folder Mode.
HPE StoreOnce deduplicates data on the target side, after the data is transported to HPE StoreOnce:
- HPE StoreOnce analyzes data incoming to the HPE StoreOnce appliance in chunks and creates a hash value for every data chunk. Hash values are stored in an index on the target side.
- HPE StoreOnce analyzes VM data transported to target and replaces identical data chunks with references to data chunks that are already saved on disk.
As a result, only new data chunks are written to disk, which helps save on disk space.