Deduplication to Tape
Deduplication to tape also known as silo storage minimizes primary disk storage usage by managing the disk space and periodically moving the data to the tape. This efficient disk management reduces the cost of storage. Silo storage can be used for long-term data preservation.
A silo is a set of disk volume folders associated with a Deduplication Database (DDB). The volume folders contain deduplicated data written on the disk storage.
When configuring a silo copy, we recommend you to follow the practices described in Deduplication Best Practices.
Deduplication-to-Tape Backup Process
Deduplication-to-tape works as follows:
- When a silo copy is created, an on-demand backup set is created in the File System Agent on the CommServe client. The default subclient in the on-demand backup set is used to back up volume folders from disk to tape.
- The volume folders are marked full, based on the following Media Management Configuration parameters, and a new volume folder is created.
- Interval (hours) between marking large disk volumes full parameter on the Service Configuration tab.
- Hours after being marked as Full for archiving Silo volumes to tape parameter on the Silo Archive Configuration tab.
For more information on these parameters, see Media Management Configuration: Silo Archive.
Note: When a DDB is sealed, a new set of volume folders is created for the next DDB. The silo associated with the sealed DDB can remain on disk or be moved to a silo copy.
- When a silo backup job is run (immediately or as per schedule), the volume folders that are marked full are copied from disk to tape.
- After volume folders are copied to tape, every 24 hours (1440 minutes), volume folders eligible for removal from disk storage (referred to as cache in silo) are identified and deleted to free up disk space.
To set up deduplication-to-tape backup, use the following process:
- Create a Silo Copy
You must create a silo copy on the deduplication-enabled storage policy copy.
- Back up the Deduplicated Data to Tape
Back up the deduplicated data to tape by scheduling a Silo Archive job.
Auxiliary Copy Process for Deduplication-to-Tape
You can perform an auxiliary copy of the silo storage. This Auxiliary Copy preserves the data deduplication in the silo storage and can be used to make additional copies of the data. This method retains the deduplicated storage environment by creating a secondary silo copy. For more information on setting up tape-to-tape Auxiliary Copy, see Create a Secondary Silo Copy.
Deduplication to Tape Recovery Process for Deduplication-to-Tape
When a restore, auxiliary copy, or data verification job is run for data in silo storage, the following process is used:
- If the data to restore is available on the disk (cache), it is restored from the disk.
- If the data to restore is available on the tape and not on the disk, then the following is performed:
- The restore job goes into Waiting state.
- A Silo Retrieval job is automatically started to retrieve the volume folders from tape to disk.
- A Silo Retrieval job copies only those volume folders that are required for the restore from tape to disk.
- When the data is available on the disk, the initial restore job resumes by using the volume folders in the disk.
- By default, volume folders retrieved from tape are written to the original source mount path. If the original source mount path is not available or has insufficient storage, any other mount path in the library with sufficient space will be used.
To ensure that a mount path with sufficient space is available in the library at all times, you can have a dedicated mount path that is reserved for the recovery process of deduplication-to-tape. For more information, see Reserving a Mount Path for Silo Restores.