How Hammerspace Uses Metadata
Hammerspace metadata is actionable in that you can use it to orchestrate data. These orchestration use cases can be broken down into three broadly defined categories: presentation, placement, and protection. In this section, we will review examples of each of these categories.
Presentation
You can do ad-hoc queries, similar to the Linux find command, but with many additional options than typical POSIX metadata. The File Metadata Example (Full) section of this document contains an export of the full metadata values for a single value. Virtually every value seen there can be used when creating queries.
| Hammerspace shares also contain a default set of collections, which are subsets of files determined by a predefined query of metadata updated on each pass of the sweeper process, a daemon that scans the files in all Hammerspace shares to determine if they are aligned (compliant) to the objectives. Collections are accessible in any directory in a Hammerspace share in .collections, for both Linux (NFS) and Windows (SMB). |
Placement
Hammerspace metadata is used by objectives to determine tiering, replication for distributed workforce, cloud applications such as analytics and rendering, as well as disaster recovery.
Protection
Objectives use metadata to apply rules for mirroring, applying and complying with 9’s of availability and durability, backup (including snapshots), and WORM.