File system tiers
A file system typically consists of files and directories. Data about the files and directories (as well as many other attributes) is kept; this information is the metadata. The data within the file system (both user data and metadata) is stored on the storage media of a storage subsystem.
Different storage subsystems have varying characteristics, in terms of both performance and cost. Based on their performance, storage subsystems are classified into "tiers," which are then used by administrators to manage the storage resources in the organization.
Storage pools can be created using storage from different tiers (up to two tiers are currently supported). These storage pools are called tiered storage pools. File system data (metadata and user data) may be stored in a single tier, or in multiple tiers. When file system data is stored on storage subsystems on multiple tiers, the file system is called a tiered file system.
For more information on storage tiers and tiered storage pools, refer to the Storage Subsystem Administration Guide.
In a tiered file system, metadata is stored on the highest performance tier of storage, and user data is stored on a lower-performance tier. A tiered file system may provide the following:
- Performance benefits: Storing metadata on the higher-performance tier provides system performance benefits over storing both the metadata and user data on the same, lower, tier of storage. The performance gain is seen because metadata is accessed more often than the user data, so storing it on higher-performance storage increases overall performance.
- Reduced expenses for storage: Storing metadata on the higher-performance storage (which is usually more expensive than the lower performance storage) and user data on lower performance (and less expensive) storage may provide cost benefits. This is because metadata typically consumes a relatively small amount of storage, while the user data consumes the bulk of the storage. Because the higher-performance storage is used only to hold the metadata, less of the expensive storage is used than if both the metadata and the user data were on the higher-performance storage. Also, because user data can be kept on lower performance storage while achieving better performance than keeping both metadata and user data on the lower performance storage, you may not have to upgrade storage as often (or you may be able to repurpose aging storage when you do upgrade).
A tiered file system has the following characteristics:
- Maintains a single file system view while providing data separation. This separation allows the file system to store file system metadata (which is critical to system performance) on very high-performance devices, while storing user data on cheaper, lower-performance storage subsystems.
- The use of multiple tiers of storage is completely transparent to applications or clients. No environmental tweaking or effort is required. All file system functionality (such as snapshots, replication, quotas, cluster name space, and virtual volumes are preserved.
- File system management activities, such as mounting, unmounting, sharing, and exporting, for tiered file systems are the same as for untiered file systems.
- The file system block size (4 KiB or 32 KiB) is maintained across all tiers of storage.
- Cross volume links are treated as metadata.