Storage tiering is a proven storage management method that moves data between different storage classes and storage tiers. The goal is to make optimal use of the storage systems.
Most companies have a large stock of so-called cold data:
- This cold data is rarely (or never) accessed anymore.
- Nevertheless, this data must be retained, for example due to legal requirements.
Tiering of this data is performed automatically based on predefined rules. For example, data that has not been accessed for a certain period of time is moved to a lower-performance, but less expensive storage system.
Storage tiering provides cost and time savings as well as efficiency gains:
- Primary storage systems are relieved of inactive data.
- Backup times are shortened and backup volumes reduced.
File Tiering and Block Tiering
Tiering can be performed according to different methods – analogous to the different storage methods. A distinction is made between file-based and block-based tiering.
- File tiering falls within the domain of data management. This method is based on file attributes (e.g. size or age of the file). It is flexibly controlled by software, independent of the storage systems used.
One such software that performs vendor-independent, file-based tiering is, for example, PoINT Storage Manager. PoINT Storage Manager moves files within a multi-tier storage architecture to the storage level corresponding to their age or usage. The rules for this tiering are defined by the user beforehand.
- Block tiering, on the other hand, is a vendor-specific storage management feature. In the context of block tiering, individual blocks of files are moved between the storage levels. The basic factor here is the frequency of access to the respective block.
Tiering and Workflows
The choice of the tiering method has a decisive influence on user workflows. This is mainly because file tiering preserves the file as a whole, while block tiering moves individual blocks of a file.
| File Tiering (by PoINT Storage Manager) | Block Tiering (by storage system) | |
| File Integrity | The files remain as such. They are stored on the primary or secondary storage according to the tiering rules and remain identifiable by attributes. | Individual fragments of a file (blocks) are stored on different storage levels. A moved block cannot be identified. |
| Tiering Rules | The tiering rules can be set individually. Tiering can also be triggered manually. | Tiering rules are predefined and cannot be customized (no manual tiering, no custom rules). |
| File Access on Secondary Storage | Once moved to the secondary storage, files can be accessed seamlessly and independently from the primary storage. | File access is only possible through primary storage tier. |
| Read Access | Restoring to primary storage is not necessary for read-only access to moved files. This maintains performance and saves storage space. | Moved blocks must be restored to the primary storage. |
Investment Protection and Future-proof
Investments in storage strategies should always be viewed in the long term.
- Files must be stored securely over long periods of time in accordance with legal requirements. In this blog article, you will learn what role the storage strategy plays in this.
- During this retention period, an archive migration may become necessary – if only because of the limited lifetime of the storage media used.
So the question here is also which tiering method offers reliable investment protection and how the data can be backed up securely and in compliance with the legal regulations in the long term, regardless of the hardware.
| File Tiering (by PoINT Storage Manager) | Block Tiering (by storage system) | |
| Vendor Independence | The tiering software moves files between storage tiers – regardless of the hardware used. | Block tiering is a functionality of the storage system itself, and thus vendor-specific. |
| Backup | The backup application identifies and skips moved files. It stores only the links or stubs pointing to the archived files. This brings significant savings in terms of storage space and backup time. | The backup application cannot identify or skip moved blocks. |
| Storage Migration | Primary and secondary storage can be migrated independently without restoring moved files. | During the migration process, a complete restore of moved blocks on the primary storage system is necessary. Migration of the secondary storage only is not supported. |
| Legally Compliant Archiving | In the context of file tiering, methods such as WORM, retention management and versioning are supported. | Typical functions that enable legally compliant archiving (WORM, retention management, versioning) are not supported with block tiering. |
Conclusion
Block tiering is a method performed by the operating system of the storage system. Transparency about the location of a file is lost, because a file is fragmented and only the NAS operating system knows where which blocks of a file are stored. In addition, migrations become very difficult because blocks of a file that have first been moved out must be moved back to the primary storage. Due to space constraints this is no longer even possible in many cases.
File tiering, as performed e.g. by PoINT Storage Manager, offers independence from the storage manufacturer and enormous saving potential. Based on automatically or manually executable rules, e.g. files which have not been accessed for a long time are moved as a unit to more suitable storage classes. In addition, compliance and archiving requirements can be met through file tiering.
Block tiering is a storage system-specific method and leads to vendor lock-in. Software-based file tiering, on the other hand, represents independent data management.