IBM FlashSystem Storage Deduplication


Two new IBM FlashSystems with space reduction features are now available – the A9000 and A9000R include pattern removal, deduplication, and real-time compression. They are architected as a modular, hyper-scalable grid-based system.

The A9000 utilizes FlashCore™ technology and, along with MicroLatency® modules, provide high density, low latency, and high I/O performance.

These all-flash storage systems are ideal for enterprise cloud environments that require fast, reliable, and redundant data storage.

Inline Deduplication

The inline deduplication eliminates and reduces duplicate data which helps you manage exponential growth. In the flash-optimized data reduction process, it is the second data reduction stage that processes and consolidates data before it is written. The benefit of this kind of deduplication is that duplicate blocks are never written to the destination.

In the illustration below you can see how the original data is reduced via the deduplication process.


In this process:

  • Hashes are calculated only once (upon write)
  • Hashes are stored in the metadata
  • Data is compared by comparing the hash
  • The block size is 8KB, although 4KB alignment is also supported and allows a higher deduplication ratio

Data reduction with deduplication alone could reach to up to 30:1 (96.7% savings).

The following illustration shows how data reduction and protection mechanisms built into the FlashSystem A9000 and A9000R work: