Grid (Scale Out) Index

Albireo grid technology enables OEMs to increase performance and scalability far beyond what is possible with traditional data optimization solutions. Albireo combines patented indexing and memory resource utilization software with a scale-out design to provide performance that expands linearly, even when dealing with multi-petabyte capacities, without the performance drop-off common with the other data deduplication solutions on the market today. An Albireo grid consists of multiple Albireo clients, embedded in each storage ingest point, connected to multiple Albireo indexes in a grid configuration over the network (Figure 1).

In a grid deployment scenario, each Albireo client performs its own content-level segmentation and hashing, then routes the resulting content fingerprints to the appropriate Albireo index. The index returns deduplication advice to the client and stores the deduplication metadata if the new data is unique. The existing storage software is responsible for data storage, while the index servers store the content fingerprints in the Albireo index.

The global index table is automatically distributed across all index servers in the Albireo grid. Each index server stores a balanced portion of the index. When servers are added to or removed from the grid, the index is re-balanced. A single Albireo grid supports multiple namespaces; separate independent storage domains (e.g., file systems, logical volumes, or systems) can share a single global Albireo index. Both unified storage pools and large quantities of independent storage volumes benefit from the performance and scalability of Albireo grid technology.

Albireo grid technology provides:

• Massive indexes for unlimited scalability
• Distributed workloads for maximum performance
• Deduplication results in just microseconds