NEC HYDRAstor is a scale-out grid storage platform with inline global deduplication, solving present and future challenges for long-term data. HYDRAstor’s grid architecture provides scalability of both performance and capacity, capacity optimization, advanced erasure-coded data protection and high availability with node-level resiliency. HYDRAstor can be further expanded and refreshed online with no data migration by incorporating newer generation hardware into the same grid system, maximizing investment protection.
The HYDRAstor architecture announcement in March 2007 marked the culmination of more than four years of intense research, development, and market validation. Nothing is more satisfying for an industrial research team than to see their innovation solve real customer problems. The HYDRAstor project started with a pair of researchers obsessed by their vision of a whole new way of providing storage before growing into a global product realization effort that drew on NEC expert resources around the world.NEC HydraSTOR is a massively scalable distributed grid storage platform, built to support long- term data retention by reducing storage capacity consumption by up to 95% or more using inline global deduplication. Making it unique in the industry is the online upgrade/expansion and data in-place technology refresh capability. Data is migrated to newer technology seamlessly with no migration planning or effort.
Scale-out grid architecture:
HYDRAstor is composed of independent nodes that work together to comprise a shared pool of storage that is self-organizing and self-healing. Simply by adding nodes, HYDRAstor can seamlessly scale to massive capacities and throughputs. Unrestricted Independent Linear Scalability of both Performance and Capacity supporting up to +100PB at +4PB/hour.
HYDRAstor deduplicates data across the entire grid and compresses data inline, yielding an effective storage capacity as much as 20x greater than raw capacity. Its inline deduplication engine cleverly utilizes a distributed hash table and harnesses the power of CPUs throughout the grid for unmatched input performance. High Performance Inline Global Deduplication – HYDRAstor DataRedux Deduplicates ALL data from ALL nodes, reducing storage capacity requirements by 95% or more.
The HYDRAstor grid supports mixing nodes of different generations. This enables an organization to retire older nodes in a piecemeal fashion, replacing them with newer nodes that have faster CPUs, more energy efficiency, and larger or faster disks. Because HYDRAstor is self-organizing, it will automatically incorporate the new nodes and redistribute data with no downtime or performance penalty. This gives HYDRAstor users a high level of investment protection as technology improves over time
Enterprises can replicate data between grids for disaster recovery and business continuity, as well as replicate data from HYDRAstor’s “mini-HYDRA” (the single-node HS3-410 model often deployed in remote offices and small businesses) to another mini-HYDRA or a larger centralized grid.
HYDRAstor includes a plethora of other features that help make it a robust storage platform for enterprise backup and long term data retention: data encryption, both in-flight and at rest; WORM (Write-Once, Read-Many) and data shredding for compliant retention; file system snapshots/clones and instant file copies.