Summary
Deduplication is a critical component to maximizing storage utilization efficiency and minimizing bandwidth requirements to offsite storage and processing. But, with shifts in solid state storage and RAID technologies, deduplication is quickly expanding beyond backup.
Analysis
Dell is expanding its services models. It has to. Otherwise, it will stay a manufacturer of commodotized products battling to keep its share of razor thin margins.
Dell's new efficient storage management services are just what the doctor ordered. Why? Because very few people know what the heck deduplication actually is! And why it is important!
Deduplication in its simplest form is removing redundant or duplicate data and indexing it in some way so that data retrievals return the right data.
People who have actually heard of deduplication assume it has something to do with making offsite backup less expensive. Foundationally, that is exactly correct. There are two primary flavors: dedup it before you send it across the wire and dedup it after it arrives at the other end.
But, deduplication can be applied in-line before the information is ever written to primary storage. The problem is that deduplication can be slow and hard disk drives are already slow enough. Plus, when you add in processor and storage virtualization, you start to create a major disaster if the servers or drives crash.
All of the problems are about to be eliminated. Prices are dropping drastically on solid state drives (SSDs). Utilization of solid state drives can cost more than traditional hard drives, but they eliminate the disk bottleneck and eliminate most of the wait that processors have historically had to do.
In addition, release of new technologies like PMC-Sierra's 6Gb/s RAID-on-a-chip further accelerates I/Os and embeds the survivability factors needed in a virtualized environment.
Dell hasn't announced services in this space yet because the combination of RAID-on-a-chip, solid state drives, in-line deduplication and virtualization is still evolving. But its time has come. The costs are coming down, the needs for compressing data centers are increasing (power, ROI, virtualization) and the technologies are all converging.
