One doesn't often hear the IT terms "mainframe" and "deduplication" used in the same context, or even the same sentence. But IBM is now connecting those dots.
IBM said Feb. 24 that, for the first time, it is adding deduplication to its System z mainframe server. The new "dedupe" appliance sports a rather verbose moniker: System Storage TS7680 ProtecTIER Deduplication Gateway for System z.
Brad Johns, IBM's Storage Program Director, told eWEEK that the new package is able to pare down piles of data by 80 to 90 percent. In testing, ProtecTier "deduped" a whopping 25 terabytes of tape application data down to a mere 1 terabyte, Johns said.
Data deduplication is a tool that eliminates redundant data throughout the storage network and makes the storage task more efficient, cost-effective and energy efficient within the network.
ProtecTier comes out of a $200 million acquisition IBM made of Diligent Technologies in April 2008. Diligent's patented deduplication software turned out to fit IBM's needs perfectly, because IBM has since customized it for several uses.
"After we acquired Diligent, we rolled out some initial implementations based on their software of what we call the Deduplication Gateway for open systems in August 2008," Johns said. "Then last year, we rolled out some pre-configured appliances using the ProtecTier technology for open systems. We added replication functions, so that was our 2009 report card."
Now IBM is taking some of that technology and integrating with some of its tape products. Many people don't realize it, but IBM's been in the digital tape business for more than 30 years. System z mainframes are designed to work directly with tape storage.
"So now we can integrate [Diligent] with z/OS [System z's operating system], and we're taking the next logical step by making deduplication available for System z customers," Johns said.
The new ProtecTIER Deduplication Gateway for System z combines a VTL [virtual tape library] with Diligent's inline data deduplication algorithm called HyperFactor, a patented technology that indexes the complete content of a repository. The repository then is permanently hosted in the System z RAM, thanks to its small footprint, Johns said.
The TS7680 also features two-node clustering and up to 1 petabyte of physical storage capacity per system, Johns added.
Diligent's in-line type of deduplication was specifically designed for high-end and high mid-range-size enterprise server and storage infrastructures, IBM Director of Mergers and Acquisitions Ari Kugler told eWEEK at the time of the acquisition.
"Diligent's software is very portable and fits perfectly into our Tivoli tool set," Kugler said. "If we hadn't bought the company, our own architects would have eventually come up with similar dedupe software at some point."
This was a strategic -- not simply a tactical -- move for IBM, Kugler said. "This will have a lot of future impact on all we do," Kugler said, rather prophetically.
IBM's ProtecTier Deduplication Gateway for System z is available now. Go here for more information.