By Kevin Komiega
—IBM kicked off 2008 this week with news that it has acquired XIV, a privately held storage technology company based in Tel Aviv, Israel, which specializes in grid-based storage systems for fixed digital content.
IBM would not disclose the financial details of the acquisition, but various media reports put the deal in the $300 to $350 million range.
XIV has developed a storage architecture called Nextra. The Nextra technology is used to create a grid-based storage system using standard, off-the-shelf hardware components. XIV claims Nextra provides high-performance storage with no hot spots via the system's ability to independently scale capacity and controllers.
Nextra is built using a series of independent modules. The modules are implemented using off-the-shelf, Intel-based servers within a customized, Linux-based architecture, and are interconnected over redundant Gigabit Ethernet switches. The modules act jointly as a large data grid devoid of a common backplane.
A Nextra system comprises data modules and interface modules. The data modules are responsible for storing data and performing all advanced storage functionality, such as redundancy, snapshots, and caching. The interface modules act as unintelligent routers and are responsible for accepting host I/O commands via Fibre Channel or iSCSI and forwarding them to the appropriate data module. Throw in some Gigabit Ethernet switches as the interconnects between modules and add some UPS units and you're off and running.
Each logical volume in the Nextra system is divided into multiple 1MB stripes. The stripes are then spread over all the disks in the system and the architecture ensures all disks and modules are used equally, even after volumes are added, deleted, or resized, or hardware is added or removed.
Nextra's design embeds the read/write cache into the same hardware module as the disks. The cache is implemented as a distributed cache, so that all cache units can concurrently serve host I/O and perform cache-to-disk I/O, thereby avoiding bottlenecks.
XIV also claims the Nextra system can survive any single failure without affecting host I/O performance, and the company touts a 15-minute rebuild time for 500GB SATA hard drives.
Andy Monshaw, IBM System Storage general manager, says the rapid creation of digital content has accelerated the need for a new way of applying infrastructure solutions to digital information.
"The acquisition of XIV will strengthen the IBM infrastructure portfolio long-term and will put IBM in the best position to address emerging storage opportunities like Web 2.0 applications, digital archives, and digital media," says Monshaw.
According to Steve Duplessie, senior analyst at the Enterprise Strategy Group research and consulting firm, XIV's specialization in storing and managing fixed digital content may help IBM move beyond traditional transactional storage architectures.
"The digital content world is different from the traditional transactional world, but most companies treat them the same," says Duplessie. "I like the move a lot, assuming IBM did it recognizing that there is a different world out there now, and even they can't make customers keep using the same architectures to solve brand-new problems."
Duplessie says transactional systems are here to stay, but he believes storage vendors will start to solve real-world problems as they move into the wild world of Web 2.0 storage. "You can eat for a long time with nothing but a knife, but sooner or later you'll cut yourself. When someone finally gives you a fork, you realize the right tool helps a lot, and it doesn't mean that you no longer need the knife," he says. "Neither is perfect alone—both have complementary roles."
Duplessie adds that architectures such as Nextra are designed to manage themselves and take the heat off IT when it comes to dealing with unknown growth, performance, reliability, and management issues.
"I'm hopeful that as more of the big vendors make plays like this, users will recognize that data growth is causing all of their problems," says Duplessie. "By looking at the type of data itself and treating the real cause—fixed digital content—differently than they treat their historic data, they can stop pushing themselves to the brink."