With an IBM contract under its belt, McData is pushing into the storage area network routing market.
Customers will be able to tier, interconnect and segment their storage networks with its new products, much as they segment their local and wide area networks today, McData claimed.
It has also filled in details of its upcoming product plans. They include the upcoming Intrepid 10000 enterprise-class director, 4gigabits per second (gbps) upgrades for the existing Intrepid 6000, a new 4Gbit range of Sphereon Fibre Channel switches and more versions of the Nishan-derived Eclipse Fibre-to-IP storage router family.
The i10000 has an eight-blade architecture with each blade carrying four paddles, said Jean Becker, McData's senior vice-president of engineering. Each paddle can carry eight 2Gbit Fibre Channel or FICON ports, or one 10Gbit port for inter-switch linking. It can support 128 full-speed 4Gbit ports, or 256 over-subscribed ports.
Next year the company will add both 4Gbit Fibre Channel and Eclipse-like routing blades to the i10000.
Becker said the Intrepid 6000 is not being retired yet, and it too will gain 4Gbit and 10Gbit blades. "Routers make it much easier to do heterogenous interoperation, that's one reason they're becoming popular," she said. "We will have routing, distance and iSCSI in one box."
The Eclipse family will be augmented next year by a 1640 model, offering two Fibre, two Gigabit Ethernet and two multi-function ports. This will include hardware encryption and wire speed iSCSI, as well as routing capabilities.
Also set for mid-2005 delivery are the Sphereon 4400 and 4700 range of 4Gbit Fibre Channel and FICON switches. These will include the 4710, with 24 4Gbit ports and two 10Gbit ports, each with extra buffer credits for long distance connectivity. They will also feature McData's FlexPort scheme for upgrading switches with blocks of extra ports, plus the ability to upgrade firmware while the switch remains online.
"The storage networking market is growing up and the landscape is changing," said Jeff Vogel, McData's senior worldwide marketing vice-president. "The level of complexity in Sans is orders of magnitude greater than it has been."
According to Vogel, the key is a network infrastructure that is tiered into a backbone, a distribution layer and an access layer. He said the core of this expanded storage network is the director, with director partitioning and routing technology allowing separate San islands to be converged and share resources, without the need for them to be merged into a single San.
However, IDC's European storage program director Claus Egge said tiered networks for storage are not the only answer to growing complexity. "A number of Sans out there are not connected, even though they're next to each other, because they have different types of data, or servers that don't mix."
Vogel disagreed, "The focus will shift from switching to routing in software. A lot more routing technology will play out in the next 12 months. Ultimately this infrastructure becomes an application grid - we're not there yet, but customers are already migrating to centralized management and network provisioning."
Bryan Betts writes for Techworld