The favourite f-word on everybody's lips these days is fabric. That shows what a sheltered life I lead, because most of the people I talk to are in IT.
The fabric they're talking about, of course is not the stuff drapes are made of but network fabric that unifies switches and servers in data centers and even across the enterprise.
Even since Cisco Systems Inc. outlined its FabricPath strategy for consolidating network layers, equipment makers have been putting out their own versions. The problem, industry analsyts say, is that there are few public case studies available to help guide organziations in implementing a fabric strategy.
In a pre-announcement interview, Mike Marcellin, Juniper's vice-president of systems strategy and marketing, told me that Bell -- a long-time Juniper customer with a traditional three-tier data centre architecture -- hasn't deployed its new converged architecture yet. And that's about as much as I can report so far.
By coincidence I was talking yesterday to Zeus Kerravala, former vice-president of enterprise research at Yankee Group, who has left to set up his own company called ZK Research, about the claims equipment makers put out about the scalability of their fabrics.
"Frankly, we don't know how big they can scale because we're in early deployments," he said. The market for fabric is "wide open" and the winner will be the vendor who can demonstrate value through detailed case studies.
"We really don't have a good industry benchmark on how to evaluate fabrics," Kerravala added: Should it be by the number of server-facing ports in one fabric? Port-to-portlatency? Do you need one fabric, or can a data centre support two to support thousands of servers?
I look forward to talking to Bell and other Canadian organizations about their fabric experiences.