My Two Cents On Cisco’s Big Data Center Announcement

Lots of technology vendors have been poking at Cisco lately with new technologies, architectures, and even out-and-out badmouthing (didn’t one CEO say that his company would “kick Cisco’s butt?). Well Cisco fired back today with a soup-to-nuts data center announcement covering virtual servers, physical servers, Ethernet switches, and storage. Cisco’s primary message here is simple: While all you other guys fight over the piece parts, we will out execute you by stiching the whole thing together far better and faster than you can. This actually makes a lot of sense because in truth, technology is not holding back further server virtualization or even cloud computing progress. What is? IT operations, performance management, IT collaboration, and security to name a few concerns. If Cisco architecture can help large organizations overcome these challenges, who cares if another vendor has a faster widget? A few other more specific observations:1. Very interesting to me that the Nexus 3000 is targeted toward “market trading” in financial services. On the offensive front, Cisco is taking its architecture up a notch from data center fabric to data center fabric for vertical industry business processes. With some effective marketing and custom professional services, this gives Cisco a play to compete with IBM, HP, and stuff shirt services firms like Accenture. On a defensive note, Cisco has felt a lot of competitive pressure from the likes of Arista and Juniper in financial services so Nexus 3000 goes a long way toward defending Cisco turf.2. Cisco’s fabric extenders provide a lot of choices for enterprises and cloud service providers that base their data centers on virtual servers. While this makes Cisco’s architecture extremely flexibile, ESG data indicates that most organizations haven’t a clue how far to push virtual switches, when to default to physical switches, or how to mix virtual and physical switches into a cohesive architecture. Cisco could really clean up here if it supports its new fabric extenders with use cases, reference architectures, and professional services help.3. A lot of my storage colleagues believe that Fibre Channel is rock solid and will remain the storage transport of choice for years to come. My response is, “that’s what the telecommunications guys said about VOIP.” I’m sharing this wisecrack because I believe that Cisco’s multi-hop FCoE for the Nexus 7000 and MDS will help accelerate the Fibre Channel to Ethernet storage transition. The storage guys won’t like this but they better acknowledge it. 4. Cisco had trouble with MDS encryption service performance in the past so the jury is out on its new offering. Cisco missed the tape encryption activity of 3-4 years ago but it is in good shape for disk encryption if its product performance is strong.5. Wish there was some news about Jawbreaker and where it fits vis a vis FabricPath, Trill, SPB, etc. 6. According to ESG Research, 28% of organizations will buy an integrated computing platform (i.e. server, network, and storage) in 2011. Most of these firms are either big enterprise or service providers. Cisco’s new UCS announcements seem very well positioned for this vocal minority of the market.If you are a Cisco shop or just interested in a more integrated data center fabric, this series of announcement has to grab your attention. If you are a Cisco networking shop, this announcement was interesting but not earth moving.