One of the first questions asked by the non-cloudy types is, "If my network infrastructure is based in the cloud, what happens when I lose my Internet connection?" The short answer is that, usually, the environment stays as it was at the time of uplink loss. Data traffic is forwarded locally anyway, but management and control functions are on hold until the link back to the cloud is restored. Among the Flex 7500's features is the ability for user authentication on a local RADIUS to ensure client access even during WAN failure, when new clients can't be verified by the central corporate credential store. This is very welcome, as more branch sites leverage high-bandwidth 802.11n access points to get closer to going all wireless.
With scalability in mind, Cisco's latest conjures up images of supporting the likes of big-box chain stores, the hospitality industry, and government and education, based on its capacities. Able to scale to 500 branch offices with up to 50 802.11n access points each, the Flex 7500 promises to centralize control and management of up to 2,000 access points when the math involved with the distributed sites lines up. Given that controllers increase cost and complexity where they are used, it's nice to think about not seeing one when you open the closet door at the branch office.
Like the competition, Cisco touts the advantages of central management and monitoring of geographically far-flung access points. Cisco also brings the crown jewels of its feature set to the branch space, with support for Clean Air and automatic RF management. Add rogue detection, guest access, and local switching of wireless voice and video, and the Flex 7500 becomes a "real" Cisco wireless solution that's not missing much.
The Flex 7500 sounds great in many ways. I hope that with its launch, and the recent introduction of Cisco's 2500 and WiSM2 controllers, that Cisco has managed to get its wireless hardware quality control act together. As a long-time Cisco thin wireless customer, I have been astounded by the product set's frequent bugginess. In fairness, the last couple of controller software versions have been mostly good, but there are still far too many items on the "caveats" list in the release notes of each new version of code, in my opinion. Dealing with a wireless system bug at your big central site is one thing, but distributing it to up to 500 sites and dealing with the aftermath would be terrible. Here's hoping Cisco has turned the corner in this regard as it forays into cloud-based wireless.