Wi-Fi Forecast: Cloudy with a Chance of Showers

by Jake Shehan on January 8, 2013 · 0 comments

in Cloud Computing,Mobility,Networking,ScanSource Security,Technology,Wireless

Ruckus

Contributed by: Marcus Burton, Technical Marketing, Ruckus Wireless

Wireless architectures are undergoing an identity crisis.

As Wi-Fi gains favor and usurps wired access, Wi-Fi capabilities are changing quickly, causing significant disparity in WLAN architectures and implementation models.

These shifts are causing customers and vendors to assess and reassess network management, monitoring, system control, and optimization of WLAN system that are compatible with yesterday’s devices, optimized for today’s devices, and ready for tomorrow’s devices.

In this state of flux, organizations of all shapes and sizes are asking similar architectural questions to find the best way(s) to deliver a wireless LAN:

  • Controller or no controller?
  • Hardware, virtual, or cloud controller?
  • Central or distributed data flow?
  • Cloud or no cloud?
  • Public or private cloud?

The only clear answer today is “yes.”

Though many industry pundits and suppliers are focusing exclusively on a single delivery model, enterprises (each with unique business needs) don’t agree which model is best or that any one model is the ultimate panacea.

Clouding the Architectural Wireless LANscape

Cloud computing is beginning to play a part in the Wi-Fi architecture debate, because—like many other segments of computing—it offers highly scalable capabilities that are difficult or expensive to deliver locally. The central business benefit to cloud networking is that a business of any size can now have access to an enterprise-class wireless solution that won’t overwhelm the IT staff or break the IT budget.

Clouds—whether private or public—are also enjoying favor in many business environments where distributed solutions are necessary (retail is a quintessential example). Cloud networks provide a graceful plug-n-play deployment model for remote sites and remote employees where IT staff resources are limited or non-existent. Because cloud management can be accessible from anywhere, distributed or centralized IT teams can easily manage and monitor distributed sites.

Distributed organizations see value in cloud Wi-Fi, but another major cloud formation is simplifying the deployment and management lifecycle by managed service providers (MSPs). If the cloud solution is optimized with MSPs in mind, it can make the business model much more effective, largely because of easy-to-access remote management, monitoring, reporting, and troubleshooting.

An additional element of the cloud’s appeal is the perception of resiliency, redundancy, and stability—in a properly designed and implemented cloud infrastructure. All the cloud buzzwords (e.g. high availability, elastic, redundant, seamless failover) make businesses feel warm and cozy. After all, mission-critical Wi-Fi demands mission-critical reliability.

Two Types of Clouds

Today, two primary cloud models are being espoused: (1) customer-owned [private] and (2) supplier-hosted [public].

Private clouds are attractive because businesses own the liability of customer and employee data. They want to own, secure, and protect it themselves, and they don’t mind accepting the responsibility for implementing and supporting it, so they deliver a centralized datacenter model where services and management are accessed from remote sites via VPNs.

Many leading Wi-Fi suppliers today are encouraging this model by offering a high-capacity centralized WLAN controller that supports “remote” or “flex” AP models. Private clouds are attractive for many large enterprises that already have significant datacenter investments, but they can lack some of the scale, resiliency, and cost advantages of public cloud options.

Yet the term “cloud” generally refers to public clouds, which provide all the benefits of releasing control, an attractive gain for smaller businesses. Someone else designs and runs the datacenter, accepts the complexity, secures the information (hopefully), provides high capacity/redundancy, and pays the power bill.  The business buys APs, signs up for a service, configures them through a simple and sexy web interface and can remotely monitor and manage the WLAN from anywhere. This changes the traditional WLAN model.  The wireless LAN becomes a service and can be effectively accounted in such a manner.

Distributed organizations are drawn to public cloud options, but despite solving the centralized management and monitoring needs, public clouds don’t solve the need for a centralized datacenter within the organization. Remote sites often need access to centralized resources via VPN, but a public cloud leaves this need unmet, minimizing the advantages of the public cloud.

When It Rains, It Pours

Despite their billowy appeal, cloud is not the be-all, end-all solution for Wi-Fi – not by a long shot. Some businesses balk at the privacy and control aspects of hosted solutions (what exactly are you doing with my information?), while others simply don’t buy the pricing ownership model—the perception is that cloud is akin to a rental model with less control and higher costs over time. The pricing reality depends, in part, on the expected lifespan of local alternatives (controllers or other management solutions). If the product lifespan of local appliances is expected to be long, customers may see more value in a “buy once, own forever” approach.

For others, the ownership hesitation comes back to a more traditional philosophy related to in-house expertise, where network staff wants to see, touch, and visibly troubleshoot their network with immediate, tangible responses to problems and outages.

Second, cloud Wi-Fi architectures either decentralize controller functions (controllerless) or they move the controller into the cloud. In some environments, this can be a plus because it removes controller hardware at each site—useful in some distributed networks. However, the same “no hardware controller” solutions must then find alternate ways to provide centralized services at each site, when desired.

In a somewhat self-defeating twist, some other local component is necessary to fill in the gap for specific features. This component is often called a gateway, concentrator or some tunnel termination device that provides scalable, centralized data tunneling, which is useful for a number of reasons (avoid LAN redesign for wireless VLANs, securely tunnel guest traffic, provide VPN termination, etc.). Some cloud Wi-Fi solutions also require a per-site appliance for centralized control functions, like roaming across subnet boundaries.

“Controllers” have traditionally been designed for central data tunneling, but new trends are focusing on distributed data planes (data breakout from the AP) while keeping the controller for management and “control” plane functions such as radio frequency (RF) resource management (channel and power settings), AP configuration settings, authentication services (802.1X or captive web portals), layer-3 roaming, and more.

For most customers, how and where system control is performed (distributed, centralized, or cloud) doesn’t really matter. What’s most important is how well system control works.  Consequently, when customers weigh various feature capabilities, the “how” argument often becomes philosophical. Moreover, enterprises want choices, flexibility, and most importantly, they want meaningful solutions for their business. Cloud or no cloud, architectural boundaries are becoming less clear.

Finally, a public cloud controller/management solution offers the reliability and redundancy benefits of cloud architectures. But architectural reliability is only one piece of overall wireless service availability. The potential benefits of cloud resiliency may be outweighed by alternative solutions that provide much better wireless stability via better radio design, adaptive RF features, antenna optimizations, interference avoidance, and the like.

Wi-Fi will always have its foundation at the radio level. Customers often understand the challenges of consistent, reliable delivery of wireless applications in high-interference or high-density environments. When customers must choose, the fundamental requirement for good wireless connections often plays a premium above the cloud’s sex appeal. In part, this is why we’ve seen some companies dwelling on the wireless component of wireless LAN equipment, optimizing features that improve capacity, reliability, and range—as well as adaptive features or RF visibility solutions. Ultimately, customer testing proves out the RF capabilities and customer priorities will always guide the decision.

Clearing Things Up

Obviously, customers want the best of all worlds: intuitive management, excellent data analytics, easy implementation, and adaptive, reliable radio performance.  What many fail to understand is that Wi-Fi reliability and performance will never be helped by anything that cloud computing offers.

Organizations must look for suppliers that offer a full range of architectural alternatives from controller-based to standalone APs, private cloud controllers to public cloud services. Despite what works best for a given organization, one fact remains clear: wireless reliability and performance must underpin any architectural choice. Without it, you’re left with an easy to manage Wi-Fi network that nobody uses.

Original source: http://www.theruckusroom.net/2012/12/wi-fi-forecast-cloudy-with-a-chance-of-showers.html

This post was written by

Jake Shehan is a Merchandising Team Leader in the Security Business Unit of ScanSource Inc. He has over 8 years’ experience in channel sales mostly with the network video and WLAN segments of the market.

Learn more about this topic at scansourcesecurity.com >

Leave a Comment

Previous post:

Next post: