skip to main |
skip to sidebar
A loud-talking ranchman applies to a banker for a loan. The banker asks a neighbor if the rancher is a good credit risk. The neighbor ponders for a moment and then replies “Big hat, no cattle.” False bravado is funny when it’s the stuff of fiction, less so in real life – especially for customers snagged by rhetorical barbs.
And yet it happens again and again. Each year the networking world is introduced to “big hat” products with features and specifications so too-good-to-be-true that we let ourselves be reeled in. Why we don’t see through the shiny veneer and ask for proof of pedigree is a wonder. But it happens all the same.
The Lessons of Wi-Fi #9: use analysts and audited financials to validate vendor claims. Neutral independent industry analysts like Burton Group, Canalys, Gartner, IDC, Infonetics, InfoTech, and Yankee Group can quickly assess vendors' technical claims.
And don't forget to check financials - audited financials - because you want your vendor to be in business should you need assistance or spare parts. If a vendor won't give up the numbers - or the numbers are substandard - then you have grounds for real concern.
A quick example will put the discussion in context. In 2008 a “big hat” four-radio 802.11n access point was announced that claimed to deliver 1.2 gigabits-per-second of aggregate capacity. The data sheet claimed that the four radios worked in tandem, enabling users to dramatically reduce the number of access points and additional security sensors, thereby reaping savings on cabling, connection and installation costs.
Still, the press ate it up. A flurry of articles expounded the virtues of delivering multiple HD streams to an entire building, with perfect coverage, at almost no cost. The world would soon be saturated with multi-adio APs, the unwashed masses blanketed with 802.11n. Wow, where do I sign up?
Fast forward to late 2009. The “big hat” super duper access point was no more. It simply vanished from the vendor’s Web site, its demise a secret. Was it ever built? No. But the company received undeserved publicity and that reeled in some unsuspecting customers.
To paraphrase Orson Wells, companies should herd no cattle before their time. Industry analysts can help you separate claims from reality. If an analyst says that a vendor can't excute well, refuses to divulge shipment numbers, and/or lacks technical vision - well, your due diligence is over.
The next time you see or hear about a product that appears to be too good to be true, separate the hats from the herds - kick the tires, test the features, validate the design. Those impressive features might be chimeras or, as with Aruba's AP-105 802.11 Access Point, the genuine article.
By any measure the California State University (CSU) system is enormous, encompassing 23 different campuses, nearly 450,000 students, and 48,000 faculty and staff.
Recently the university system was faced with a massive and potentially hugely expensive wired network refresh to upgrade infrastructure that was approaching the end of its service life. At the same time, the CSU system was experiencing a surge in the demand for network access across all of its campuses. In the absence of a budget for a Wi-Fi solution, which would have allowed one wired port to be simultaneously shared among many users, the IT staff was concerned that the need for Ethernet ports and switches would double.
What would you do in this circumstance? Expand the wired network? Seek additional funds for a wireless initiative? Restrict access to the network?
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #8: you can fund your Wi-Fi deployment by rightsizing your wired LAN.
Cisco suggests that the right solution was to expand the wired network with perhaps a smattering of wireless in lecture halls. Why? In a paper titled True-Sizing the Network, Cisco claims that Ethernet is future proof, more secure, and more reliable than wireless networks. In fact it marginalizes Wi-Fi, relegating it to situations in which Ethernet cannot otherwise be used.
The twisted “true-sizing” message short changes end users because it fails to take into consideration changes in user preferences, markets trends, and technology that have occurred in recent years:
- iSuppli reports that shipments of laptops surpassed desktops (38.6M vs. 38.5M) in 3 Q 08;
- Yankee Group estimates that enterprises with no Wi-Fi access will drop from 43% in 2006 to just 3% in 2012;
- Burton Group states that 802.11n marks the beginning of the end for wired Ethernet as the dominant LAN access technology in the enterprise;
- Best-in-class Wi-Fi networks sport WPA2 encryption, wireless intrusion detection, policy enforcement firewalls, and FIPS 140-2/Common Criteria/DoD validation - making them equal or more secure than most wired networks.
The best solutions for end users originatefrom understanding how and where they want to use the network, and then designing networks that meet those needs.
Aruba's network rightsizing program defines just such a process - measure wired port utilization, consolidate ports in use into fewer switches, and deploy 802.11n wireless to address mobility needs. Use Wi-Fi everywhere you can, wired networks only where you must. If savings are to be had, the rightsizing analysis process will tease them out. If not, then that will also be made clear. Either way, the network rightsizing analysis will offer insights into network and port utilization that might not be intuitively obvious.
Returning to CSU, what the IT staff decided to do was to obtain more data by measuring wired port usage. What they found surprised them: wired ports across all 23 campuses were consistently underutilized. More than half of the wired ports had passed no packets during the previous six months.
Armed with these data, the team decided to embark on a new approach. Instead of upgrading the entire wired network, something they had historically done every 4-5 years, they looked at the opportunity before them with fresh eyes.Wi-Fi was determined to be a reliable, low-cost option for delivering pervasive campus connectivity. Several campuses had already deployed some Aruba wireless LAN equipment, mostly for coverage in selected high-usage areas, and San Diego State University had built a relatively large WLAN on their campus. The Aruba WLAN had proven to be highly secure, scalable and reliable. It also allowed for a scaled-back refresh of the wired network, saving money by limiting upgrades only to the wired ports that were actually used.
CSU's IT staff created a database that included every telecommunication room, the number of ports in each room, and the number of those ports that were actively used. A formula was developed to define the refresh requirements of each of the 23 campuses based on this measurement.
By applying this formula across all 23 campuses, CSU was able to save approximately $30 million by reducing the scale of the wired network refresh and enhancing network access with Aruba’s Wi-Fi solutions.
The CSU system still uses wired networks but they've been rightsized to address actual and projected utilization. Wireless network utilization has risen sharply, because users are taking advantage of the mobility afforded by the expanded 802.11n network. And CSU saved a whopping big chunk of change that can be applied to other programs and opportunities.
Network rightsizing is a proven method of assessing and adjusting your network infrastructure. The California State University rightsizing program is a testament to the validity and value of the rightsizing model.
While the rightsizing mantra is to use wireless wherever you can, wired only where you must, the model makes no presumptions about the right mix of wired and wireless access. Proponents of “true-sizing” maintain no such neutrality. Their bias towards Ethernet marginalizes Wi-Fi, and in so doing deprives end users of the potential cost savings and mobility/efficiency gains that organizations like CSU have obtained.
Introduced in July 1979, the Zilog Z80 was an 8-bit microprocessor that operated on 1, 4, 8, or 16-bit data, had a 16-bit address bus, generated its own RAM refresh signals, and would run programs originally designed for Intel’s 8080 CPU. The flexibility of the design made it suitable for a very wide range of consumer, industrial, and military applications spanning from the Tandy TRS-80 computer to programmable logic controllers to naval weapon systems. Prices fell as volumes rose, and the Z80 was one of the most popular 8-bit CPUs for many years following its original introduction.
One of the wonders of semiconductor technology is that a standard part like the Z80 can find its way into so many different applications. The very same CPUs, memories, amplifiers, voltage regulators, and/or transceivers found in consumer products in your home might be found in automobiles, office equipment, factory production lines, airplanes, or ships. What differs is how the part is applied, packaged, and tested. In other words, you don't always need custom parts made of unobtainium to perform specialized tasks in demanding environments.
What happened to the Z80 in the 1970s is happening today with 802.11n chip sets. Chip set vendors are designing a common set of 802.11n parts for use in enterprise, SMB, gateway, and home access point and router products. Doing so drives up the volume of sales, resulting in production economies that boost profit margins for chip vendors even as prices fall for end users.
One of the largest Wi-Fi chip vendors – Atheros – sells its AR9002AP-4XHG chip set for all of the above referenced applications. The chip set features extensive component integration, a small form factor, and low overall cost. The fact that the AR9002AP-4XHG finds its way into such a diverse range of applications speaks volumes about the potential flexibility and robustness of the design. I say potential because whether the objective is realized or not depends on the implementation of the final Wi-Fi device.
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #7: you don't need unobtainium to build great Wi-Fi products.
Just as naval weapon system vendors leveraged a common Z80 design to create very unique and rugged products, so, too, has Aruba leveraged an 802.11n chip set targeted at a broad market in the design of its unique AP-105 802.11n Access Point. The AP-105 was tailored to demanding enterprise applications, and special care was taken in the design of the packaging, antennas, power supply, and security features to make the product both robust and exceptionally fast. A great AP, with a great standard 802.11n chip set, selling for a great price.
The result is an enterprise-class 802.11n access point that has higher throughput and more features than Cisco access points, yet sells for roughly 40% less money. So much less that Cisco felt compelled to pull apart the AP-105 to find out what makes it tick (they did the same when Aruba's high-end kick-ass AP-125 802.11n Access Point was released).
Their conclusion? The AP-105 is unobtainium-free and therefore no better than a consumer product. You know, like that cell phone you rely on for emergency calls 24x7, or that iPod that has delivered faithful service every day at the gym. Comparing the reliability of the AP-105 to that of a consumer product is not an insult. At the end of the day, Cisco still has to explain why the AP-105 is faster, more feature rich, less expensive, and easier to install than its own run-of-the-mill, over-priced, unobtainium-based access points.
So with the wind of good design at our backs, and unobtainium nowhere to be seen, the AP-105 is flying off the shelves, charting a path the Z80 followed.
Would you ever strap a PC to your ceiling and run it there? Probably not. What about inside the plenum space above the ceiling? Nope.
Accessibility aside, the ceiling and plenum are hostile environments for electronics that aren't specifically designed for the vibration, temperature extremes, and blown dust typical of these locations.If you look inside devices designed for this environment - smoke detectors, passive infrared sensors, quality Wi-Fi access points - what you WON'T find are vibration-sensitive connectors (like SIMM sockets), moving parts (like fans), and modular circuit boards that could wiggle loose. These devices are typically designed to have high mean time between failure (MTBF) ratings, something impossible to achieve with commercial SIMMs or fan-based power supplies. It seems so intuitive...and yet.Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #6: sleight-of-hand is no substitute for good product design. Wi-Fi access points need to be designed from the ground-up to withstand the rigors of ceiling and outdoor mounting environments.
Consider Wi-Fi arrays, which are effectively PC motherboards with a fleet of sockets, add-on modules, plug-in connectors, and memory SIMMs. They even conjured up a fan-based PC-like power supply - no standard 802.3af Power-over-Ethernet here. And when it fails you've lost 4+ radios at one time. The only workaround is to double-up the number of arrays, a real budget sink. Arrays just aren't designed with long service life, energy efficiency, or network resiliency in mind. That's the reason why no leading vendors in the Wi-Fi market sell arrays.
Aruba Wi-Fi access points have no fans, no SIMM sockets. Our 802.11n access points are designed for the rigors of ceiling and plenum mounting, and run from standard 802.3af PoE. MTBF ratings are excess of 250,000 hours - more than 28 years. And should an access point go down, Aruba's Adaptive Radio Management adjusts the power of near-by access points to self-heal the coverage gap. Automatically.
They'll provide years of reliable service and are backed by a lifetime warranty. And they cost less than an array-based system. A lot less.
So the next time you consider upgrading your wireless LAN, think about the environment in which the equipment will be used. Reliable products don't happen by magic - they happen by design.
Let's consider an alternate ending to Lesson #4. You need wireless access across an entire floor of your building, and a Wi-Fi vendor with shiny white tasseled loafers planted on your desk says he has just the solution: a single16-radio access point that will provide coverage across the whole floor and will save you a bundle in installation costs. How can you go wrong? Think of the cost savings: only one access point to buy, only one access point to wire.
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #5: eggs break - don't put them all in one access point.
What appears alluring at first glance is really false economy. One single failure and there's nothing between you and a totally dead network - you'll have lost the entire floor. A 16-radio access point on a single cable sounds cool but it only gives you coverage – not capacity (you'll need a lot more radios, cables, and switch ports for that. And it offers no redundancy against failures like a dead CPU or memory.How about just throwing in a second 16-radio access point for redundancy? Even if you could align it to deliver the same coverage pattern, your hardware costs would be blown sky high. And if you're using 802.11n, you’ll further drain the bank by needing additional expensive power supplies and even more cables and ports.
With a multi-access point, multi-channel design, any coverage gap created by the loss of a single access point is mitigated by nearby access points. Load balancing handles high density scenarios while airtime fairness handles different mixes of 80211a/b/g/n clients. And using separate access points allows you to cover rooms and labs and lathe walls and metal-foil wall paper that can't be penetrated from outside - even by a single, centrally-located 16-radio array.
The question to ask yourself is what is the cost of a failure? How much will you lose if the entire office wireless network goes down for a day? Or students can’t access the Internet? Or a trade show network stops running? For most users, the cost of putting all of your eggs in one access point is too high.
You've now discovered why no major wireless LAN vendors pack so many radios into a single access point. It's false economy because it puts your business at risk should a failure occur.
And as far as cost differences, they've all but evaporated with Aruba's newest 802.11n access points. You don't need to take my word for it - Gartner's 2009 Wireless LAN Infrastructure Magic Quadrant spells it out in black and white.
If you'd like to get the whole picture on Wi-Fi architecture you've only to download our free white paper, WLAN RF Architecture Primer. And leave it to someone else to relearn the lessons of Wi-Fi.
You've invited Wi-Fi vendors to your facility to discuss a new Wi-Fi project. You need wireless access across an entire floor of your building which includes open plan seating, conference rooms, and executive offices. This will be the primary form of network access and it needs to work. All the time.
It's late afternoon. A Wi-Fi vendor sits across from you in his white suit and black shirt, the very model of semi-neo-avant garde stylin. His shiny white tasseled loafers are firmly planted on the corner of your desk. He looks you straight in the eyes and says that his access point transmits radio signals farther than anyone else's. "It uses special technology. Yes, it's expensive, but by packing sixteen super duper radios in one unit you'll save a bundle because you only need one access point to cover the entire floor." Wow! How can you go wrong?
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #4: we all live by the same laws of physics, and no Wi-Fi vendor has yet bent them to their will.
The maximum output of a radio at any given frequency is dictated by local regulatory agencies. In most countries 100 milliWatts is the upper limit of what an indoor access point is permitted to output. Regardless of vendor and irrespective of Wi-Fi chip vendor - Atheros, Broadcom, Intel, etc. There is a level playing field when it comes to building radios.
What vendors can do is twiddle with antennas, using directional antennas to focus the allowed radio energy into more well defined beams. And, indeed, doing so can project radio signals longer distances.
The issue is that Wi-Fi networks are bidirectional - there's something on the receiving end of those directional antennas. Low power clients like iPhones and netbooks aren't equipped with directional antennas, much less ones that are easily focused on access points. They may be able to hear distant access points but the access points may be unable to hear them - even if directional antennas are used - because they don' use high power radios.
Additionally,as we learned in Lesson #3, bit rate is inversely proportional to range. In a shared medium like 802.11 where only one device transmits at any one time, lower data rates mean less available air-time for data on that entire 802.11 channel. So even if an access point and its clients can communicate, the throughput from the clients to the access point will be relatively low. Not good for voice. Not good for video. Not good for you.
You don't get something for nothing, but you can find yourself with nothing from something. The Wi-Fi standards anticipated the use of multiple access points, and that's how clients are designed to work. Pushing the limits of how far a Wi-Fi signal can be made to propagate has heuristic value, but when it comes to real-world deployments it can jeopardize the functionality and reliability of your network.
It's best just to tell the vendor to take his shoes off your desk and sell his wares elsewhere - you're having none of it.
If you'd like to get the whole picture on Wi-Fi architecture you've only to download our free white paper, WLAN RF Architecture Primer. And leave it to someone else to relearn the lessons of Wi-Fi.
You're excited - the two bids you were expecting for your new Wi-Fi network have just arrived. You rip open the envelopes and then stare in disbelief.
The first bid - the low bid - includes fewer than 100 access points and a note stating that the access points are specially designed to operate at full power at all times so fewer are required. The second bid includes 135 access points and a note about meeting bandwidth capacity requirements and providing resiliency in the event of failure. Both vendors had the same set of plans to review, both did a walk-through of the facility. How could their bids be so different?
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #3: wireless coverage ≠ wireless capacity. Designing for coverage means providing a discernible Wi-Fi signal everywhere without regard for network speed. The access points on these networks are typically run at full output power so the signal coverage is max'd out. They're also spaced with minimal or no overlapping coverage. As a result fewer access points are required.
The downsides of designing for coverage? Many. Consider these two:
- Bit rate: There is an inverse relationship between bit rate and range. The farther away a Wi-Fi device moves from an access point, the lower the bit rate. Wi-Fi devices operating at the fringe of the coverage area will be very slow indeed. Too slow for voice, streaming video, electronic white boarding, and many other applications;
- Failure happens - but this design can't deal with it. If an access point fails, nearby access points can't increase their output power to fill in the coverage gaps.
Designing for coverage is okay if consistent network performance and resiliency are unimportant. Otherwise it should be avoided.
In networks that are designed for capacity, the required bandwidth is available throughout the coverage area. Application performance will therefore be universally uniform.
Planning for capacity requires more access points because the distance to laptops, iPhones and other clients needs to be more limited (remember rate vs. range) for robust, high-speed operation. They're also needed to ensure adequate load balancing, a feature especially important in areas with densely packed clients such as classrooms, lecture halls, and trading floors. The benefits far, far outweigh the cost - you end up with a resilient network on which you can consistently depend for years of service.
Some vendors play on customers' lack of familiarity with the difference between coverage and capacity. When it comes to reviewing bids and proposals, take note of differences in the number of access points and claims about "unique" features affecting coverage. If you fall for the coverage
If you'd like to get the whole picture on Wi-Fi architecture you've only to download our free white paper, WLAN RF Architecture Primer. And leave it to someone else to relearn the lessons of Wi-Fi.
One the reasons for creating technical standards is to ensure interoperability between devices that need to work together. In the Wi-Fi world, the 802.11 standards serve this purpose, and encompass a very extensive set of guidelines that manufacturers of infrastructure and devices must follow to create a cohesive wireless system. Why then do we encounter situations in which Wi-Fi infrastructure is incompatible with Wi-Fi devices?
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #2: not all Wi-Fi networks are standards based. Some use proprietary technology that is not compatible with the way other Wi-Fi device manufacturers have designed their products.
Take, for example, Glenelg Country School and Frances Xavier Warde School, both of which experienced dropped connections with wireless classroom multimedia projectors. At Raytown C-2 School District radio interference affected laptops on rolling computer carts, while at Prairie Cardiovascular Consultants interference was so bad that it affected both office and clinical operations. Others have reported issues with different models of PCs or Apple Macintosh computers and iPhones.
What's interesting about these cases is that the problems were traced to one common source: the wireless LAN infrastructure. Once the infrastructure was upgraded - in these cases to Aruba wireless LANs - the problems went away.
All of these sites had used a non-standard, proprietary single-channel wireless LAN architecture. There are only two companies in the industry that make such systems, and both are small niche players with shrinking market share. So why would anyone buy such non-standard products in the first place?
Simple - product differentiation can be very alluring. It offers the opportunity for the adventurous to tout themselves as early adopters of what they hope will be "the next big thing." Wanting to be the first to use a new Apple iPad, Alienware laptop, or Google Nexus One makes perfect sense. These products embody innovative designs that redefine their markets. But they're also designed to work with existing networking infrastructure like 802.11 Wi-Fi - that they didn't redefine.
Where you run into serious trouble is deploying non-standards based infrastructure. That's akin to
being the first to try a 156 Volt, 76 Hz electrical system in your house. Some devices might work, but you run the very considerable risk that others will crash and burn.
And that's what happened to the single channel wireless LAN customers. The reason single channel architecture hasn't caught on isn't because it's a secret waiting to be discovered. It's because there's a secret to what makes it run, and therefore interoperability is not assured.
When it comes to living on the bleeding edge of technology, consider the importance of interoperability. If a new technology has to be seamlessly integrated with other existing devices - as is the case with Wi-Fi networks and devices - then using a non-standards based product is just asking for trouble.
If you'd like to get the whole picture on Wi-Fi architecture you've only to download our free white paper, WLAN RF Architecture Primer. And leave it to someone else to relearn the lessons of Wi-Fi.
You've invested thousands - tens of thousands - in new educational software, a fleet of new Wi-Fi enabled laptops, and even computer carts to chauffeur computers between classrooms. But when the students fire up the machines and try to access the shiny new instructional video you're trying to stream wirelessly, they get nothing. Nothing but static. Or jitter. Or dropouts. What went wrong?
Those who forget the lessons of Wi-Fi are doomed to repeat them. Lesson #1: not all Wi-Fi networks are created equal. They all have access points, and they may even be Wi-Fi Alliance certified. But the similarity ends there.
Streaming real-time video is a demanding Wi-Fi application that requires additional processing above and beyond the Wi-Fi standard. The main technology enablers for video over Wi-Fi are adequate bandwidth, quality of service (QoS), and multicast support.
While 802.11n - the newest high-speed Wi-Fi technology - provides a significant bandwidth boost, RF management algorithms are important to ensure continuous, high-rate coverage. These algorithms must include control of the access points and the laptops (clients) - a feature provided by Aruba's Adaptive Radio Management technology - to automatically calculate the optimum channel and transmit power assignments, move clients to the most appropriate access point, and optimize the network’s use of available radio spectrum. This function is especially important for mobile clients - like iPhones - and in the presence of densely deployed clients such as you would find in classrooms and lecture halls.
QoS for video uses the same mechanisms as for voice, however, the bandwidth requirements of video applications vary widely. It is therefore important that any content that requires special handling be correctly flagged. Aruba's integrated stateful firewall does just that.
Since video can account for a large percentage of network bandwidth, determining when to broadcast to multiple clients - multicast streaming - is essential. Here again, Aruba incorporates technology to monitor multicast group members, and only delivers multicast streams to access points whose clients require it.
If you'd like to get the whole picture on video over Wi-Fi you've only to download our free white paper, I Can See Clearly Now. And leave it to someone else to relearn the lessons of Wi-Fi.
One of the challenges of distance learning is how to replicate the "campus experience" for remote students. Doing so encourages collaboration with other students, and improves study opportunities, by leveraging the same electronic learning applications, library reference materials, and server resources as campus students enjoy.
It also builds school loyalty because if these services remain in place post graduation, it improves the chances of continued participation once students become alumni.
Providing secure access to your school's electronic learning resources is a challenge. Open access or password-controlled access won't protect against network attacks, password-sharing, or excessive bandwidth consumption by mischievous students.
A secure virtual private network (VPN) requires your IT staff to load and manage client software on every device a student might wish to use. This is an on-going burden because incompatibilities may be introduced as students upgrade operating systems or other applications on their computers.
Virtual Branching Networking (VBN) solves distance learning connectivity and security issues. Using a small, very inexpensive device called a Remote Access Point (RAP), VBN enables remote students to connect securely to your data network. RAPs enable students to use any IP-based devices with an Ethernet port or W-Fi - MacBooks, iPhones, iTouches, iPads, PCs, VoIP phones, printers - without loading any software clients.
A built-in firewall strictly enforces access policies set by your IT staff, and can even control how much bandwidth a student uses. All access policies are centrally managed and then pushed over the network to the RAPs. The same is true of software updates: they're pushed automatically to every RAP in the field.
New RAPs are shipped unconfigured. To connect one to your network the student pushes a button on the front of the unit and then enters the IP address of your data center.
A RAP controller in your data center then exchanges security certificates with the student's RAP and voila, the student is on-line. No IT staff involvement is required for this process to occur, meaning that it's possible to economically support a very large distance learning program without adding IT staff.
Since RAPs are shipped unconfigured, they can be sold or rented to students through your bookstore or by a third party with zero-touch involvement by your IT staff. If a student leaves your distance learning program, or fails to pay tuition, a simple change in the access policy will completely disable the RAP.
Alternately, when a student graduates the RAP settings can be changed to disable distance learning and enable Internet access using your alumni site as the home page.
VBN has been field-proven in enterprise teleworker deployments around the world, and is the ideal solution for distance learning applications of any size. To find out more please visit our Web site.
A loud-talking ranchman applies to a banker for a loan. The banker asks a neighbor if the rancher is a good credit risk. The neighbor ponders for a moment and then replies “Big hat, no cattle.” False bravado is funny when it’s the stuff of fiction, less so in real life – especially for customers snagged by rhetorical barbs. 
Posts
