The Open Compute Project (OCP) - started two years ago by Facebook to promote open-source hardware solutions for data centers - announced on Wednesday that it has started a new project to develop a specification and "reference box" for a "open" networking switch.

(See also Facebook's Group Hug Frees The Microprocessor From The Motherboard.)

The project will be led by Najam Ahmad, who runs the network engineering team at Facebook, and high-profile companies including Intel, VMware and Broadcom - among others - have already signed on to participate. 

The new project's goals include helping software-defined networking (SDN) "evolve and flourish" and creating "flexible, scalable, and efficient" data center infrastructures. Just as the Open Compute Projects' simple, generic and open server designs competed with server vendors like HP and Dell, this new initiative challenges networking leaders like Cisco. 

(See also Software-Defined Networking: What It Is, How It Works, Why It Matters.)

The project is expected to kick off at the inaugural OCP Engineering Summit to be held at MIT next week.

Image courtesy of Shutterstock

Beta testing can be a bitch - especially when you're working with complex business technology that doesn't make sense for consumers. It can be incredibly difficult to find good test subjects with enough of a knowledge base to give you intelligent feedback on these kinds of sophisticated products.

That’s exactly the issue facing Pertino as it prepared to launch its cloud-based network launch last fall.

Where To Find Qualified Beta Testers?

Pertino’s concept was to build a cloud-based global network, requiring no specialized hardware or virtual private networks. The company envisioned a network affordable enough for small and midsize (SMB) companies with the security and performance of an enterprise network.

How the heck do you beta test a product like that?

Pertino CEO Craig Elliott turned to the Spiceworks.com community of more than 2.4 million IT professionals, centered around the company's free, ad-supported IT management tools for SMBs.

Elliott and many Pertino employees were already Spiceworks members, and they started with a 20-company private beta program that grew into “a community-exclusive public beta” involving 250 “Spiceheads.”

Spiceworks’ co-founder Jay Hallberg says three to four years ago the Spiceworks team was “dreaming big that someday we’d have a company launch within Spiceworks.” Pertino turned out to be that company.

Beta Testing Feedback Is Essential

While the usual point of beta testing is to find out if your product is good enough to launch, most initial offerings end up requiring signficant tweaks. “If you’re not thoroughly embarrassed by the first product you release," Elliott says, "you’ve overthought it, and you’ve come to market too late.”

Beta testing in Spiceworks enabled knowledgeable IT professionals to actually use the product and offer Pertino “incredible, first-hand feedback and insights,” says Elliott.

Todd Krautkremer, Pertino's VP of marketing, explains that, “Since so many members of the Spiceworks community work IT at small and mid-sized businesses, it was a way to treat SMBs as consumers… The Spiceheads provided feedback in real time [that] shaved months off what the normal development timeline would be.” Beta testing in Spiceworld gave Pertino “validation and the ability to go back to the drawing board based on the feedback,” Krautkremer adds. If you can’t make it in the Spiceworks community, how can you succeed in the broader market?

Speed Wins

Pertino didn't worry about launching its “private” beta to such a large community. ""In the world of open-source tech,” says Krautkremer, “to rest your laurels on defensible IP is not a recipe for success.” Patents can't protect you.

Instead, seizing the market as early as possible is the best way to become a dominant leader, says Krautkremer. It’s not necessarily being first to market,” Krautkremer continues, “MySpace was there before Facebook.” To win, your idea has to be novel and simple, and you have to pursue it aggressively.

So far, that approach is working for Pertino. The company publicly launched its product in February: "6,000 people downloaded it on day one,” Krautkremer says, and more than 300 Pertino networks were built.

Beta Test Tips

Sharing what he learned from the Spiceworks beta, Krautkremer offers tech companies 4 quick tips:

1. Use the freemium model: make it easy for potential customers to try your product.
2. Keep it simple: “Click, click, done wins. Click, click, click, done loses.”
3. Eat your own dogfood: use and test your own product.
4. Get to market first and then grow fast.

Oh, and find qualified beta testers to provide useful feedback before you make your product publicly available.

Cloud computing is all about "abstracting servers," turning actual hardware into virtual machines and moving them out of your organization so you don’t have to worry about them. Big data’s non-relational databases and Hadoop clusters perform a similar level of abstraction on database administration. Software-defined networks (SDNs) may do the same thing with networking.

SDN promises to make high-capacity networks cheaper to build and especially to re-configure on the fly - as well as potentially faster and more efficient. As more and more computing moves to the cloud, those network improvements will be critical to keeping everything affordable and available.

It won't be just cloud environments that will get SDN benefits, either. SDN could enable corporate networks to be reconfigured on-the-fly. Imagine being able to plug in a multimedia-intensive device in a conference room and the network adapting to effortlessly handle the sudden new load without grinding the rest of the building's network traffic to a staggering halt. Or re-arranging cellular networks during a disaster to deliver connectivity when and where it matters most. That's no doubt why analysts like IDG are predicting the business of SDN hitting $2 billion by the year 2016. 

But Software Defined Networks are incredibly complicated, so even as many networking professionals look forward to SDNs as the best thing since sliced bread, others are scratching their heads and wondering if the IT hype machine has gone completely off the rails.

Who's right? To answer that question we need to look at exactly what is a software-defined network, and how do you create them?

Here's The SDN Theory

Think about a traditional network and everything that entails. You have your routers, your switches, and lots and lots of CAT5 and CAT6 cable strung around: all physical hardware that, when connected in a certain way, defines the flow of data in the organization. Like laying down a network of highways, planning a network takes time; it has to be done right the first time because shuffling things around afterward is expensive.

A network has to do two big things: deliver data and manage the flow of that data. If I am downloading a video from California, the network knows to get it to me here in Indiana. Shunting the data through India and Europe would not be the most efficient way to do it – unless, of course, some big physical failure occured between here and the West Coast that required the signal to be sent the long way around the planet.

Inside a company, the same thing happens on a smaller scale. Data is passed back and forth, and that traffic is usually managed by software inside the physical devices - software that knows how to manage the day-to-day operations of the workplace.

With cloud computing, the physical servers that hold the virtual machines are still networked together with the same routers and switches that are used in a workplace network. But the demands on that physical network can be much, much greater - at times - than anything your employees can dish out. (Which, really, is the whole point of using the cloud in the first place.)

What SDN does is this: Assume you have the network cable laid out between every physical server in the cloud environment and all of the optimized routers and switches. The SDN layer essentially acts a virtual software switch or router in place of (or in conjunction with) the physical network devices.

So instead of software embedded in the routers and switches managing the traffic, software from outside the devices takes over the job. The network layout, or topography, is no longer rooted in the physical. Instead, it's flexible and adjustable to the systems’ needs on the fly.

Properly implemented, this means an application running inside the cloud itself can take over the job of directing networking traffic. Or a third-party cloud-management application could do the job. That could make it easier to perform tasks such as load balancing devices across servers and automatically adjusting the network architecture to deliver the fastest and most efficient data paths at the right time.

Rules Of The SDN Road

There are risks involved in this kind of networking - namely those stemming from how complicated these kinds of operations can be.

Traditionally, networking "decisions" have been left in the hands of the the devices on which the network actually runs. That's what they are meant to do. Taking the control away from these specialized devices and the embedded software that runs them could be the prelude to a networking disaster, unless everything is done exactly right.

Done improperly, this would be akin to letting every driver in Cleveland have independent control of all the traffic lights in town. Chaos. This is why there's a gap between theorizing about SDN and actually implementing it.

Right now, the Open Networking Foundation's OpenFlow protocol and the Linux Foundation's new OpenDaylight project are two open source projects working to establish a set of SDN traffic rules that applications can use to prevent such chaos.

These protocols would also vastly simplify the work of application developers, who would not have to learn the nuances of networking control, but rather just call on one of these tools to handle the heavy lifting for planning the right path through a network, based on an application's need.

Think of it like getting a police escort through a strange city. You don't have to know anything about the city or worry about stops - you just get in your car and follow the vehicles with the flashing lights, trusting that they know what they're doing and where they're going.

SDN Benefits Could Be Widespread

Because of its complexity, SDN won't take off right away. But once SDN is implemented, the benefits will be immediately apparent. Cloud computing environments controlled by SDN will see significant increases in speed and efficiency, since their networks will be optimized for the applications running in that cloud environment. Corporate and mobile networks are also likely to gain benefits from SDN implementations.

IT will have to lay down the rules of the road first, but no one can deny there is big potential in software-defined networks to bring more adaptable networking resources to businesses and consumers alike. 

At one time, Twitter's most familiar icon wasn't its perky little bluebird. It was the Fail Whale, the image that Twitter displayed when the service was down or disrupted. 

The Tale Of The Whale

The Whale was the work of artist Yiying Lu, whose now-iconic image was picked up by Twitter in 2007 to jokingly commemorate the site's accumulated days of outages. Since then, the image of a whale borne aloft by several tiny birds has  inspired its own Twitter fan club and Facebook page, along with an online repository of dozens of illustrations.

But now that Twitter's online performance has dramatically improved, there's a new Twitter icon that you probably will never see: the Success Loch Ness.

On the evening of April 8, following the Shorty Awards, Lu and other employees from Twitter were hanging out at The Pony Bar in New York City, when Tom Spano, the events coordinator at Twitter, asked Lu, "Since Twitter’s now became more and more stable, there’s less chance for folks to see your image. How about something opposite from the Fail Whale? Success Whale?"

The proposal was shouted down, Lu describes on her blog, because it didn't rhyme. Instead, Lee Semel, the founder of the Shorty Awards, suggested the Success Loch Ness - named after the infamous Loch Ness Monster, often called Nessie. And the rest, as Lu suggests, is history.

Twitter Gets Its Act Together

Go back far enough in Pingdom's logs of Twitter's uptime, and you'll find what could be the nadir for the site: an abysmal 92% uptime in March 2007, equivalent to being down more than two days in that month alone - just after Twitter hit the big time at South by SouthWest (SXSW). For the next few months, Twitter's uptime averaged about 98% - better, but still not great. In Dec. 2007, Twitter moved to a new data center, which helped significantly. Twitter still suffered occasional outages during the next few years, including slowdowns when Michael Jackson died in 2009, a denial-of-service attack that same year and more slowdowns during the 2010 World Cup.

But over time, Twitter slowly improved its performance to the point where the service now reports stellar uptime results. In seven out of the last twelve months, for example, Pingdom credited Twitter with a perfect 100% uptime.

All that has put Twitter in a celebratory mood.

Here's Twitter's Tom Spano holding both the original Fail Whale and the new Success Loch Ness:

The launching of the "Success Loch Ness", taking over from the "Fail Whale" vine.co/v/btLajquBZg0

— Jason Seed (@jasoncseed) April 12, 2013

Birth of the #SuccessLochNess, Vined, with Musika! vine.co/v/bUOzVuvwQlM

— Yiying Lu (@YiyingLu) April 18, 2013

Too Bad "Success Loch Ness" Doesn't Make Any Sense

It's great that Twitter has been able to move beyond the Fail Whale, but while its replacement may be far more positive, the new creature has its own grammatical, style and content issues. Loch Ness is a lake, after all. The Loch Ness Monster, if it exists, lives in the lake. More to the point, Success Loch Ness just doesn't trip off the tongue with quite the same ease. 

Image courtesy of Yiying Lu.

Just as consumer technology invades the enterprise, business-class technology is becoming available in the home market. A great example of this counter-trend - call it "the enterprization of the consumer" instead of the consumerization of the enterprise - is Skydog, from Palo Alto-based PowerCloud Systems.

Skydog is a dual-band Wi-Fi router that also incorporates patented cloud-based intelligence to enable homeowners to optimize their network resources and manage bandwidth remotely from their smartphones. The idea is to make it relatively simple for parents to limit their children's Web usage by time of day, total amount of time, site, site type, device and activity. With the Skydog router and the company's HTML5-based data service app, users can monitor and manage their home network via iPad, iPhone and Android. They can also get personalized alerts about current usage or problems via text message. 

Skydog: The Router For Parents?

These opportunities for parents to take control of the home network could be Skydog's "killer app." If a parent is alerted via text that a child is streaming video after bedtime, for example, the Mom or Dad can temporarily remove the offend device from the network. If a user has multiple Skydog boxes set up, one at home and one in the home of their in-laws, say, both can be managed through the same account. No more hours on the phone helping out remote family members.

Not interested? That's not too surprising. Few people pay much attention to their home networks these days, except when it goes down. That said, in my brief demo of Skydog - both the hardware router and the data service - one thing quickly became clear: When home networking data across all devices is exposed in an intuitive, visual format, network management and troubleshooting can become empowering.

Still not sold?

Skydog claims to allow busy parents to see and control home Internet usage, even when they are at work or otherwise out of the house. Parents can be alerted, for example, if a guest attempts to access the network. They can optimize network bandwidth so that a Skype call from the home office always takes precedence over a child's YouTube habit. Children can easily be allotted special network access and usage privileges for the weekend or holidays. 

Why Skydog Chose Kickstarter

I asked PowerCloud Systems founder and CEO and founder Jeff Abramowitz  why a company with experience in the enterprise networking world - and located in the heart of venture capital country - would choose the Kickstarter route? "The sales channel (for networking hardware like Skydog) is dominated by incumbents," Abramowitz explained. "In addition, the Kickstarter campaign gives us buy-in with our target audience and allows us to work directly with those who want this solution." 

He also noted that a Kickstarter campaign, allows early adopters to provide valuable user feedback - and not only to the company itself. Skydog has also set up online forums so users can share best practices. 

If Skydog's Kickstarter campaign fails, what then?

It's at least possible that the company's patented cloud-based algorithms, which deliver network information to its users, can be embedded in other company's devices. "We are keeping our options open as it relates to licensing software, and we will learn more about the users and prospective partners interest in this area through the Kickstarter campaign," said Vivek Pathela, PowerCloud's VP of Marketing. 

Image courtesy of Skydog.

Networking giant Cisco predicted Wednesday that as we move into a "fundamentally mobile and video" world, the "Internet of Everything" — which combines the so-called Internet of Things with the Internet used by people and their mobile devices — will create $14.4 trillion in value and boost overall corporate profits by 21%. All by 2022.

Those are some pretty big numbers, shared by Cisco executives at a press event in San Jose on Wednesday. But while the vision makes sense, quantifying the changes to be wrought by growth of the Internet of Everything seems, well, fairly abitrary. To say the least.

What Goes Into $14.4 Trillion?

Rob Lloyd, Cisco President, Sales and Development, broke down the $14.4 trillion figure this way:

• $2.5 trillion in better asset utilization
• $2.5 trillion in employee productivity
• $2.7 in supply chain logistics
• $3.7 trillion in better customer experience.
• $3 trillion in enabling new innovations.

Perhaps we can start by seeing which industries benefit first and most dramatically. According to Lloyd, the top candidates include manufacturing, the public sector, energy and utlities, healthcare, finance/insurance, transportation and wholesale/distribution.

The Internet of Everything combines several trends, including the growth of connected devices, the increasing use of video, cloud computing, Big Data and the increasing importance of mobile apps compared to traditional computing applications.

Lloyd did lay out numbers to support the importance of the trends. But though these are also all giant numbers, connecting them to the $14.4 trillion figure still requires a leap of faith.

In terms of connected devices, he said, we've gone from 200 million in 2000 to 10 billion devices today, to a predicted 50 billion by 2020. On the mobile side, Lloyd said, 20 billion mobile apps were downloaded last year alone. By 2017, he added, two-thirds of mobile traffic will be video.

New Levels Of Complexity To Support New Uses

That complexity will make today's issues "look very, very minor," and pose historic challenges to manage, Lloyd said. That statement, at least, is easy to grasp. "We've been warming up for this for the last five years." The company already has a number of projects in the works demonstrating key elements of the trend, including installing smart meters and pole-top routers for BC Hydro in Canada, and a single auto plant with 50,000 IP devices.

Cisco's Sean Curtis demo'd live data from San Carlos, Calif., showing a heat map for mobile connections using "dwell time" metrics to track how efficiently pedestrian traffic was moving through the suburb's commuter train station. Curtis said similar information mashups have been applied to San Carlos' farmers market, offering insights into how many shoppers showed up, how long they stayed and which stalls they visited — information that would be of great use to both retailers and city planners.

The next step, Curtis said, is to link that kind of data with store data as well as parking and traffic information to help shoppers optimize their experience. The idea is that eventually shoppers could see the best route to the least crowded store with the best prices on the items they were looking for.

As apps like that come online, the Internet of Everything should indeed spur growth. Maybe even trillions of dollars worth of growth. Exactly how much and when, though, seems a Big Data question of the highest order. 

Rising Expectations, Bigger Security Issues

The rise of the Internet of Everything is already changing corporate expectations, Lloyd said, not to mention who pays for technology advances. "The Internet of Everything will be driven by business funding, not just IT funding," Lloyd said.

What about security for all this connected information? Padmasree Warrior, Cisco's chief technology and strategy officer, said "the data will be collected whether we want it to be or not. How will it be used? That is the security question."

Photos by Fredric Paul

Guest author Kris Barker is co-founder and CEO of Express Metrix Asset Management Software.

Even the smallest network is under threat from botnets, hacking, Trojans, denial of service (DoS) attacks and information leakage. Malicious or criminal attacks, the most expensive cause of data breaches, are on the rise and the consequences of poor network protection are harsh.

A Ponemon Institute and Symantec study published in March 2012 shows a jump in data breaches caused by malicious attacks from 31% in 2010 to 37% in 2011, with an average cost of $222 per incident. Negligence accounted for a further 39% of reported breaches. The majority of serious breaches result from failings in people, process and technology.

The majority of threats originate from within an organization. The U.S. Computer Emergency Response Team (Cert) estimates that insiders - whether malicious or merely careless - are responsible for almost 40% of IT security breaches. Security technology such as firewalls, content security appliances or desktop programs can’t entirely compensate for people’s ability to deliberately or innocently bypass the rules.

Meanwhile, changes in workplace habits like mobile working and the use of multiple devices have upped the security ante. Outside the office, employees connect to corporate systems and programs via VPN tunnels or Web-based remote access applications, using corporate, personal or even public computers and devices. With so many access methods, the network perimeter remains porous, leading IT security managers on a constant search for additional protection and monitoring capabilities.

The situation is exacerbated by the rise in employees’ use of their own devices for work, whether authorized or as an under-the-radar aid to productivity. Despite increasing acceptance of BYOD (Bring Your Own Device) practices, there’s a growing gap between what employees actually do and what organizations have accommodated into their security and corporate best practices. Research by Information Law LLC from March 2012 indicates that 31% of companies surveyed had no company policy governing employees’ use of their own devices at work, while a further 26% said they ‘sort of’ did.

The Case for Deeper Software Insight

In addition to securing the network perimeter, corporate desktops and mobile devices, IT departments need to quickly and easily monitor the software that users are installing and accessing, and ensure that only authorized individuals are using programs with access to sensitive information.

To this end, software asset management (SAM) tools add a valuable weapon to the IT security arsenal. SAM helps tackle potential risks from the software usage perspective, helping IT managers detect and halt threats in four major areas:

1. Identifying malicious programs, hacking tools and other unauthorized software
2. Preventing the use of suspect or malicious applications
3. In the event of a security breach, examining application usage data to see who was running suspect applications
4. Identifying and reducing the number of underused software titles so IT can support and patch fewer applications

Acceptable Application Matrix

It is much easier to maintain a robust security posture if acceptable software titles and types are defined and documented from the outset. Maintaining a matrix of tested, validated, approved and documented software helps strengthen policies and support existing technology. Establishing a matrix helps IT set policies preventing workers from using unauthorized software.

Application Control

Despite the most stringent software usage policies, portable storage and mobile communications devices can insert unwelcome software behind the organization’s firewall at any moment. But disabling unacceptable programs can be a powerful weapon against potential security breaches. Application control also helps ensure that only authorized users can gain access to specific programs.

Access Insight

While most applications through which sensitive data can be accessed are protected by authentication controls, SAM solutions add a further layer of security by providing an instant snapshot - at any time - of which employees are accessing which program. The ability to retroactively trace the origins of a breach is an important reporting tool - especially for companies subject to regulatory compliance.

Improved Patch Management

Better SAM tools give IT a streamlined way to identify and eliminate underused or redundant software titles, and to restrict access on a needs-only basis. IT departments no longer have to act as detectives, and they can save time by supporting and patching fewer applications. They can also help ensure that all devices on the network are running the appropriate security software, a huge time saver.

Knowledge is power - and security. Your level of protection is significantly higher when you know exactly what software your organization authorizes, see who is accessing which programs, prevent the use of unacceptable programs and identify any breaches.

In 2010, Cisco Chief Executive John Chambers told reporters at the company's reseller conference in San Francisco, "We don't focus on other companies. We focus on market transitions."

The statement was a half-truth. Chambers should have said companies other than Microsoft.

On Monday, the eve of Microsoft's first Lync User Conference, Rowan Trollope, general manager of Cisco's Collaboration Technology Group, posted a blog that that explained why Lync was inferior to Cisco's platform for unified communications and collaboration.

"I'm quite sure some of it will generate controversy but that's OK - it's a conversation worth having in our opinion," Trollope writes.

But as sometimes happens when brands or political campaigns "go negative," the whole thing is blowing up in Cisco's face, as analysts point out the weaknesses in Cisco's arguments.

The real takeaway, in fact, is that Cisco seems to be scared of what Microsoft is selling.

Cisco's Claims

Trollope's post isn't super nasty, at least not by Apple-v-Android standards. But he takes some shots at Microsoft Lync, calling it "a solution that's primarily been developed for a desktop PC user experience" and thus "less able to meet these wider post-PC requirements than one that has been designed and optimized for them from the outset."

An example of the latter, Trollope says, would be Cisco's UC&C, which is a set of integrated products, such as messaging, Internet telephony, video conferencing and data sharing. All the products are accessed through a single user interface.

Another of Trollope's criticism is that with Microsoft, customers need to go out and buy all sorts of different devices instead of getting everything from a single vendor. "And, in our opinion, that could lead to increased complexity, cost and risk, not to mention the hours spent trying to figure out `who's on first' when troubleshooting an issue."

And finally this:

"There are other important topics that we think should also be discussed. Does your collaboration vendor have any conflict of interest with other BYOD device vendors? Can you move from an in-house deployment to a cloud-based service and get the same functionality? We would encourage you to explore these points with us and any other vendors you are considering."

This is all pretty garden-variety competitive marketing, and certainly far less aggressive than what Microsoft does with its anti-Google "Scroogled" campaigns.

Nevertheless, analysts were quick to cry foul and to point out flaws in Cisco's arguments.

Cisco's Hypocrisy

A large part of what Trollope called a "frank and direct conversation" was a "little far fetched and hypocritical," Gartner analyst Steve Blood says.

Cisco claims Microsoft's Surface tablet represented a conflict of interest, since Lync would also support competing tablets from Apple and Google. Cisco seems to have forgotten its own entry into the tablet market with Cius, which failed miserably and was pulled last year. "It wasn't worried about a conflict of interest then," Blood says.

Cisco also has other conflicts when it comes to hardware. While its UC&C products work on other vendor's systems, they run best on Cisco's Unified Computing System. And when it comes to partners offering Cisco UC&C in the cloud, its UCS server is the only hardware option, Blood says.

Trollope claims Lync is more complex and expensive because customers need to get phones, video equipment, voice and video gateways and networking gear through hardware partners since Microsoft doesn't make those products, while Cisco sells its own integrated hardware and software.

Art Schoeller, analyst for Forrester Research, isn't buying Trollope's argument. "Each account is different in what they have, what they want, and what capabilities are important to them and what model appeals to them more," he says.

While Cisco arguably has a stronger hosted platform than Microsoft, Cisco's biggest resellers are also selling hosted Lync and Office 365, which is "a recognition by Cisco's partners that in some instances, the Microsoft solution is something they would want to propose in place of Cisco," Blood says.

The biggest problem Microsoft has in offering Lync in the cloud is with voice communications. In many countries, as soon as voice hits the cloud, it becomes a regulated service, much like that of a carrier. Microsoft and Cisco are solving the problem by partnering with carriers. "Currently, Microsoft promotes Lync on premise, if a customer wants deeper voice capabilities like conferencing," Schoeller says.

Cisco Feels The Competition

Cisco is going on the offensive because Microsoft is becoming a serious competitor, which is good for companies in the market for unified communications products. However, Cisco would do better to focus on customers, rather than spend time attacking the competition with "ill-prepared, and weak arguments such as this," Blood says.

In a recent interview with AllThingsD, Chambers said, "We love to compete, and we try to always compete with class."

If Chambers believes Trollope's blog is class, then he needs to look up the definition.

With all the recent teeth gnashing about startup investment shifting from consumer to enterprise technology, it's worth noting that successfully cracking the enterprise market is no easy task:

• 70% of the U.S. economy hinges on consumer spending. Even with the pending fiscal cliff, it's kind of hard to ignore the numbers. 
• Enterprise technology is not a short game.

Unlike most consumer technologies, enterprise infrastructure and applications run on a much longer upgrade cycle: 5-7 years. While you might ditch your smartphone every year or two for a newer model, few companies are willing to swap out their CRM systems, storage or security technologies that quickly.

Switching behavior is both the most complicated and important subject in the enterprise technology market. Even if enterprise customers have good reasons to be unhappy with their technology vendors (e.g., lack of innovation, price gouging, poor support), their business runs on that technology. This makes them highly incentivized to see existing vendors address any issues and continue the relationship. As we all know, moving's a bitch.

Of course, enterprise tech is a rich, rewarding game, so it's worth exploring the strategies startups can use to overcome the barriers to switching in the enterprise market:

1. Transformational Technologies. The ultimate startup is the one that changes the game on an incumbent in such a way that the latter neither can block nor retaliate. Classic examples include Virtualization and Software-as-a-Service (SaaS). Because virtualization decouples compute functions from hardware (while running on top of the hardware), it is the ultimate disruptor because it's non-invasive. SaaS eliminates the stickiness of packaged software - and the lucrative support contracts that go along with it. Interestingly, while there tend to be many attackers in Virtualization and SaaS, only a few players tend to win big. Very big: witness VMware and Salesforce.

2. Changing Product Cycles. Catching technology giants in product transition cycles is one of the most effective ways to insert new technologies. However, this usually requires an outside force to speed insertion. Earlier in my career, Intel Centrino drove the need for enterprise Wi-Fi and forced an architectural change. In 2013 you can see many great examples of this idea, including Palo Alto Networks, Splunk, ServiceNow and Workday. These transition cycles don't last forever, though. Over time the incumbents typically build or buy their way into the new product segment and the situation stabilizes until a new cycle begins.

3. Trojan Horses. Sometimes a new enterprise IT category emerges in an indirect way. Cloud infrastructure eliminates the need to buy IT hardware and software; the rental model emerged as form of shadow IT for specific projects that could not wait for corporate IT to respond. It also became the preferred approach for brand new businesses (Netflix streaming). Amazon Web Services and Rackspace, two big early winners in cloud computing, sell computing cycles by the month, payable with with a credit card - often bypassing traditional IT purchasing processes. Once established, Cloud and SaaS vendors can then turn their attention to selling to mainstream IT.

4. New Buying Centers. The multi-hundred billion-dollar enterprise IT game now pivots on competition for the IT "stack," as we shift from the Client-Server/Web mobel to cloud computing. This change has created a new class of IT decision makers such as the "cloud architect." As companies move more to the cloud, this new IT leadership category drives key decisions for enabling new applications, also driving the buying all of the underlying IT components. And these new buyers may not be as wedded to the incumbent suppliers as were the decision makers they supplant.

5. The Consumerization of IT. The iPhone led to a watershed change both in enterprise mobility and computing. Not only did it challenge corporate purchasing patterns ("I buy, you enable," also known as BYOD, or Bring Your Own Device), it eliminated a final barrier to what constituted a business device. This is less about "consumerizing" enterprise IT, but rather, adapting enterprise IT to leverage consumer technologies. In addition to mobile devices, apps are challenging the application market for business software.

6. Coalitions of the Willing. For most small companies, hiring a large enterprise sales force and entering a year-long acquisition cycle is likely to be an expensive exercise in futility. Sure, you might be able to make a living selling to universities, hospitals and niche verticals, but attacking the Fortune 500 requires friends who need another reason to re-engage in a selling conversation. Manufacturing and strategic partnerships with hardware makers made a lot security companies rich during the client-server era (e.g., McAfee, Symantec). Today, companies like Box are changing the game through new kinds of partnership integrations.

Frontal assaults are the hardest attack strategy for an enterprise startup. Attacking a powerful technology company's profit sanctuary tends to piss them off. If you can pull it off, it might just get your company acquired, but run a big risk of perishing in the attempt.

That's why this tends to be the strategy of large companies (e.g., HP's acquisition of 3Com to attack Cisco) and does not have a great track record. The assault on the business PC by iOS and Android tablets and smartphones may turn out be a more successful example, but, Apple and Google and Samsung are hardly startups.

It can be done, of course. Many decades ago, Microsoft's PC operating system was such a technology and for a generation, a small company in Redmond changed the world. (With a big initial boost from IBM, of course.)

Current technologies that might have the power to force enterprises to switch and create hugely successful startups include Apache Hadoop, Network Virtualization, Flash Storage, and Cloud Storage and Collaboration. That's where I'd look for the next big thing.

Image courtesy of Shutterstock.

Software-defined networks (SDNs) are poised to move from market hype to real technology in 2013. International Data Corp. predicts the market size of this disruptive technology will soar from barely existent this year to $3.7 billion by 2016. When a market is expected to grow that much in so little time, big technology players take notice and start shopping for what they won't have time to build for themselves. IDC is marking five startups that will be high on shopping lists: Big Switch Networks, Embrane, Plexxi, Vello Systems and Midokura.

To get an idea of why this market is so hot, let's review SDNs and why they are so potentially disruptive. Software-defined networking places control of network resources within a software layer that sits on top of routers, switches and other physical and virtual network devices. This solves the problem of having to use a control panel for each individual device in order to configure, program or perform other management tasks for the network. Making significant changes to a network today can take one to two weeks using the standard tweak-each-router method, but an SDN holds the promise of reducing the time to a few hours. 

For an enterprise market where server elasticity is becoming a hot commodity, imagine what a enterprise infrastructure could do if the network were elastic, as well.

A key enabler of SDNs is the OpenFlow protocol created at Stanford University and the University of California at Berkeley, a standard now under the control of the Open Networking Foundation. Board members of the foundation include Cisco, Juniper Networks, Brocade, Citrix, Hewlett-Packard, Dell and IBM. Not coincidentally, these are the same companies that IDC believes could go on an SDN shopping spree next year.

Here's a look at each of the company's that IDC believes could be on the short list of some of these vendors:

Big Switch Networks

In November, Mountain View, Calif.-based, Big Switch Networks released its product suite called Open SDN. The suite includes a controller that can sit on top of roughly 1,000 switches to handle programming and set policy-based functions. Other suite components include a network monitoring application and data center network virtualization software for automated network provisioning.

Since March 2011, the company has raised $39 million from investors, including Redpoint Ventures, Index Ventures and Khosla Ventures. Its founders are Guido Appenzeller, who worked on the OpenFlow standard at Stanford, and Kyle Forster, who was the vice president of product management at Joost before starting Big Switch. They founded Big Switch in 2010.

Embrane

Founded in 2009, Santa Clara, Calif.-based, Embrane released its "heleos" software platform a year ago. The product can be used to control a variety of network services, including load balancers, firewalls and virtual private networks. Heleos, which also provides wide-area network optimization, is targeted at cloud environments, whether public, private or hybrid.

Former Cisco executives Dante Malagrino and Marco Di Benedetto founded the company, which has raised $27 million in funding. Investors include venture capital firms Lightspeed Venture Partners, New Enterprise Associates and North Bridge Venture Partners.

Plexxi

Cambridge, MA-based, Plexxi introduced its product strategy this month. The company has built switches that can communicate directly with each other over high-speed fiber optic interconnections. This is meant to replace traditional network architectures that have an access switch communicating with other switches in order to connect two computers, according to MIT Technology Review. The use of access switches can create bottlenecks and unnecessary overhead; something Plexxi is seeking to work around.

Plexxi Chief Executive David Husak founded the company in 2010. Before Plexxi, Husak co-founded Reva Systems and founded C-Port Corp. and Synernetics. Plexxi investors include North Bridge Venture Partners, Matrix Partners and Lightspeed Venture Partners. As of June, the company had raised nearly $48.5 million.

Vello Systems

Founded in 2009, Menlo Park, Calif.-based Vello Systems has been profitable from the start. In each of its first two years, it had roughly $10 million in sales, according to Dow Jones & Company. Vello's first and only round of funding was for $25 million in 2011. The money came from financial institutions that were also Vello customers.

Vello focuses on using SDN technology for internetworking between cloud data centers. Essentially, the company provides cloud-switching software that optimizes latency sensitive applications, such as content delivery, storage replication, big data connections and cloud services. The company first focused on carrier networking and later expanded into corporate data centers and cloud service providers.

Midokura

Midokura is a Japanese startup that entered the U.S. market this year. Company Chief Executive Tatsuya Kato and Chief Technology Officer Dan Mihai founded the company in 2010.

Midokura's flagship product is called MidoNet, which is an intelligent software abstraction layer that manages the internetworking between the hardware infrastructure in an enterprise and the OpenStack cloud-computing platform used in public and private clouds. The company claims its technology reduces complexity, improves fault tolerance, optimizes network traffic and delivers higher availability of servers and services.

Consolidation in the SDN market has already started. In 2012, VMWare bought Nicira for $1.26 billion. In October, Cisco bought SDN startup vCider, which has technology similar to Nicira.

But this is just the beginning. If IDC's predictions are right, then major tech vendors will be moving a lot faster this year to see who can grab the biggest share of this fast-growing market.

Image courtesy of Shutterstock.

Wardrivers take note: You may not have as many Linksys routers to hack if Cisco's rumored sale of its consumer-facing Linksys unit actually happens. Well, you will, but the technology may be in a whole other form.

The name "Linksys" was synonymous with home-network routers throughout the world, as driving with a Wi-Fi device down most streets will demonstrate. I can go out into my backyard right now and find three Linksys networks in range (along with a couple of Netgear routers and a whole bunch of AT&T U-Verse and Comcast routers).

The proliferation of AT&T U-Verse and Comcast routers may be one of the reasons why Bloomberg is reporting that Cisco has hired investment-bank Barclays Plc. to sell the Linksys unit, which Cisco bought with much fanfare for $500 million in 2003.

Cisco is not commenting on this news, so one should take it with the requisite grain of salt. But the move would match Cisco's pattern of dumping its consumer businesses. In April 2011, Cisco announced it would shut down its Flip camera business. The idea was -- and still is -- to get back to Cisco's priorities: "core routing, switching and services; collaboration; architectures; and video."

Presumably Linksys would not fall into core routing, but one would imagine that Cisco might make an exception for its subsidiary if Linksys were bringing in significant money.

When Cisco picked up Linksys for a mere half-billion dollars, it was estimated that router maker had something like 39% of the home-networking market. Orders for Linksys products have since weakened significantly. In 2010, In-Stat said Linksys' revenue amounted to just 9% of total revenue for the market segment.

One reason why Linksys may have stumbled so far could be broadband providers policy of leasing equipment to customers -- equipment that comes as part of the deal for most broadband contracts. While tech-savvy consumers are more likely to save the monthly equipment rental fee to go buy something else on their own, a majority of subscribers won't bother.

Linksys devices were occasionally part of the packages offered by the broadband firms, but not often enough to change its fate.

For these people, the rental cost is transparent; just something that's part of the monthly bill. Plus, setting up a router on your own, when the one that the broadband provider gave you is working just fine seems silly to them. Normally, I could get haughty about security, but when my provider came in to install its box, the installers actually gave me a good lecture about changing the router's default password. (I sent the router back, but just to save the money.)

This will not be the end of Linksys, of course. There is already speculation that the technology will shift to the set-top box sector, and that's a pretty good guess. It's a very sure bet that Cisco won't get nearly the price that it paid for the company nine years ago. But that will have to be the price Cisco pays to for its realignment.

Any company that's using components made by Huawei and ZTE ought to "be very concerned." according to one U.S. cyber-security expert who's familiar with past Chinese efforts to infiltrate private and government IT operations.

No Transparency About Even Basic Details

"They should be very concerned that there is little transparency and [the companies] avoid answering questions about the most basic details, such as who is on the board of directors," said Dmitri Alperovitch, CTO of security firm Crowdstrike, referring to any end users of Huawei or ZTE products.

The two Chinese firms were specifically mentioned by yesterday's report from the U.S. House of Representatives Permanent Select Committee on Intelligence that recommended that businesses and individuals should avoid doing business with ZTE, Huawei and other Chinese firms due to a potential national security threat. (See How U.S.-China Tensions Could Affect Your Next Smartphone.)

Alperovitch highlighted Huawei's diverse product line as one particular example of how far reaching the potential threat could be. Huawei makes and sells smartphones used by various carriers here in the U.S, such as AT&T, MetroPCS and Cricket. Huawei also makes high-grade and high-capacity network switching gear, which is where the bigger risk lies, Alperovitch explained.

Specific risks could include

• Smartphone remote access - though the individual risks are small, this could present pesonal threats to high-value individuals
• Firewall breaches
• Remote access to servers, which could lead to data breaches
• Remote access and re-routing of network switches and routers - with unknown implications
• As yet unknown cyber attacks
• Backdoors to automatically send information back to China - where it would presumably be shared with competitors and/or the Chinese government

Backdoors Are The Big Problem

That kind of equipment, which handles high-volume voice and Internet traffic, would be especially vulnerable to backdoors that would enable malicious organizations to intercept and re-route traffic at will.

This is not exactly a new position for Huawei and ZTE. Alperovitch related an incident a few years ago when Sprint reached out to possibly partner with Huawei on networking infrastructure. "The U.S. government called them up and said 'no.' Their efforts to get into the U.S. market have been stymied for years."

The U.S. is not the only nation concerned. In March, Australia announced that was blocking Huawei from bidding on a national high-speed Internet project because of concerns about cyber-attacks from China, Alperovitch said.

Not The Same As Google And Apple

As news about the House Committee's report percolated through the wires yesterday, many commenters, including me, pointed out the irony of complaining about information "phoning home" to China when U.S.-based Google and Apple have been doing just that for years.

But Alperovitch asserts its not at all the same thing. "Google and Apple are not beholden to any Western government," he indicated. In the past, both companies have made it a point to tell the government to take a hike. But, given its suspected ties to the People's Republic of China government, "do we think Huawei would ever say no?"

So what would it take for ZTE and Huawei to reestablish trust? Alperovitch said the companies should be willing to be forthright and unevasive about the most basic aspects of its infrastructure and governance, such as who is on the boards of directors and when were the companies even founded? Such transparency would not be the only thing needed to completely establish trust.

"But it's a start," Alperovitch concluded.

Lead image courtesy of Shutterstock.

The United States Air Force is one very high-tech organization, and we’re not just talking about jet fighters. The Air Force’s latest mission is a high-stakes, high-speed migration to Internet Protocol v6 (IPv6). Chances are most corporate networks aren’t as extensive or complex as the Air Force’s, but the service’s planning operations offer worthwhile lessons for many organizations.

The Air Force began its transition to IPv6 earlier this summer, and expects to have its entire network migrated by the end of September 2014, the deadline self-imposed by the US government for all of its network operations. The move to IPv6 will also let the Air Force support more ad hoc networks in the field - making it more operationally agile and better able to support machine-to-machine communications.

A Complex Mission

Several years ago the Air Force established a Transition Management Office (TMO) at Scott Air Force Base, located outside of St. Louis, to help coordinate the effort. ReadWriteWeb visited with Doug Fry, Network Engineer, Air Force Network Integration Center and engineering lead for the TMO. His role is to develop network policies and operational procedures that will be carried out by the various Air Force base engineers around the world. Fry is giving a talk at the upcoming New York City Interop this fall.

One of Fry’s biggest issues is maintaining the security of the network as it makes its transition to IPv6. “We can’t let unknown traffic traverse our networks, of course, but the security tools that we have in our inventory aren’t fully v6 compliant yet.”

The Air Force has 130 bases and about 100 of them are IPv6 capable and ready, according to Fry. He is working on the rest right now.

The Air Force base furthest along in the transition process is Eglin in the Florida panhandle, which also happens to be the service’s largest base - covering more than 600 square miles and employing more than 30,000 people. To give you an idea of the size of the base, it has 30,000 individual IP addresses assigned, to a wide mix of both computing and embedded equipment. There are two core networks, 14 access layer devices, and 5000 in-building switches. That is a lot of gear to migrate over to the new networking protocols.

But Eglin’s lead role is more a matter of circumstances than anything else: the base’s aging Cisco routers and switches were due for a major refresh at the same time that the Air Force was planning the IPv6 transition.

8 IPv6 Lessons Learned

So what are some of the lessons the Air Force has learned so far?

1. Don't go with your first address plan, but think about ways that you can make it more hierarchical and improve it. "We are on our fourth iteration of our address plan," said Fry.
2. Make sure your core and IOS routers are all IPv6 compatible and can run dual stack protocols. This seems obvious but it is worth mentioning.
3. Make sure all your monitoring equipment is up to snuff. Eglin uses homegrown IP address assignment and monitoring programs, and of course these will have to be upgrade to handle the longer IPv6 addresses.
4. Now is the time to make sure your entire network documentation actually reflects what is actually deployed. "Some Air Force bases are better documented than others," said Fry.
5. Upgrade your router firmware or replace them to handle IPv6.
6. Build a test lab that replicates your entire network if you can afford to. "I wish we had the budget to build a lab from the beginning, it would have been helpful to learn more about IPv6 before we got down the road," said Lee Tran, a technical advisor for the Operational Infrastructure Branch and part of the Communications Squadron for Eglin. (You can read a ReadWriteWeb a white paper about this topic here.
7. Understand how things will change when you add new desktops or network infrastructure to your IPv6 network. "You don't want to introduce any new vulnerabilities," said Tran. One issue for the Air Force is being able to automatically push out security patches to its routers over an IPv6 network. "Right now we have to do this manually," he said. Another implication is how your desktops will come with support for IPv6, and whether you want this active or not before you actually cut over to IPv6.
8. Finally, participate in the next World IPv6 Day in June and other experiments to prove out your installation and deployment plans."This was incredibly helpful for us, and I was glad to see that our IPv6 servers didn't have any issues then," said Fry.

Good luck with your own IPv6 transition plans.

Lead image and Air Force medallion courtesy of Shutterstock.

Bottom image courtesy of the US Air Force.

In the game of Internet connectivity, it appears that size does indeed matter: size of the Internet pipe, that is. And in this game, the phone companies keep losing.

A new report from the Leichtman Research Group reveals cable operators acquired 330,000 new subscribers in the second quarter of 2012, while the telephone companies posted a net loss of about 70,500 subscribers.

The source of the bleeding was right smack dab within the phone carrier’s DSL service: the once-mighty broadband Internet channel that has seen a lot of subscribers flee in favor of faster fiber- and cable-based services.

AT&T and Verizon, the top two phone companies providing Internet service, did add 669,000 fiber subscribers via their respective U-verse and FiOS services, but that wasn’t enough to compensate for their combined loss of 763,000 DSL subscribers.

No one should be terribly surprised by such small numbers - the second quarter is seldome stellar for Internet providers. College students, for instance, typically drop their subscriptions when they go home for the summer or graduate.

For telephone companies, though, this was only the second time they had a net quarterly loss against the cable Internet providers like Comcast, Time Warner and Cox. In the second quarter of 2010, they dropped back by a mere 1,600 subscribers.

The Rise of Cable Internet

Cable companies increasingly seem to have the upper hand over their phone competitors. “Over the past year, the top broadband providers have added nearly three million subscribers, with cable providers accounting for 83% of the net additions,” reported Bruce Leichtman, president and principal analyst for Leichtman Research Group.

Explaining this trend comes down to three major factors:

1. Size

2. Price

3. Content

Size refers to broadband speed, which is almost always much faster on cable services than DSL. Even when ADSL can match cable Internet for download speeds, cable technology enables higher upload speeds, something that more Net-savvy customers are looking for if they are using more cloud-based services for storage and application use.

Pricing can work in the cable providers' favor, too. Thanks to bundling of services and regional pricing differences, it’s hard to estimate which option is cheaper in a given region, but anecdotally in northern Indiana, the non-discounted price for a local telephone company’s fiber-based Internet service is $63/month for 24Mbps download service, while a local cable provider charges $58.95/month for a comparable 30Mbps service. (Not counting modem rental and other fees.)

Assuming these providers actually deliver these advertised speeds, this breaks down to $2.63 per Mbps for the phone company and $1.97 per Mbps for the cable company. Worse for the phone company, this is the highest download speed it offers. The cable provider has two more tiers of service, up to a 105Mbps download pipe for $199.95/moth that comes in at $0.53 per Mbps.

Cable companies typically have the leverage on content, too. Having already made their deals with the television and movie studios, it’s easy for the cable providers to make a few more arrangements to provide content like the Olympics online for their TV subscribers. Services like FiOS and U-Verse can make similar deals, too, but any phone company that relies on DSL as their broadband option will be at a disadvantage.

Don’t Forget Wireless Broadband

One area where the phone companies might recoup some losses against the cable providers for Internet subscribers may be rural customers. As faster 4G wireless service covers more square miles, there could be a new crop of rural customers taking advantage of broadband via that channel, especially in areas where physically stringing cable or DSL wires is impractical. If wireless broadband is factored into the phone companies' overall Internet subscription numbers, things might not look as dire.

Intel, Cisco and Verizon are pouring $3.3 million into research at five prominent universities to improve video delivery over wireless networks. The first goal of the Video Aware Wireless Networks (VAWN) program? Find a good way to measure mobile video quality.

Understanding Video Quality

As consumers watch ever increasing amounts of video on their mobile devices, network congestion threatens to ruin the experience for mobile users - and give carriers and equipment makers fits trying to accommodate all the traffic. To figure out the best ways to address the problem while still delivering a great picture, though, you need to know - in great detail - what a great picture actually is.  Therefore, VAWN's initial aim is to assess subjective video quality in quantitative terms, according to execs from the three companies who spoke at a roundtable in San Francisco this week.

It turns out that question is a lot more complicated than it sounds, and includes as many perceptual issues as technical concerns. Because viewers perceive quality differently depending on what they're watching - sports versus talking heads, for example - quality isn't about throughput but experience, explained Jeff Foerster, principal engineer and wireless researcher at Intel Labs. That's why VAWN researchers partnered with psychology departments to better understand how the brain comprehends different kinds of video on various devices.

The research is important to finding ways to deliver the best video experience to the most people when the networks get overloaded; that is, minimize the problems that annoy people most. That could mean adjusting the streams' algorithms so the network knows how to deal with particular kinds of content and devices, and understanding the impact of data compression, caching, and storage on video quality. One key is better cooperation between different parts of the network. Video, like other network traffic, is made up of packets of data, but “not all packets need to be treated the same,” Foerster said. “Some packets are more important than others to maximizing perceived video quality.”

Chris Neisinger, Verizon’s executive director of network planning, explained that “right now, we take video in forms that have been created for wired delivery, without care for what the [available network] bandwidth is.” The goal is to figure out “what parts of the video can I change, so that the cognitive perception is still high quality? How can we create a version of the video that’s really highly suited for delivery over wireless?” 

Neisinger said cellular networks already do this on voice calls, making smart tradeoffs to make sure maximum number of users have the best experience. “We don’t have that in the video world.”

To make things more complicated, the measurements must vary by device: “You can’t just take the subjective score for a TV and directly apply it to mobile devices,” Foerster said.

The Bigger Picture

Measuring subjective video quality is only the first step toward the larger goal of delivering high-quality video while boosting network capacity to handle the ever-increasing flood of data. The issue can’t be ignored: In 5 years an estimated 90% of Net traffic will be video, and 66% of mobile traffic will be video. Video traffic is expected to grow 66 times based on the Cisco Visual Networking Index (VNI), but carriers simply can't afford to spend 66 times the cost to boost network capacity. “When you look at a number like 66X, we have to find more efficient ways to do that,” Verizon’s Neisinger said.

Different kinds of video have different requirements and there’s no intrinsic need to treat them all the same, Foerster pointed out. For example, streaming video can use a long buffer - several seconds, say - but that doesn’t work in videoconferencing. Similarly, fast-action sports video needs a higher bit rate, but you may be able to get away with a much lower bit rate for relatively static talking-head newscasts.

Some of VAWN's approaches to solving these problems have to do with the video characteristics - how the stream is compressed, for example. But others are all about network management. “Instead of letting [the streams] all fight to get best possible bandwidth - which is basically what they do today… We need to manage that to increase the satisfaction of the largest number of users instead of some who get really high quality while others get lower quality.” 

In addition to the technological and perceptual aspects, these discussions also have a political element. “It starts getting into the discussion of Net Neutrality,” said Flavio Bonomi, Cisco Fellow and Vice President and Head of the Advanced Architecture and Research Organization at Cisco Systems. “When you treat different content in different ways… It’s a very difficult discussion, but it comes up when you allocated different bandwidth to different streams that might have different importance for users.”

A New Model For Research

The program's third goal is to improve the way companies work with universities. “With the pace of change in the network, we need to provide input to the universities so their research can keep up with the pace of change,” said Verizon’s Neisinger. Basically, to tell them, “here’s how we would use that technology.”

The companies sent out a request for proposals to major universities in 2009, and out of 25 responses chose to work with five: the University of Texas at Austin, Cornell, University of California San Diego, University of Southern California and Moscow State University. The universities devote three to four faculty members to the project, each with four to five graduate students. All results are posted on a public website and the research is not patented. Professors are encouraged to work with each other, and instead of just tossing the stuff over the fence and then forgetting about it, the team holds regular reviews and sharing sessions. “We’re very results-oriented,” Foerster said.

The project is now in year two of its three-year, $1.1 million-per-year plan. By then end of the program, the VAWN team hopes to have a good method to measure perceived video quality, as well as algorithms and software tools to use that measurement to make network decisions. 

The cost of bringing fiber to the home is so high that even the largest ISPs can’t afford large-scale deployments. According to the FCC, a typical deployment costs the provider a minimum of $2,500 per subscriber. Google is apparently approaching the project intelligently, and it will probably find a way to cut deployments costs. But there is a limit to what even the Monster of Mountain View can do. “I don’t believe you can turn the economics of fiber on its head,” says Steve Timmerman, senior vice president of ASSIA, which builds management systems for DSL providers.  

The biggest expense is that of running the fiber itself. Digging trenches in an urban environment is labor-intensive. Google is using existing telephone poles to carry its fiber in Kansas City, but that’s not possible in many cities and suburbs where utilities have long since moved underground. What’s more, cutting deals to use overhead poles and underground cableways could be difficult once the companies that own them come to view Google as a competitor. 

Google, of course, is very large and can use its scale to buy fiber and equipment at excellent prices. But Verizon had the same capability and ultimately lost approximately $800 per subscriber on its fiber deployment, Timmerman says. 

Less tangible, but still significant, is the difficulty of running a consumer business, an area that Google has little experience in. Indeed, when it tried to enter the cell phone business directly with the Nexus One two years ago, it failed miserably. Although carriers like Verizon and AT&T aren’t winning any popularity contests, they do have huge networks of highly experienced technicians and customer service people who know how to keep a network running and how to manage a customer base. 

It’s also worth noting that AT&T’s much-hyped U-verse fiber service isn’t purely fiber; it’s actually a hybrid of fiber to the pole and DSL from the pole to the home - a cheaper, but slower, alternative to 100% fiber deployment, says ASSIA CEO John Cioffi. DSL still serves 70 percent of broadband customers worldwide, Cioffi notes. 

Will the carriers react to Google’s fiber deployment by redoubling their own efforts? That’s still unclear. But unless Google succeeds in proving that large-scale fiber deployment is economical, you’ll have to move to Kansas City to enjoy those cheap, 1GB downloads.

Oracle did not release financial terms of the agreement to buy Xsigo, a maker of virtualization technology that moves data more efficiently in and out of data center servers, but VMware will pay $1.26 billion for Nicira.

The two acquisitions are a reflection of how seriously major tech vendors are taking the emerging network virtualization technology and its potential to shift network intelligence from proprietary hardware from companies like Cisco and Juniper and move it into software running on commodity systems.

Nicira vs. Xsigo

Besides selling into the same market, though, Nicira and Xsigo have little in common. Xsigo sells hardware that provides input/output virtualization that replaces a server’s multiple Ethernet and Fibre Channel interfaces with a single high-speed Ethernet or InfiniBand link.

Nicira provides an intelligent layer of software that sits on top of data-center servers and manages and controls the resources and capacity of networking gear. Nicira’s technology covers a much broader segment of network virtualization, also called software-defined networks, or SDN.

Because Xsigo and Nicira are so different, some analysts say the acquisitions have little in common. “This won’t effect Juniper or Cisco, and it’s not the same type of news as VMWare plus Nicira,” Andre Kindness, tech analyst for Forrester Research, said of Oracle’s purchase.

Others disagree. Michael Genovese, analyst for equity research firm MKM Partners, sees the acquisition as the start of Oracle getting deeper into networking in the data center. “I call it more of a warning shot than a full battle shot across the bow of Cisco’s and Juniper’s businesses,” Genovese said.

Oracle in Network Hardware

While Xsigo does not make Oracle a competitor in that arena today, the company is giving notice that it could head deeper into the network virtualization hardware business. That would happen If Oracle goes after a bigger fish, such as Brocade or Extreme Networks, Genovese said. And at that point, Cisco and Juniper would have much more to worry about.

“Oracle hasn’t done very well in hardware to date, but they want to be a bigger player in the infrastructure (of data centers),” he explained. “This (acquisition) is somewhat of a networking play.”

Oracle became an enterprise hardware vendor in 2010, when it completed the $7.4 billion purchase of server maker Sun Microsystems. Since then, Oracle has bundled its databases and business applications on Sun hardware to provide all-in-one hardware/software packages.

Xsigo is important to Oracle because customers using its server virtualization technology, called Oracle VM, would benefit from having network connections sitting in a virtualized platform.

“The proliferation of virtualized servers in the last few years has made the virtualization of the supporting network connections essential,” John Fowler, Oracle executive vice president of systems, said in a statement. Xsigo is already used today with Oracle databases in the data centers of HiFX, a foreign-exchange broker; Investec, a specialist bank and asset manager; and Bitbrains, a managed service provider.

What About Cisco and Juniper?

With startups and tech giants leveraging network virtualization to close in on their core businesses, Cisco and Juniper can’t afford to stand still. Juniper has launched its own software-defined networking platform called Q-Fabric and Cisco has invested $100 million in SDN startup Insieme. Cisco Chief Executive John Chambers has said the company plans to eventually absorb the startup, which is led by three Cisco engineers.

However, Cisco and Juniper have not embraced the open standards used in other software-defined networking technologies favored by Web businesses and cloud computing companies such as Google, Facebook, Microsoft, Yahoo and Verizon. In fact, Cisco has been trying to slow the adoption of SDNs by confusing customers with claims that its products are compatible with other SDN technologies, said MKM Partners analyst Genovese.

While that strategy may work in the short term, Cisco won’t be able to stop the movement away from switches and other proprietary hardware that has powered networking for years. “They can’t stop the trend, and the trend is towards programmable, open, software-defined networks,” Genovese said. 

Business customers want more efficient networks in which capacity is maximized and they have control instead of vendors of proprietary hardware. Cloud computing has accelerated the need for such networks and old guard networking vendors will need to adapt or watch their businesses erode.

Oracle expects to close the Xsigo deal in the fall.

Does your broadband connection seem a bit faster lately? If so, it’s not a fluke. Many ISPs are delivering speeds that are close to, or even exceeding, the download speeds they’ve advertised, according to a report by the Federal Communications Commission. A few of the big ones, unfortunately, continue to lag.

Most startling was the huge improvement by Cablevision, which now delivers an average of 126 percent of its advertised speed, compared to just 76 percent last year.

“Almost across the board, the July 2012 Report shows that ISPs are doing a better job of delivering what they promise to their customers today than they did a year ago,” the commission concluded. Why the improvement? More investment in network capacity by the ISPs, the FCC says. 

However, that "almost" glosses over significant underperformance from two of the largest Internet service providers, AT&T and Verizon. AT&T improved by four percentage points and now delivers 89 percent of advertised speeds. Verizon, which has talked about abandoning its copper DSL network in favor of fiber, saw its service deteriorate from a robust 115 percent last year to 89 percent this year. 

It’s important to note that tested speeds are averaged over an entire week and generally slow down significantly during peak periods. And because download speed is affected by numerous factors, your own connection may be much slower than average. 

Software-Defined Networking

Nicira’s software acts as an intelligent layer that sits on top of data-center servers and manages and controls the resources and capacity of networking gear. The company’s technology, called software-defined networking (SDN), makes it possible to run data-center networks much more efficiently. That may be good for networking customers, but it means less hardware purchased from vendors like Cisco.

VMware is so confident in Nicira’s technology that it agreed to pay a hefty premium: $1.26 billion in cash and equity, which equity research firm MKM Partners estimates is 40 times the sum of Nicira’s next 12 months of revenue. The 5-year-old company had only recently shipped its first production-ready systems.

Analysts have said Nicira is a perfect complement to VMware’s leading virtualization software that makes it possible to run business applications on multiple operating systems on a single server. While Nicira makes networks more efficient by making better use of capacity, VMware does the same for servers.

“By combining Nicira’s technology with VMware’s market-leading server virtualization and cloud products, VMware today creates a clear path to becoming the most important infrastructure company across servers, networks and storage for the next 10 years,” said Ben Horowitz, a Nicira board member and a partner in Andreessen Horowitz, which was an investor.

A Risk To Cisco

Horowitz may have been giddy from all the money his company just made, but independent analysts also believed the acquisition was a big deal. MKM analyst Michael Genovese said the acquisition is a threat to Cisco, because it takes in roughly $14 billion annually from the sale of high gross-margin switches, which may become less important as technology like Nicira’s gets adopted in data centers.

“The risk to gross margins stems from the idea that the vendor-and-switch-specific software Cisco sells will become less important over time as SDN proliferates,” Genovese said in a commentary on Barron’s. “It is particularly worrisome for Cisco that VMware is the buyer of Nicira.”

Investors agreed with Genovese. Cisco shares fell nearly 6% to a 2012 low the day after the acquisition was announced. It surely didn’t help that on the same day, Cisco reported that it would cut 1,300 jobs, or about 2% of its workforce, in order to reduce costs. The company said the layoffs are needed to maintain growth in the weak economy.

Already Struggling

Even before VMware’s move, Cisco was struggling to maintain growth - especially as the trend toward cloud computing threatens to consolidate demand for its switches.

Last year, the company eliminated 6,500 jobs, or 9% of its workforce, to cut $1 billion from annual costs. The day of the Nicira deal, Cisco’s shares were down 11% for the year.

Of course, Cisco is not the only old guard maker of networking gear threatened by Nicira. The startup also worries Juniper Networks, which has a $500 million a year business in enterprise and data-center switches.

Juniper has tried to defuse upstarts like Nicira by launching its own SDN platform, called Q-Fabric. “However, competitors have done a good job of painting Q-Fabric as a proprietary SDN solution,” Genovese said. Nicira is pushing a movement toward open standards for SDN, such as Openflow and OpenStack.

“We think that Nicira’s ‘openness’ has engendered a lot of customer interest in the startup’s technology,” he said.

Those customers have included AT&T, eBay, Google and Rackspace Hosting. Such deployments have shown that Nicira has technology that works in some of the largest data centers.

Nevertheless, software-defined networking is still in its very early stages. Virtual networking still needs a certain number of physical switches on whih to run, so Cisco remains in the game. But the changes have begun, and Cisco is definitely going to be affected.

Cisco sign image courtesy of Prayitno.

You might think your existing broadband Internet connection is fast enough. It’s not. When it comes to Internet speeds, more is always better. That’s why we all owe some sincere gratitude to the intrepid men and women who bring truly high-speed Internet into our homes.

This post is a message of sincere appreciation - a love letter if you will - to the cable guy who recently upgraded the Internet connection in my San Francisco home. Whether you know it or not, you’ve made my life better in so many ways.

I’ve had broadband access at my home since DSL came to San Francisco in the 1990s. So I didn’t think getting faster service would make all that much of a difference in my life. Boy was I wrong.

My family and I just upgraded our cable Internet service from about 10 Mbps to 50 Mbps. And then we bought a new Wi-Fi router to extend that service to all our wireless devices. Now, 10Mbps isn’t that slow, and 50Mbps is far from the fastest service around (heck, ReadWriteWeb’s headquarters clocks in at an awesome 100Mbps). But I am still stunned at how much the change is affecting how we all use the Internet. And how much I want to hug the Astound cable guy who brought it to us.

Easy Upgrade

Compared to the early days of broadband, the process was amazingly simple. The Astound technician came out to our 115-year-old Victorian with a new Cisco DPC3010 cable modem (actually showing up in the first half hour of the promised 4-hour window!) The tech replaced our old unit and checked out the cabling in less than an hour. Bam, the speed of our hardwired connections instantly quintupled! No fuss, no muss.

Except that the increase was only visible on wired connection, not the fleet of smartphones, tablets, laptops and other devices where we do most of our work (and play). They got a bump to about 20Mbps. Faster, but suddenly pokey next to the wired connections.

Even though we had a relatively recent Belkin router using the modern 802.11n Wi-Fi specification, it simply couldn’t keep up. The tech - remember how much I love him? - recommended getting a new router that supported the DOCSIS (Data Over Cable Service Interface Specification) 3.0 standard. And because we have a mix of Apple, Windows and other devices in the house, my spouse decided to choose simplicity over economy and we splurged for an Apple AirPort Extreme Wi-Fi router.

Although its $179 price is almost double that of competing devices with similar specs, it was by far the easiest router to install and configure that I’ve ever used. Everything was up and running within five minutes, with none of the false starts and geeky questions I’ve encountered setting up other wireless systems over the years. I wouldn’t have chosen it, but I can’t say I missed the headaches.

More importantly, though, suddenly every device in the house was testing out at 50Mbps downloads. 

Faster Everything

The conventional wisdom holds that just about any broadband connection is sufficient for browsing the Web, and that faster connections don’t really provide much benefit in this regard.

Thanks to our cable guy, I can confidently state that conventional wisdom is dead wrong.

Web browsing at 50Mbps is noticeably faster and less annoying than browsing at 10Mbps. In most cases, pages begin loading faster and images show up along with the text, not a second or two later. Downloading large files, from software applications to data sets to high-resolution images is now something we do in real time, rather than a process we start and let run in the background. 

If general browsing got a mild boost from the faster speeds, working with Software as a Service (SaaS) applications delivered over the Internet enjoyed a serious kick in the pants. Gmail and Google Docs suddenly seemed almost as fast as email or productivity software running on a local machine.

Better Streaming

Perhaps the biggest, most noticeable improvement came when consuming streaming content. At 50Mbps, YouTube and other online videos leap into action, instantly jumping ahead with plenty of buffer room. Nice to see on the desktop, but positively intoxicating on an iPad or other tablet, which now seems seamlessly connected to the entire Internet. (I murmur soft words of thanks to the cable guy every time I watch anything on a tablet.)

I now find that I want to have the iPad close at hand at all times, because it’s just so darn easy to watch anything online as soon as I can type it in. Just as important, I’m now wondering why I need a tablet with 64GB of storage space when I can grab stuff from the Net just as quickly. (That makes my new Google Nexus 7 tablet seem more inviting.)

Not surprisingly, that holds true when using streaming media services - whether on a computer, iPad or big-screen TV. Services such as Hulu or HBO Go perform almost as well as our satellite TV service - and our Apple TV box delivers a more TV-like experience than ever before. If it weren’t for live sports, I’d already be considering cutting the cord. (I worry that the cable guy wouldn't like that, though.)

Better Backups and Sharing

All of the members of my household rely on Dropbox to sync and share files, and some of us even pay for extra space. And one of us relies on Apple’s iCloud to sync huge chunks of data among many devices. But syncing all that data to new devices used to take hours, and it churned through much of our 100GB per month data cap. No more. At 50Mbps down and 6Mbps up, those syncs and backups happen much faster. Syncing and backing up to the cloud now seems like a much more reasonable option than it used to.

Upload speeds are often the Achilles' heel of cloud services, but 6Mbps is fast enough to help ease the bottleneck. Still, if the cable guy wants me to buy him chocolates, it would be nice to have upload speeds closer to the downloads.

Bigger Data Limits

When you add up all this stuff, it’s pretty clear that my family is likely to churn through a lot more bandwidth every month - and we were already incurring fees by exceeding our old plan’s limit of 100GB per month. The new plan ups our data transfer limit to 300GB per month, but with the extra speed encouraging all these new uses, we’re actually worried we may blow past that figure as well!

We made the switch because we cycle through a lot of data in our house, and it seemed to make sense. But I think we were all surprised at how much a five-times boost in speed changed the quality and quantity of our Internet usage. I’ve become an instant convert to the idea that the future of the Internet requires that everyone get not just broadband, but really fast broadband.

I hear that Verizon FiOS now offers up to 300Mbps. A week ago, I would have said that’s ridiculous. Now I’m wondering if those speeds will ever be available in San Francisco (if not from FiOS, which apparently won't be built out any more than it already has been, then from another provider). Sorry cable guy, I appreciate how much you’ve done for me, but if the phone company guy shows up with six times faster service, I’m going with him.

Lead image courtesy of Shutterstock.