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

Guest author Ryan Murphey is VP of Facilities and DC Operations for PEER 1 Hosting.

Data center computing demand grew 63% in 2012, requiring enterprises and data center operators to build new facilities to accommodate the market.

Although some data centers may seem to be placed at random, selecting a data center location is a little more strategic than throwing darts at a map. In fact, there are many different factors that affect this decision-making process. The goal is  to ensure that new facilities address local service demand efficiently and sustainably while still making a profit (or meeting corporate needs). Whether the plan is to build a new facility or retrofit an existing building, a number of factors must be considered to ensure that the potential $1 billion investment will deliver strong returns, including type of service, proximity to end users, potential for disaster and climate.

(See also Facebook To Build Huge New Data Center In Iowa - Here's Why)

While all of these factors come into play for data services and hosting providers, many of them are also concerns for enterprises building their own data centers for corporate use.

Is There Enough Demand?

Typically, the first step in hosting companies choosing where to put a data center is talking to customers, prospects and partners to determine where companies are looking for hosting support. This may seem obvious, but for providers to ensure the success of a new data center, they must fill a substantial portion of the facility before the doors even open to guarantee profits during the first month.

If a provider can’t fill enough of a new facility, one way to reduce the financial risk is to use a modular approach. Many companies, including Dell and IBM, use this approach to accommodate growing data center demand quickly, as it allows for the gradual buildout of infrastructure. Additionally, going modular means that providers don’t have to dedicate resources to power and cool unused aisles and racks – they just pay for what’s being used.

Who Needs These Services?

In addition to demand, the location must also match the services to be offered. For example, if a provider is receiving numerous requests for co-located trading equipment on Wall Street, then a co-location facility as close to Wall Street as possible – ideally on the same block – will best serve the demand.

Alternatively, if a provider is predominantly seeing non-latency-sensitive demand across the greater New York area, it can build a hosting facility anywhere nearby. Building a data center on Long Island would cost much less in rent and utilities, and would still be able to meet this market’s hosting needs.

How Likely Are Natural Disasters?

Another element that plays a role in the decision-making process is the number and severity of natural disasters common to the region. For instance, areas prone to tornados, flooding or hurricanes raise a red flag because they could knock out power and damage the facilities. Similarly, operators may shy away from building new data centers along turbulent coastlines and instead look at real estate further inland to avoid water or salt damage.

Hosting providers haven’t always considered volatile weather a determining factor, though. For instance, despite the likelihood of hurricanes and tornados in areas like the Southeast and Midwest, both these areas have a high concentration of data center facilities. But as data center infrastructure and functionality rise in importance, this factor can no longer be ignored – especially since data center outages can cost an average of $5,600 per minute – that’s $336,000 per hour!

What About Free Cooling?

Average temperature is also very important to keep in mind when choosing where to construct a new data center, as it can greatly influence utility costs. Power accounts for an estimaed 50% of data center operation costs, which is why many operators choose temperate environments that won’t add to the heat generated by servers. In a milder climate, operators can also take advantage of “free cooling,” such as open-air cooling, to further cut cooling costs. Large companies like Facebook, Amazon and Apple have been opening data centers to the Northwest region of the U.S. to take advantage of the area’s cool climate and potential for free cooling.

(See also Say Goodbye To The Traditional Data Center.)

Selecting a new data center location follows complex formulas that may not result in the same outcome for every every hosting operator or enterprise.

Every company building a data center will prioritize different goals and concerns. For example, data centers located in Los Angeles face big bills when it comes to climate control, while New York City data centers must grapple with extremely high rent. It’s all a matter of finding an location that offers the greatest number of benefits - without costing a fortune.

Image courtesy of Shutterstock.

Facebook is planning to build a massive data center in Altoona, Iowa, the company said on Tuesday. That's right, Altoona, Iowa, a suburb of Des Moines.

With more than a billion users around the world to support and just three wholly owned data centers (Forest City, North Carolina; Prineville, Oregon; Luleå, Sweden, with the latter two still being built out) Facebook may have needed another location. (The company has also stashed servers in at least two co-location facilities owned by other companies, on both the East and West Coasts.) But why Altoona, Iowa?   

Why Iowa?

According to The Des Moines Register, which deserves credit for breaking the story on Monday, Altoona officials sold Facebook on four key selling points:

1. The site sits on the nexus of an interstate fiber optic system, providing connectivity to the rest of the nation.
2. A power substation sits within half a mile of the campus.
3. Transportation access.
4. Environmental stability.

The last is an increasingly important consideration. Data-center providers that went down during Superstorm Sandy in New York last year learned that lesson well; hurricanes, earthquakes, tornadoes and other natural disasters can bring a cloud services down just as effectively as a power outage. 

A Facebook blog post, meanwhile, cited "an abundance of wind-generated power" as well as proximity to "a great talent pool that will help build and operate the facility" as reasons for building in Altoona. Apparently, Des Moines and Ames are the new Silicon Valley and Boston when it comes to technical skills. The new facility will break ground this summer and begin serving traffic in 2014, Facebook said. According to the Register, Facebook's facility "will join what’s becoming a data center corridor of sorts in Altoona. LightEdge was built in 2006, and Enseva will break ground this spring."

Facebook hasn't confirmed the size of its new data center, but the Register earlier this month claimed that planning documents put it at 1.4 million square feet and said Monday the total investment could hit $1.5 billion. That's about four times the size of the company's Prineville facility - and 50% larger than Apple's $1 billion investment in its new data center in Maiden, North Carolina. 

Apple's data center in Maiden, N.C. (Source: Apple)

"In the coming years, as our service continues to grow and people share and connect in more ways, we need to make sure that our technical infrastructure also continues to scale," Facebook's Jay Parikh said in the blog post. "Our goal is not just to deliver you a fast, reliable experience on Facebook every day – we also want to help make connectivity a universal opportunity. Our data centers are essential for making that happen."

How Facebook "Hacks" Its Data Centers

Facebook has put almost as much technology effort into its data centers as its core services. Earlier this year, Facebook disclosed that its Luleå facility would be entirely built on hardware constructed by no-name server manufacturers using designs developed by the Open Compute Project, which shuns "vanity" hardware sold by traditional server vendors like Dell and Hewlett-Packard in an effort to minimize cost. Rather than pay top dollar for the most sophisticated and powerful equipment, this kind of "open source hardware" approach adds capacity by just adding ever more cheap, generic servers.

(See also Can Servers Save PC Manufacturers? Sadly, No.)

Facebook's Lulea, Sweden data center. (Source: Facebook)

Facebook also has been a pioneer in using natural or ambient cooling its data centers. Traditionally, data centers place servers on raised floors cooled by mechanical "chillers," or air conditioners, that push away heat from the servers to keep them running properly. 

Facebook's Prineville facility uses a combination of evaporated water and ambient air to cool the servers without the need for energy-hogging chillers; its Swedish site uses the frigid near-Arctic air to do the same thing. (Google, meanwhile, is building a data center in Hamina, Finland, which pumps water - and exchanges heat - from a nearby canal.) Although Facebook hasn't disclosed how its Altoona servers will be cooled, it's likely to employ some form of evaporative cooling.

Last week, Facebook was the first to offer a near-real-time look at the Power Usage Effectiveness (PUE) — the all-important batting average of a data center's energy efficiency  — of both its Prineville and Forest City facilities. A few years ago, a PUE of 1.8 was considered average; the Prineville facility's PUE now regularly pushes below 1.10, close to the 1.0 ideal.

Lead image via Facebook.

As part of its Digital Service Efficiency effort to reduce power consumption in its data centers, eBay slightly changed some of its software code so that the code would require less memory. Less memory meant more operations could be performed on the same server over a given amount of time.

The end result? eBay cut its power consumption by about a megawatt and took 400 servers out of its data centers, saving some $2 million in equipment costs.

“The ripple effect was gratifying,” said Dean Nelson, eBay's vice president of Global Foundation Services during a break at the annual conference of The Green Grid. “We just changed the application a bit to save power."

What Is Digital Service Efficiency (DSE)?

Digital Service Efficiency (DSE) is a metric that the auction giant hopes to popularize in the industry. In a nutshell, DSC divides the work accomplished by the power consumed. In eBay’s case, it divides the number of transactions and/or listings by the energy consumed. Energy used in searches is amortized across the entire operations. It is similar to the PUE (Power Use Effectiveness) rating developed by The Green Grid a few years ago, but it arguably is more targeted toward measuring power consumption and actual operations.

eBay’s figures don’t include all the crucial energy data — like how much gas gets consumed shipping a 1973-era Mattel Electronic Football game from Salt City, Mo., to a collector in San Jose, Calif. — but the numbers are still compelling. If anything, they underscore how data centers represent a more efficient way to conduct commerce than driving to the mall:

• eBay conducts 45,914 transactions per kilowatt hour. A medium-sized window-based air conditioner can consume a kilowatt hour in an hour.
• The company generates $337 million per megawatt hour.
• eBay has 52,075 servers that serve 112 million active users.
• It gets $116,716 in revenue per server.
• eBay racked up 7.3 trillion “transactions”, i.e. URL requests to buy or sell something, in 2012. That’s more than 1,000 transactions for every person on the planet.

The numbers come from eBay, so one can take them with a grain of salt, but the overall picture is pretty clear. The company serves a lot of customers fairly efficiently. At a minimum, the company is certainly doing its homework to make it as efficient as possible.

The Dell MDC server racks before the cooling unit is attached.

How Can DSE Help Save Money In The Datacenter?

eBay, for instance, has reduced the number of server configurations it will deploy down to two. Previously, it had 200 to 300 server configurations, with 15 of them accounting for 80% of the total population.) eBay now has a High Performance Computing (HPC) server designed to handle transactions. The HPC servers contain 72GB of memory and 4 hard drives. They are tuned, says Nelson, for rapid processing. A single rack can hold 96 of the servers, brining the total RAM per rack to 6,192GB.

Complementing the HPC servers are eBay’s Big Data servers, of which 48 can fit into a rack. Each Big Data server comes with a dozen 2TB drives. The Big Data servers can fit 1.2 petabytes of storage capacity per rack. The next Big Data servers may come with 3TB or 4TB drives, which would boost the total storage capacity in a rack to between two and three petabytes. Two petabytes can hold the same amount of information contained in all of the academic research libraries in the U.S. (Unlike Google or Facebook, however, eBay is not designing its own equipment in datacenters. Instead, it will buy from computer vendors.)

It has also created a showcase datacenter in Phoenix, dubbed Project Mercury, that utilizes modular containers to isolate equipment and pack it more densely along with liquid cooling. (If you can lower air conditioning bills in Arizona, you can lower them anywhere!)

An HP POD being installed next to another POD

The challenge now lies in balancing efficiency with necessary growth. Storage capacity, in particular, is set to explode.

“It is the storage, stupid,” Nelson said. “The volume of growth of storage is insane. We added 100 petabytes of capacity in the last 12 months.”

In 2013, traditional data centers will begin to lose their dominant status within the data-management food chain. They will increasingly be replaced by big-data software and lower-cost, ARM-based systems-on-chips. 

When thinking about the future of data centers, the problem is one of scale. For the past few decades, relational databases and the attendant hardware that runs them have been able to manage pretty much anything a company could throw at them, but those days are coming to an end.

When Relational Ruled The Land

In the beginning, and for the first 20 years or so, data was heavily transactional, and was managed in discrete and very secure ways. Speed was less important than making sure the data was safe as houses.

In the 1990s, data began to be used in a slightly different way, as comapnies placed analytical demands on the data being gathered. Instead of being retreived in discrete packages, data became as a strategic asset to be analyzed, leading to the disciplines of business intelligence. Databases grew into massive data warehouses, and parallel querying arose as the only way to effectively manage the staggering workloads placed on information technology.

Through the early years of electronic data, growth in the volume of data may have been rapid, but data tools and infrastructure were pretty much able to keep pace.

That's not so true anymore. Software soon will not be able to cope with the overwhelming volume of data being generated, says Mike Hoskins, chief technology officer of Pervasive Software. What's coming is a real break in how data is managed.

Breaking The Old Model

To give an idea of what kind of scale we're talking about, Hoskins points to U.S. retailer Wal-Mart's estimated 1-petabyte data store.

"That's the accumulation of 40 years of Wal-Mart sized business," he said. "Facebook? Facebook generates that much data in a week."

There's always a collection of data behind each transaction. But in e-commerce today, a customer can be clicking around quite a bit before buying, which leads to useful data sets tens, hundreds or thousands of times larger than "so-and-so bought widget X with credit card Y." Add the fact that the machines handling these activities are also recording machine-to-machine transactions, and the data workload explodes beyond the capacity of any traditional data center. 

"We are reaching the end of the useful life" of our data centers, Hoskins said. "The bottom line is, it's a death march."

Even if conventional software could manage this explosion, no company could afford it. Not to mention the energy costs invovled in buying, running and cooling the hardware.

Indeed, it is innovation in hardware that's going to provide the evolutionary break that Big Data requires. Servers with ARM-based processors, which absorb something like 20 times less power than Intel-based processors, are the next wave in data center infrastructure. Lower power requirements, after all, mean less resistance and less heat. Less heat means less money wasted on cooling and the ability to compress ARM-based systems closer together.

As energy and general hardware costs coem down, hardware is lined up to take care of the new data workloads of this new massive scale of data.

First Hardware - Then Software

On the software side, Big Data will increasingly be handled by Hadoop systems that can store data and manage and analyze Facebook-scale loads.

If you're wondering why this is supposed to be big news, think about it this way: Relational databases have been handling data of all shapes and sizes for decades, and now there will be a certain level of data that the traditional data center architecture will simply be unable to handle. It's the first stratification of data management. 

On one level of data management, relational databases will still be around, supporting smaller, less complex and more tactical workloads. But on this new level, whole new architectures will be created to deal with this scale.

What's The Data Center Endgame?

This half-way approach is not the end game, though.

Eventually, Hoskins believes, tools will be built into the Hadoop framework that will enable data managers to run applications and analysis right where the data lives, inside the Hadoop clusters.

It's no accident then that the latest iteration of one of Hadoop's core components - MapReduce 2.0, code-named YARN - includes the beginnings of a framework that will let developers build exactly those kinds of tools inside Hadoop. This is something that the VP of Apache Hadoop Arun Murthy confirmed to me early this year at the Strata Conference in Santa Clara, California. When the YARN application framework is robust enough, Hadoop will be able to let developers code those applications.

This will be the new way of working with data as it gets too big for relational databases to handle: a new architecture of low-cost, low-power servers that will keep applications and data as close to each other as possible, in order to maximize efficiency and speed.

"Relational database technology has had a good run," Hoskins said. But the days of the relational database being a part of every data solution are fading fast, as a new kind of data center becomes the new sheriff in town.

Image courtesy of Shutterstock.

Not surprisingly, IBM execs believe they will soon be selling servers and supercomputers that process data much faster than today's technology allows. But they’re talking about much, much faster.

Rather than use electricity to move data between processors, IBM is preparing to use light, giving high-end customers enough power to keep pace with the growing information flowing through data centers.

IBM's Tiny Milestone

After more than a decade of research, IBM scientists have put nanophotonics technology onto 90 nanometers of silicon and integrate it with ordinary microprocessors.

That last part is critical. If IBM can integrate the new optical technology with ordinary processors, it avoids a layer of manufacturing expense and complexity, and it can make the new systems in conventional factories.

IBM is attacking a bottleneck in data centers and supercomputers. Moving data as electrical impulses - even over the comparatively small distances between chips - chokes off performance.

Solving the problem now means adding more processors, which raises costs.

Practical Nanophotonics

A proof-of-concept just two years ago, IBM’s nanophotonics system is a transceiver that receives electrical impulses from one chip, converts them to staccato light signals, and transmits those to another transceiver sitting alongside another processor. The receiving transceiver reverses the process.

Even with the conversions, the process creates a much higher chip-to-chip data flow - 25 gigabits per second per channel -and can also be used for chip-to-memory communications, another common bottleneck.

Just as with the far bigger fiber-optic systems used for global telecommunications, each beam of light can carry multiple data streams, each on a different wavelength of light. While not a feature on the current proposal, this would make it possible to send terabytes around otherwise conventional electronics with almost literally blinding speed.

"The technology is very universal, very flexible and versatile," Solomon Assefa, a nanophotonics scientist for IBM Research, said.

Short, Sharp Signals In Data Centers

IBM is preparing to put nanophotonics in servers, data centers and supercomputers, Assefa said. The technology is also expected to play an important role in exa-scale computing, a benchmark expected to be reached before the end of the decade. Today's supercomputers are measured in petaflops, equal to 1 quadrillion calculations per second. An exaflop is 1,000 times faster.

There will be millions of transceivers needed for an exa-scale system, he said.

Martin Reynolds, an analyst for Gartner, said IBM's technology appeared flexible, but noted that it's destined to remain in high-end systems for a while.

"The device has to have optical power [i.e. light] fed to it, as there is currently no practical way to generate light on the silicon," Reynolds said. "This demand requires sophisticated and expensive packaging techniques that, for now, constrain these devices to high-end systems."

So it will be some time before nanophotonics make its way into laptops or tablets. In the meantime, IBM hopes to beat competitors by getting it into data centers and supercomputers first.

While some cloud providers build great press about their energy consumption, the truth is data center carbon emissions can actually be much higher than the greenwashing hype would lead you to believe. Most data centers measure energy use, not emissions produced. So one Icelandic start-up cloud provider plans to open source its monitoring software to help cloud providers clear the air about carbon.

Clouding Sustainabiliy

A lot of press has been generated about how cloud computing is a green solution. Intuitively, it certainly feels like increasing energy efficiency should cut energy consumption and carbon dioxide (as CO₂) emissions. That's why cloud services spend a so much time on a stat known as Power Usage Efficiency (PUE), working to achieve a higher utilization of virtual machines on each physical server. Indeed, a recent study from WSP Environment & Energy and the National Resources Defense Council (NRDC) would seem to back up this assertion, as shown in this graph:

Key values for determining the carbon emissions of cloud computing. (Image courtesy of NRDC.)

In practice, though, there are some wrinkles. For one thing, besides PUE and server utilization, the location of the data center most definitely matters and is intertwined with the other data-center-emission variables - an uncomfortable fact often overlooked in the press releases from major cloud providers and their clients, says Tom Raftery, lead analyst, of the GreenMonk energy and sustainability practice within analyst firm RedMonk.

That raises the second and larger problem: For all the green hoopla surrounding the openings of the new high-efficiency data centers that reportedly sip power, there is a distinct lack of transparency into just how much CO₂ emissions these data centers are actually responsible for.

Facebook Is An Open Book

Take, for instance, Facebook's much ballyhooed Prineville, Oregon, data center, for which the company has open-sourced the highly efficient plans as part of its Open Compute Project. While most new data centers have an average PUE value of 1.5, with older data centers carrying PUEs of around 2, the Prineville facility has a PUE range of 1.07 to 1.08 throughout the year.

But according to Raftery, the Prineville facility's utility company, Pacificorp, gets 58% of its energy from coal and 12% from gas - that's more than 70% from fossil fuel directly. And since 22.5% of Pacificorp's power is purchased from other suppliers, there could be even more fossil fuel hiding in the background. Pacificorp burns an estimated 9.6 million tons of coal per year, Raftery said.

And while Prineville is a model of energy efficiency, it is only one part of Facebook's total energy consumption. According to Facebook's own data, in 2011, Prineville used 71 kilowatt-Hours (kWh) of power, just 14% of Facebook's total of 509 kWh across all of their data centers.

It's important to note that Prineville did not open until April of 2011. Using Facebook's carbon-emissions data from the same report, though, we can figure out Facebook's ratio of carbon emissions to its total data-center power, displayed in this table:

Of the four data centers outlined by Facebook, the West Coast center seems to be doing the best, pulling in the most power and only pushing out a third of Facebook's total data-center CO₂ emissions. But even at low PUEs of 1.08, Prineville is just barely improving upon the carbon-emission/power-consumed ratio, and the new North Carolina data center that opened in April of 2012 is responsible for more carbon emissions for power consumed than even the presumably dirty East Coast colocation facility.

Just as Raftery pointed out this weekend in his keynote address at the CloudStack Collaboration Conference in Las Vegas, the source of a data center's power is as important to the data center's total production of emissions as its efficiency ratings. Too often, he added, "energy use" and "emissions produced" have been conflated to mean the same thing. But as outlined here, that's very much not the case.

Clouding The Picture

Facebook, it should be emphasized, is a very good example of transparency when it comes to data-center (and other) power used and amounts of carbon emitted. We are taking the social network's reported findings at face value, of course, but at least it is making those reports. Many companies don't even bother with the details, and still make big claims.

Amazon, for instance, claims that "both the Oregon and GovCloud Regions use 100% carbon-free power," presumably from a completely renewable energy source. Ask Amazon to explain those claims, however, and all you'll get is the sound of crickets, Raftery told his audience, like this non-response from Amazon Web Services senior evangelist Jeff Barr.

Salesforce.com also likes to play the sustainability game a little fast and loose. Raftery lambasted the company's Carbon Calculator, which lets customers select their current location and cloud type to generate a picture of what their annual carbon savings might be if they migrate to Salesforce's cloud. Raftery said the Europe setting (apparently continents are as granular as Salesforce gets) promises up to 86% annual carbon savings.

"Of course that's complete horseshit," Raftery stated. Europe is not homogeneous in its use of fossil fuels for electricity production. While some cloud providers build great press about their energy consumption, the truth is carbon emissions can actually be much higher than the hype would lead you believe.

Clearing The Air

Transparency is key when cloud providers talk about energy and sustainability issues, but Raftery notes that can be hard to achieve even if a company is willing to share its actual data. The problem is that there are few tools and few standards to measure and rate cloud emissions.

In Iceland, a nation that's fast becoming a data-center hot spot thanks to a 100% renewable energy grid and lots of cold seawater for cooling, start-up cloud provider Greenqloud has made some strong efforts in building software that can monitor the the emissions and energy stats for any cloud system.

Raftery was particularly excited about Greenqloud's intention to open-source its monitoring code in the second quarter of 2013. Raftery believes that opening this code and incorporating it into open-cloud systems like CloudStack, OpenStack and Eucalyptus will make monitoring emissions easier for cloud providers around the world.

It will still be up to those providers to choose how much information they want to share. But they'll have one fewer reason not to make their emissions stats more transparent.

Lead image courtesy of the Open Compute Project.

Oracle seems to have a problem with truth in advertising. Since April, the tech giant has had to pull three ads that claimed Oracle computers performed much better than IBM's.

Each time, Oracle offered no proof of its claims and the ads were dropped after IBM complained to the National Advertising Division (NAD) of the Better Business Bureau (BBB). This sleazy behavior, called "strategically stupid" by one analyst, comes at a time when Oracle's hardware business is only limping along.

Oracle's Bad Behavior

The latest ad, pulled in November, said Oracle's Exadata server would run five times faster than IBM's Power Server "or you win $10,000,000." The NAD found that the ad did not provide "any speed performance tests, examples of comparative system speed superiority or any other data to substantiate the message. NAD determined that the advertiser’s decision to permanently discontinue this advertisement was necessary and appropriate," the organization said in a statement.

Oracle did not respond to a request for comment, but the NAD said that the vendor disagreed with the group's findings. Oracle told the organization that it would take its "concerns into account should it disseminate similar advertising in the future."

Oracle's creativity in advertising appears to be more than the result of an overzealous ad exec. In July, the company got slapped by the NAD for saying a giant European retailer moved its database from an IBM Power server to Exadata and found it ran 20 times faster. Three months earlier, Oracle pulled an ad that made unsubstantiated claims that its SPARC SuperCluster T4-4 computer system ran Oracle apps and Java faster than IBM's speediest computer. All the challenged ads ran as full-page spreads in The Wall Street Journal and other publications.

"Three similar rulings in a period of eight months seems like a trend to me," said Jeff Cross, spokesman for IBM's Systems and Technology Group.

Lying Is "Strategically Stupid"

Oracle's decision to sign off on the ads was "strategically stupid," Rob Enderle, president and principal analyst for the <a href= "http://www.enderlegroup.com/"target="_blank">Enderle Group,</a> said in a column for CIO magazine. Enderle argues that Oracle is risking its credibility with large corporations, where IT managers put their jobs on the line in trusting vendors to deliver on major technology deals.

"The difficulty with false advertising is that it speaks to trust and enterprise vendors must be trusted to be successful," Enderle told ReadWrite. "But it really depends on how many customers become aware of this problem. Not many watch the BBB, but in this age of social media, this stuff is circulating and if you don’t trust what you are being told by any vendor, and particularly in the enterprise class, you tend to avoid buying from them."

Oracle's Sales Slumping

Oracle's shameless behavior comes at a time when its hardware sales could be better. Worldwide factory revenue from its servers fell more than 23% in the third quarter, compared to the same period a year ago, according to IDC. The market as a whole dropped 4%. IBM led the pack in terms of revenue.

Oracle's hardware strategy is to sell servers with the company's database software and business applications. The hardware is based on technology acquired through its $7.4 billion purchase of Sun Microsystems in 2010. So far, Oracle's success has been modest.

Sun hardware has not been a favorite of Oracle salespeople. According to text messages released in an unrelated legal dispute with Hewlett-Packard, one Oracle sales official wrote that Sun hardware "baaaallllloooooooows" and another called it a "pig with lipstick."

In the fiscal first quarter ended Aug. 3, Oracle sold $5.7 billion in software, but just $1.35 billion in hardware. Oracle has said that its hardware business is in transition as it drop some Sun systems to focus on high-margin products such as the new Oracle Exadata server.

Meanwhile, sales and earnings are expected to slow. Since 2006, Oracle earnings, excluding some items, has been growing an average of 20% annually, with sales rising 18% on average, data compiled by Bloomberg shows. During the next two fiscal years, ending May 2014, analysts expect per-share profit growth to slow to less than 10% annually and revenue growth to no more than 7%.

As long as Oracle remains silent on the ads, it's impossible to know the company's motives. But the longer Oracle's hardware business fails to take off, the more the sordid behavior looks like desperation.

Oracle plane image by Kunal Mehta / Shutterstock.com

Larry Ellison image by drserg / Shutterstock.com

Emirates Airlines can have up to 90 aircraft in flight at any given time. So what happens if something goes wrong on one of them? If everything works the way it's supposed to, airline personnel may know about it even before the pilots do, via terminals and a massive video display that’s slightly smaller than the screen in a movie theater.

Emirates controls the in-flight safety of its aircraft from inside its operations center, close by the main terminals at the Dubai airport, within the United Arab Emirates. The company opened the facility to reporters last week, as part of a sponsored tour of its latest aircraft, the Airbus A380-800, as well as its Microsoft-powered crew-management application, known as KIS. 

(Disclosure: Emirates flew this reporter to Dubai, United Arab Emirates, to cover the airline's technology as well as the Middle-Eastern launch of Windows 8. It also paid for a hotel room and meals, some of which it split with Microsoft.).

Inside Emirates' Dubai facility, employees have access to a small collection of shops and restaurants, open 24/7, Nearby are the corporate offices, whose lobby is studded with jerseys from teams like London’s Arsenal, which the company sponsors. Security is tight, naturally, with each member of the airline's staff carrying an RF badge with varying levels of access.

By fleet size, Emirates doesn’t rank in the top ten of the world’s airlines, at least according to Wikipedia; its 188 aircraft put it at about half of the 362 aircraft operated by Air Canada, the ninth-largest airline. But by another metric, passenger-kilometers, Emirates ranked fourth in 2011, with 153.2 million. The airline’s hub is in Dubai, and relatively few aircraft fly long distances each day, reaching San Francisco, Tokyo, Buenos Aires, and other far-flung destinations.

The first priority of any airline is safety. But with Emirates striving for top marks in customer satisfaction, minimizing delays is nearly as important. A repair that grounds an aircraft for just 30 minutes may mean the difference between a passenger making or missing a connecting flight, making the center’s operations even more critical.

(Airlines that make up more than 1% of U.S. “deplanements” are required to report on-time flight information to the federal government; because of its relatively few, long-haul flights touching down in the States, Emirates doesn’t fall under this category. A list of its overall complaints by passengers to the Department of Transportation is included in the document, however; simply search for “Emirates.”)

Flight Monitoring - Plus A Big-Screen TV

One might imagine that a network operations center might be constructed with glass and steel, perhaps with network cables strewn about. Instead, the facility’s workstations are embedded in the type of old wood (or at least paneling) that one might find in a reading room, with busts of old composers surveying the room. The workers used traditional mice and monitors, overseeing sophisticated management applications, including AirMAN, for managing Airbus aircraft, and Airplane Health Management, for Boeing airplanes. I saw Windows 7 and Windows XP both being used, decorated with the traditional post-its and tech detritus (“double facepalm” images, for example) taped alongside the monitors.

The room is dominated on either end by sweeping views: to the north, two-story picture windows look out over a pair of Emirates terminals, where aircraft taxi in, park, and then debark. On the other side of the center is a massive video display, slightly smaller than a movie screen. The display shows the local air traffic around Dubai’s airport, followed by a larger, worldwide view of location of each Emirates aircraft on duty. On the day we visited, most of the aircraft clustered around Dubai, either recently departed or waiting to leave. But other flights were en route from Tokyo and other destinations.

On the left, however, Emirates had tucked a video window, tuned that day to CNN. Why? Because maintenance crews and other teams need real-time information. “A perfect example would be Hurricane Sandy,” an Emirates spokeswoman explained. “We would have been touch with the [New York] Port Authority, monitoring the impact from that... whoever has the latest update, we can stay on top of that.”

About 100 employees man the operations center, working 12-hour shifts, 6:00 to 6:00. “We have all the stakeholders... ensuring that the aircraft sticks to the departure that is on the schedule,” said the manager of maintenance control from Emirates’ engineering department. (Emirates allowed the manager to speak with reporters, but did not allow him to be quoted by name.)

“So we have the controlling vice president, who oversees all the aspects of departures and things like that, the maintenance control center, preventative maintenance, health monitoring, scheduling to make sure that the aircraft is not unduly delayed, then hardware, flight dispatch, crew scheduling and Dubai hub control,” the manager said.

If there’s a problem, everyone’s in the room: “It makes it easy to call out, ‘Cabin crew scheduling!’” when that comes ups, said the Emirates spokeswoman.

Maintenance Owns The Planes

Each morning, the manager of maintenance control reviews the last 24 hours, reviewing any delays, discussing service issues and planning the day ahead. Representatives from Airbus and Boeing attend, as well as staff from engine makers Rolls-Royce and other component manufacturers. Even aviation systems manufacturers like Honeywell have offices in the building, the manager said.

According to Emirates, the company owns 188 aircraft in all, with an additional 212 on order. Some 27 of its existing planes are the Airbus A380-800 superjumbos Emirates uses on its New York (JFK) route, while another 80 are the Boeing 777-300 that Emirates uses to service San Francisco.

As each plane takes off, supervision of the plane moves to the maintenance staff. Over the United Arab Emirates, crews can communicate with the ground via radio. Farther out, satellite phones or text messages are used, via a global communications partner that Emirates declined to name.

Whatever the medium, the maintenance staff is on the lookout for are error messages, specifically automated failure messages that can indicate trouble. Emirates can also “ping” newer aircraft, such as the 777, for real-time updates that allow engineers to communicate directly with the aircraft. If there is a failure warning, those messages flash on to the screens of the Emirates engineers, who then can alert the crew, if necessary. There’s even a simulation showing the actual cockpit of aircraft like the A380, with each instrument in its proper place.

At that point, maintenance must make a series of decisions: does the aircraft need to be repaired? Does it need spare parts? Are those spare parts available, or do they need to be shipped in from London, for example? If repairs do need to be made, Emirates can ensure that the necessary manuals and documentation are electronically transmitted, so that the repair crews have the information before the plane reaches the ground. In critical situations, Emirates even has what it calls a “tiger team” of engineers, ready to deploy to any spot on the globe.

Emirates can also decide to allow the aircraft to fly even with “deficiencies” - a scenario where performance is degraded in a non-critical area. “That means that this aircraft will not be compromised on safety because of this situation,” the maintenance manager said.

I was pretty impressed by Emirates’ ICE system, a sort of “Netflix in the sky” that offers 10,000 movies and TV shows to its passengers while in the air (check ReadWrite on Wednesday for more on that). But the maintenance manager made it clear where its priorities lay: “The most important point of having this technology is to fly passengers in a safe manner at any given point in time,” he said.

A new industry consortium launched Thursday is putting Intel on notice: Your hold on processors for servers is about to be challenged by ARM, the processor family that has exploded on tablets and smartphones in the past few years. The Linaro Enterprise Group bands together some big newer players in ARM space, including AMD, Facebook and HP, working together to try to unify the fragmented ARM processors enough to prep them for a new breed of servers that should drastically reduce the size and energy consumption of current servers and datacenters. 

To understand the significance of this news, here's the nickel tour of processors and semiconductors, the hardware in a computing device that actually does the computing.

Intel vs. ARM

For "traditional" computing devices like laptops, desktop computers, and servers, there's Intel, AMD, and well, that's pretty much it. Intel has not only developed the x86 architecture that has been powering PCs and other devices for 30-odd years, it also builds the chips itself, licensing the specs to only one other company: its competitor AMD.

Intel and AMD processors, since they share many of the same features and architecture, have a lot of things in common. For one thing, they are very powerful and, since they have been around for such a long time, plenty of operating systems can run on machines with these chips, including Windows, Mac OS X and Linux.

Contrast that with ARM processors, which are not made by just one company like Intel. Instead, ARM develops only the core architecture and licenses it to any hardware maker that wants to use the design to make its own processors. Because of the way it is designed, ARM processors use very little power, which makes them perfect for personal electronic devices where battery resources must be frugally maintained.

But all of these vendors licensing ARM and building their own version of the processor has led to a serious fragmentation problem. An ARM processor or system-on-a-chip (SoC), a piece of hardware that essentially miniaturizes all the components to make an SoC a stand-alone computing device, will carry the fingerprints and design used by each ARM vendor. That means software that runs on one ARM SoC may not run on another ARM SoC.

Why was this done? Apparently, it was the old scratch-the-itch problem, according to George Grey, CEO of Linaro, a non-profit group working to build standards into the ARM ecosystem - and the parent organization of the Linaro Enterprise Group.

"Engineers will differentiate if they can," Grey told me in a recent conversation. And so, true to their natures, every ARM vendor has a different spin on how its ARM chip was built.

In the past, as messy as this was, the model worked. Manufacturers that used ARM hardware would work with that particular SoC and ensure that an operating system like Linux, Android or iOS could be installed on it, Grey explained. Once the common operating system is in place, the problem of software compatibility was reduced: applications like Angry Birds and FaceTime talk to the operating system, not directly to the processor.

The Appeal Of ARM In The Cloud

Now that the cloud is becoming perhaps the biggest force in enterprise IT, a lot of people are noticing that maintaining racks and racks of servers can get expensive. Intel processors, for one thing, get hot. If they get too hot, things start melting, so you need to use even more energy to keep all those servers cool. And that means you can't make these servers too compact, or you can't fli ub air or water to pull the heat away.

The average Intel server processor pulls in something like 80 watts of power just to run, Grey explained. But a multi-core ARM SoC draws only about 4 watts of power. That's a whole system on a chip, too, not just the processor. Less power means less resistance and less heat. Less heat means less money wasted on cooling and more capability to compress ARM-based systems together.

That's why ARM is getting a lot more attention from hardware vendors, software vendors and really big cloud computing users. Today's announcement of the Linaro Enterprise Group features exactly that kind of line up: AMD, Applied Micro Circuits Corporation, Calxeda, Canonical, Cavium, Facebook, HP, Marvell and Red Hat.

For AMD, this is the second half of a double-punch. On Monday the semiconductor company announced it would release ARM-based 64-bit chips for servers in 2014. Today's inclusion within the Linaro working group solidifies its intent to do this right.

More Than Just Paper Shuffling

Because doing this right is what AMD and the rest of the members of the new group plan to do: Instead of building servers in a system where every different ARM chip or SoC maker has its own unique architecture and muddling through it like the phone and personal device makers, the intent is to start the game right and create common architectural features that will let server vendors using ARM chips more readily install software.

This is not an attempt to unify ARM processors into one single standard like Intel, Grey emphasized. There will still be diversity between ARM systems. But common features like boot architectures, which enable software to work on processors, will be standardized across various chip designs, so software can be more universally installed.

The establishment of a working group like this may just seem like a lot of paper shuffling, but the impact is potentially huge. Grey outlined a new kind of server environment where entire boxes of ARM-based SoCs sit in server rack shelves, instead of one Intel server per rack shelf, delivering the same or more computing power as an Intel-based system in less space, and for far less power and operational expense. The actual servers could be quite small, the size of a smartphone, or even smaller.

Intel will no doubt rise to meet the challenge, but ARM's success in the mobile sector has made the low-power processor a very appealing alternative in the enterprise space.

Image courtesy of Shutterstock.

A business' data is still its most important asset, but that doesn't mean it every company needs to spend millions of dollars building and maintaining its own datacenters. The cloud has grown up, and the days of the company-owned-and-operated datacenters are numbered.

The Basics

In the pre-PC days, computers were sprawling and temperamental beasts requiring constant care and feeding. Mainframes lived in climate-controlled, secure facilities with round-the-clock surveillance and redundant power and bandwidth connections. Over time, most mainframes were replaced by more modular, inexpensive servers, but the operational requirements remained: a cold, secure room, redundancy and 24/7 management.

During the prosperous '90s, IT managers with datacenter blueprints and wads of cash set out to build their own data fortresses. Data was a critical asset that could not be left to chance, so businesses wanted to keep it safe and close. The number of in-house datacenters exploded, and to IT managers of the 1990s or early 2000s, datacenter vendors like Avocent, Veracode, and EMC were every bit as familiar as Microsoft or Oracle.

Until the past five years or so, outsourced data management was generally thought of as a "little guy" option, conjuring all-to-accurate visions of system outages, sloppy security and spiky performance. But datacenter complexity has risen and the cloud has grown up, so more and more businesses - even really, really big ones - are getting out of the in-house datacenter game and leaving it to the pros.

The Problem

Datacenter management has always been tricky, for two related reasons: servers are hot, and all technology eventually fails. All other datacenter issues (cooling, power consumption, physical space management, redundancy, and monitoring) stem from these. If anything, progress has made things worse. Every year, server manufacturers pack more processing into smaller footprints, and datacenters pack more servers into each square foot. Higher processing density has tripled heat levels in recent years.

Headaches

Balancing heat density and floor temperatures is not a game for the faint of heart. 60% of datacenter costs are electrical and mechanical, and that adds up to real money. Upping your floor temperature just one degree can save 4% in power costs, and companies have reclaimed millions of dollars per year simply by tweaking cooling. But those gains aren't free. Raising temperatures just a few degrees can reduce your time to respond to outages and increase the risk of catastrophic failure. That's a lot of responsibility to place in the hands of an IT department that's probably better at (and more interested in) software and networks than HVAC.

Even a datacenter operating at peak efficiency might cost more than its worth. Few businesses have the kind of predictable, linear growth best suited to a datacenter construction plan. Mergers, opportunities, new regulations or unexpected setbacks can all influence the scope of data under management, and that can change on a dime. Not enough capacity is a logjam. Too much capacity is wasted money. The standard logic regarding outsourcing goes like this: "If it's core to your business keep it in-house." But if you're an information company, are refrigeration and construction really core to your business?

Quality

Once upon a time, managing your own data center was the only way to ensure adequate Quality of Service and data security. Businesses assumed the painful realities of in-house management to keep their customers happy and their data safe. These days, that claim is pretty tough for most businesses to make. Netflix, which already runs its consumer business in Amazon's cloud, plans to move more than 95% of its internal servers to the cloud, dropping from 2,500 in-house virtual servers to about 50. The quality is there, and unless you're NORAD, so is the security. In fact, the US federal government, will close more than 1,000 datacenters by 2015 through consolidation and migration to the cloud.

Larry Tabb, founder of the research firm TABB group, summed it up nicely in an interview with Advanced Trading: "Just because it's a shared datacenter, that doesn't mean it's any less safe than your own datacenter. If Citi can get hacked, and the big banks can get hacked, and Sony can get hacked, it can happen to anybody."

The Prognosis

Cloud vendors have advantages due to economies of scale and domain expertise, and they've reached parity or better on security and QoS. In the long run, in-house datacenters in all but the rarest circumstances are done. As Axcient's CEO Justin Moore pointed out, though, migration to the cloud is a gradual process. More entrenched companies with massive investments in datacenters won't move overnight, but newer businesses that have not made major investments will be hard-pressed to build any infrastructure. Over five to ten years, as aging hardware and support systems in established datacenters come due for replacement, most larger businesses will begin to consolidate in the same manner as the federal government – pulling key data close, migrating non-critical data to the cloud, and eliminating excess capacity.

Can This Technology Be Saved?

The datacenter itself isn't going away – just the paradigm of running it internally. As storage technology continues to improve, datacenter management will become increasingly prohibitive, and frankly, fewer and few CIOs will want the headache. It's possible that a few high-profile hacks of cloud service providers might slow the tide for a while, but ultimately, outsourcing management just makes sense.

Previous Technology Deathwatches

Point-and-Shoot Cameras: No change

Video Game Consoles: The utility of bundles apps like Netflix and Vudu seems to be slipping. An NPD Study showed that one in five consumers who view streaming video on their TVs do so without a peripheral device.

Blu-Ray: The same NPD study reveals that "online video is maturing” as users migrate to watching streaming media on their TVs.

QR Codes: It's been a mixed bag. While Bank of America is testing QR codes for mobile payments (good news for the technology), a security researcher demonstrated how a malicious QR code could be used to wipe a Samsung smartphone.

Company Deathwatches

For an update on our baker's dozen of company Deathwatches, check out our updated ReadWriteWeb DeathWatch Update: The Unlucky 13.

Map image from Whitehouse.gov.

High-tech companies like to portray their industry as the greenest of industries - saving everything from paper to business travel. But tech companies continue to be major sources of pollution and waste.

These facts have to the fore once again in a yearlong New York Times investigation that found huge amounts of wasted electricity in powering warehouse-size data centers driving the explosion in digital information. Add these findings to the industry’s failure to control the spread of the electronic waste it creates, and it’s easy to question the validity of its commitment to protecting the environment.

What The Times Found

In looking behind the industry’s green mask, The New York Times found that data centers can waste 90% or more of the electricity they draw from the grid. That’s a staggering amount when you consider the world’s digital warehouses use 30 billion watts of electricity, equal to 30 nuclear power plants, according to The Times. Because burning fossil fuels produces a lot of today’s electricity, the tech industry’s contribution to global warming and air pollution is significant.

The Times article has attracted critics, including Dan Woods, an author and contributor to Forbes and CITO Research. “The New York Times failed in its mission to accurately explain the important issue of improving efficient use of power in data centers, and instead wrote a confused and incomplete article that is unworthy of its reputation,” Woods wrote for Forbes. Michael Gartenberg, a Gartner analyst, tweeted “Wow. So much assertion with so little facts. Wow.”

In general, critics said The Times ignored the high efficiency rates of data centers run by Internet companies such as Amazon, Facebook and Google. In addition, the newspaper was criticized for not giving enough space to efforts underway to improve data center efficiency.

But I found the article did a credible job laying out the problem of wasteful energy consumption by the information industry. And the piece is the first in a series, so critics may get the details they want in future articles.

(Is Microsoft the worst offender? See Brian Proffitt’s Microsoft’s Energy-Wasting Strong-Arm Tactics Show That Redmond Hasn’t Changed.)

The Green ‘Crisis’

In the meantime, environmental protection groups that have been battling the tech industry for years remain fierce critics, despite acknowledging some progress. A major complaint is the failure of consumer electronics companies to design products that are easier to recycle so hazardous electronic waste doesn’t end up in dumps in China, India and Africa.

“Your average consumer electronics are, unfortunately, still in crisis, in terms of really having an outlet for responsible and sustainable recycling,” said Sheila Davis, executive director of the Silicon Valley Toxics Coalition.

The problem is in product design, Davis said. Many consumer electronics are too difficult to tear apart to extract and either recycle or properly dispose of the toxic chemicals they contain, including lead, mercury and cadmium. As a result, even though more than two-dozen states have mandatory laws governing the handling of e-waste, recyclers find a lot of the products too expensive to dismantle. So they simply ship the junked electronics to other countries, where the gear is salvaged for components and the rest tossed in dumps. Often, the toxic chemicals contaminate the ground and eventually end up in water supplies.

“That’s a pretty horrendous outcome for these high-tech products,” Davis warned.

Apple’s Nearly Non-Recyclable MacBook

Apple offers a good example of what Davis is talking about. In July, environmentalist denounced the company’s decision to drop 39 products from certification by the Electronic Product Environment Assessment Tool. The system measures the environmental friendliness of computers, monitors and laptops.

Essentially, Apple believed its products were green enough and didn’t need EPEAT approval. Critics said it wanted to continue with its hard-to-recycle design that reached an apex with the MacBook Pro with Retina Display. In a teardown analysis of the new system, Kyle Wiens of iFixit wrote in Wired that it was the “least repairable laptop we’ve ever taken apart.” Its construction, along with that of the iPad, was also a nightmare for recyclers. (Not to mention tinkerers.)

“The design pattern has serious consequences not only for consumers and the environment, but also for the tech industry as a whole,” Wiens added.

Apple reversed its decision after organizations that required EPEAT certification said they would either drop Apple or review upcoming purchases.

Despite the change of heart, the incident raised questions about the tech industry’s commitment to the environment when it might affect profits. The Electronics TakeBack Coalition released a report in August that found many makers of ultrabooks were heading in the same direction as Apple by not using removable batteries.

Whether it is consumer electronics or data centers, the tech industry is going to find it increasingly difficult to defend the green image it has so carefully crafted. Let’s hope it fights back with solutions and not just “greenwashing” PR.

Lead images courtesy of Shutterstock.

What can you do with a ubiquitous metropolitan gigabit Ethernet connection? Google has recently gotten lots of attention with the metro fiber network that it is beginning to build in Kansas City. Welcome to Chattanooga, Tenn. The city has laid its fiber network just about everywhere, and is beginning to reap the rewards of ultra-fast Internet service. What lessons can Google and others learn from the experience?

Chattanooga's gigabit fiber network wasn't installed in the name of civic progress, or as a calling card to attract IT-related entrepreneurs, or to improve city services or to encourage telecommuting - all things that are happening as a result of the network.

Instead, it began as a project from the municipal electric utility, EPB, to improve power delivery to its customers. Chattanooga suffers many violent storms that can knock out its power grid for hours or days. The utility wanted to increase the reliability of its operations through having a smarter grid that could minimize these outages.

As part of the effort, EPB automated 1,200 power switches and added technology capable of anticipating potential transformer overloads by measuring power flows every 15 minutes using the fiber network. This smarter grid has cut the number of power outages by more than 40%. The utility says it has also saved money.

But the same infrastructure that provides the control network for the utility can also be used to deliver Internet connectivity, and once the fiber network was in place, the utility became a fast Internet service provider.

Take Me To The (Digital) River

Chattanooga Mayor Ron Littlefield says, "Think of what we did as putting in place a digital equivalent of the Tennessee River." That's an apt analogy for the city.

Chattanooga has always leveraged the Tennessee to its advantage. Back in the early 1900s, for example, it used its place on the Tennessee River to attract the first bottling plant for Coca-Cola as well as smokestack industries. The fiber network is just a different kind of river.

The utility's smart-grid efforts have made the area more of an employment magnet and given it new ways to attract talent. The city's IT department, for example, has filled its past 10 jobs with out-of-towners, a post-fiber development. Major employers are encouraging telecommuting.

"We now have a very balanced economy between industrial and clean jobs," said Littlefield. "We have something no one else in North America has, and something that will sustain our future development."

New Uses for Fast Internet

The city has continued to build on its gigabit fiber network. For example, it put together a series of initiatives to monitor and control downtown areas. At one downtown park, the police can adjust the lighting to discourage flash mobs from gathering, as well as scan license plates on cars that are parked in the lot. This helps increase the perception of safety, not to mention discourage potential criminals. "People now know not to park in the park if they have a stolen car," says the mayor.

Speaking of street lighting, city engineers are in the process of replacing the 28,000 traditional halogen lights with LED lights and sensors that adjust their output based on ambient light. And traffic signals can be controlled by the police or first responders to move emergency vehicles through the city.

In the works is the installation of more than 400 different wireless road sensors. In the past, the city needed to send out construction crews to dig up the road and install the common wire loops that are seen across cities around the world. The newer battery powered sensors are the size of hockey pucks and take just minutes to bury.

All told, the city has built more than 50 apps to use the fiber connections, and more are on the way. "Fiber makes bring-your-own-device strategies possible," says Mark Keil, the city's CIO. "We will have three times more devices on our network next year than before we had the fiber, and we have made it easier to monitor and manage them, too."

Gigabit Takeaways

Here are five lessons to be learned from the gigabit experience of Chattanooga:

• Don't build a fiber network just for Internet connections. What made Chattanooga's gigabit fiber network work was the backing of its electric utility. Once this physical plant was in place, the utility was able to offer gigabit service for $350 a month to residential and business customers.
• Symmetrical service is key. Having both a gigabit up and download speed is important for a variety of applications that rely on user-generated content to receive the same benefit as downloaded Web pages. A local group of radiologists built their own app so that doctors could view digitized scans whenever and wherever. That wouldn't have been possible without a symmetrical network.
• Focus on both big and small employers. The region was able to attract a new Volkswagen auto assembly plant and an Amazon.com distribution warehouse, but these success stories were matched with smaller firms. The mayor is effusive in his support for the various entrepreneurial efforts around the region in bringing in smart, tech-savvy people. City CIO Keil mentions that the city asked for some programming help from several Google developers from Atlanta. By the time the project was finished, at least one of them was packing up to move to Chattanooga because of the gigabit network. And this summer several private companies put together the city's first Demo Day to feature eight tech companies who agreed to move to the city in exchange for a chance to win a $100,000 grant. One of them moved from Ireland to participate in the program. Banyan, the ultimate winner, provides integrated productivity tools.
• Find or create a university-based commercialization partner. Chattanooga was fortunate in having a branch of the University of Tennessee, and was able to establish a supercomputing center and a non-profit commercialization entity to help license the technologies developed by academia. Several of their apps are being used in disaster management and large-scale urban planning simulations, for example.
• Finally, don't rule out many unexpected benefits. "We got into robotics and energy development when they were popular many years ago. But our fiber network is like having the first city that discovered fire," says Littlefield. The city is just beginning to see lots of new apps on its network and is still discovering new uses for the universal connectivity. 

Iceland's nascent data center industry, powered by cheap, renewable energy, could be a growth driver in that nation's economic recovery and the cloud computing revolution. It could also save your company money.

If you're looking to host data between Europe and the U.S., it's worth taking a look at your options in Iceland.

Because of Iceland's unique climate, cooling costs are drastically lower than they are in places like New York and London. And the low cost of electricity helps push pricing down further.

Don't miss: Iceland's Cheap, Green Cloud: How a Tiny Island Could Drive Big Changes

Data Centers

The two big data centers right now are Verne Global and Thor. Both offer a range of services, including co-location. Verne is closer to the main Keflavik airport, while Thor is closer to the capital city of Reykjavik. (They're all pretty close to each other, though.)

Hosting, Cloud & IT Service Providers

Datapipe, Colt, and Opin Kerfi have announced a presence at Verne's facility. GreenQloud, a Reykjavik-based startup, is offering an suite of cloud services similar to Amazon Web Services, hosted in both Verne and Thor's facilities. Others may follow.

What kinds of services are a good bet for hosting in Iceland? The bandwidth situation is much-improved over a few years ago, but if latency is your top concern, Iceland may not be your best option. But Verne Global CTO Tate Cantrell - biased, of course! - estimates that companies could host 75% or more of their services in a facility like his, ranging from ERP and data analytics to backups.

What about taxes? Once upon a time, Iceland's nasty 25% value-added tax on servers was a big roadblock, but it's gone now.

If You Visit, Don't Miss...

One of the fringe benefits of hosting in Iceland is that you get to go there to check out your data center. Reykjavik is a neat place to visit for a couple of days, and Iceland's countryside is beautiful. (Also, everyone speaks near-perfect English.)

I recommend the Icelandair Marina hotel right next to the waterfront in downtown Reykjavik. It's brand-new and super cool. For dinner, check out Grillmarkadurinn or Tapashusid; for breakfast or lunch, Aldin; and for coffee, Kaffismidja Islands.

If you're traveling with an unlocked phone, you can get a prepaid, data-only Vodafone SIM card with 5GB of data for about $15. (Check out the mobile phone shops at the main shopping mall in Reykjavik.) If you want to rent a mi-fi device with service in Iceland, check out Tep Wireless.

And be prepared for sunlight all night long in summer and darkness almost all day in winter!

Image courtesy of Shutterstock.

Iceland. A fascinating, strange place. And misunderstood. Did you know that the famous Blue Lagoon is actually a man-made pool filled with runoff from the power plant next door? Or that Iceland could become the world's next data center capital?

There's something magical about a 2 a.m. stroll down Reykjavík's party street watching burly dudes brawl under blue skies. But I'm not in Iceland to gawk, booze or hunt for trolls. I'm here because these guys really screwed up their economy a few years ago, and now they're trying to fix it. And if all goes according to plan, the Internet is going to play a role.

Reykjavík isn't a tech hub yet, but a lot of people are hoping that data centers can drive a nice chunk of its future growth. Moreover, if the promise of Iceland's inexpensive, sustainable data center industry comes to life, it could accelerate the migration of all kinds of businesses into the cloud. That could be good news for thousands of Web companies and billions of users around the world.

Iceland's Big Rebound

Iceland torpedoed its economy by playing way over its head in the global banking system. Legendary shenanigans, well worth reading about. Absurd deals, big cars, birthday-party-appearances-by-Elton-John kind of stuff.

Now, as Iceland rebuilds and stabilizes, part of its growth plan hinges on cooperation between Mother Nature and the Internet: low-cost, green data centers that are popping up outside Reykjavík, Iceland's capital.

Iceland, where the dominant industries are fishing and aluminum smelting, has a few advantages that could eventually make it an important player in the massive and growing data center industry. 

It's Cheap Because It's Green

The biggest costs of running data centers are electricity and electricity. Running the servers and keeping them from overheating consumes a lot of power. 

Lucky for Iceland, it has tons of electricity: inexpensive, practically 100%-green power generated by hydroelectric and geothermal plants around the country. Iceland generates more renewable energy per capita than any other country in Europe by far, the power company boasts.

As for cooling, well, it's almost as simple as leaving the window open.

"This is a center-of-the-fairway, right-down-the-middle, perfect location for a data center," says Tate Cantrell, chief technology officer for Verne Global, one of Iceland's two main data center companies. (The competing Thor data center is across town.) "We wanted to anchor our company in a place where it just made 100% sense." 

Verne got the ball rolling in Iceland several years ago, and earlier this year, its first data center opened for business in an old NATO base, not far from Iceland's Keflavík International Airport.

Welcome to the Rock

Inside, it looks pretty much like any other data center: clean floors, rows of server cages, neatly organized cables, LCD status monitors, gaseous fire suppression systems and the like - including laser air sensors that can "sniff" a server's defective power supply days before it starts smoking.

The first working server room itself, made by a U.K. company called Colt, was constructed elsewhere and shipped to Iceland in container-sized "Lego blocks." Verne can order them on-demand for relatively fast expansion.

But what makes Verne's facility special - like other Icelandic data centers, now and in the future - is the way it's run. 

Thanks to all those hydroelectric and geothermal plants, the power is 100% renewable - no nuclear or coal. (There's a diesel generator out back for emergencies.) This is nice for the planet, sure, and it may help some clients scratch the green itch. But it also makes running a data center super-affordable.

Iceland doesn't have an easy way to ship its abundant electricity off the island, so it's inexpensive if you can do something with it right there. Historically, that's been aluminum smelting, which consumes some 70% of Iceland's power. But most everyone is hoping that data centers take off. They'll pay higher electricity rates, create better jobs and bring better visitors with them.

Cantrell shows me the Icelandic equivalent of a massive A/C unit: an opening in the wall that lets in the outside air but keeps out wind, rain and volcanic ash.

The air is monitored and circulated, of course. But even in late June, it's chilly enough to cool the equipment without refrigeration. It's actually mixed with warm, recirculated air to reach the target temperature. (In winter, Cantrell says, it never gets too cold - unlike, say, northern Sweden, where Facebook is setting up shop.) And that helps keep costs down even further. In Verne's marketing materials, the company promises reduced cooling costs "by 80% or more."

Verne doesn't specify prices. But the point is that hosting in Iceland is designed to save a lot of money over markets like New York and London, where Verne will be looking to snag clients. 

"In Manhattan, for example, based on our marketplace statistics, you're going to pay around $500 per kilowatt of usable power," says Jason Evans, CEO of Stackpop, an Internet infrastructure marketplace based in New York. "In Iceland, you're going to pay slightly above $125 per kilowatt usable. Getting four times more power for your dollar for large server deployments could be enormous in terms of cost savings."

Why Iceland Now?

So why isn't Iceland a data center capital already? It has always been there, near the U.K. and a short flight from the East Coast. If you do business there, you get to go to Iceland "for work" whenever you want. Hello, Blue Lagoon and grilled lobster!

In the past the biggest hurdle was bandwidth: Iceland didn't have the right underseas cables to support high-speed, low-latency data centers. Now it does, and new cables may eventually connect it with Canada, New York and Ireland, adding more capacity and quicker speeds. Today, a major U.S. or U.K. bank still isn't going to house its trading servers in Iceland. But Cantrell estimates that an enterprise could host 75% or more of its applications at Verne, ranging from data analytics to email to backups.

Taxes were also a problem. Iceland stuck companies with a 25% tax on servers they brought in - a nonstarter that would have negated most cost savings. But the country's parliament has resolved that issue: no more 25% tax.

Then there are Earth-related concerns. Will a volcano blow? Will an earthquake destroy everything? No, probably not. But Iceland is an island. You have to fly there - it's not like taking a car from Manhattan to New Jersey. Add it all up, and no one has been fired for not betting on Iceland as a data center location.

After an official launch this spring, Verne's server racks are starting to light up. Datapipe, a big hosting and colocation provider, is in. GreenQloud, a startup based in Reykjavík, offers Amazon-like cloud services hosted at Verne and Thor. CCP Games of EVE Online fame, founded in Reykjavík in the 1990s, is a Verne client. And Cantrell says a few potential European and North American customers were kicking the tires during my visit. No Facebooks or Googles, but that sort of thing takes time. Earlier this year, Bloomberg reported that Microsoft may be looking at Iceland.

The good news: Because Iceland is so small - and because the $100-plus billion global data center industry, per Gartner, is so huge - even modest success would make a meaningful change for the country.

The Verne campus might eventually top out at only 100 employees, Cantrell estimates. But "100 jobs are just great, really, for us," says Magnús Orri Schram, a fast-rising member of Iceland's parliament who played a role in getting the server tax taken care of. "We're a very small community." (So small, in fact, that Schram invited me to his house to chat on his sunny porch during his vacation. Great country!)

Iceland's largest and state-owned power company, Landsvirkjun, hopes to command 1% of the European data center market by 2020. In terms of power consumption, that could amount to 1.5 terawatt hours, or about 10% of the energy Iceland generates today. So data centers wouldn't be the biggest customers, for sure, but they would be important ones. And, again, they would be paying better rates - and creating better jobs - than the aluminum plants. 

Remember, this is an island the size of Kentucky, with the population of St. Louis, in the middle of nowhere, coming back from financial collapse. Creating 100 jobs in Iceland is like creating 100,000 in the U.S. So if these data centers take off, they could be a great thing for Iceland.

And if it also means lower-cost and more sustainable service for Internet companies, and less strain on the planet, it's an easy win-win-win.

If ever there was a time that seemed to demonstrate the vulnerability of cloud computing, last weekend was certainly it. Mother Nature’s capricious whims put the smackdown on an Amazon Web Service (AWS) data center in Ashburn, Virginia, Friday night, bringing down hundreds of websites and quite a few popular online services.

But human judgement and a reliance on what some call Amazon’s “sell it cheap” service performance may also have played a role in the storm’s effects on Netflix, Pinterest and Heroku.

The story seems simple on the surface: A massive line of thunderstorms plowed through the U.S. Midwest and Eastern seaboard regions on June 29. They reached Virginia with powerful straight-line winds - a phenomenon known as a derecho - that killed over 20 people and left millions without power. A huge disaster any way you look at it.

But should the storm have wiped out a big swath of the Internet as well?

Failure to Failover

It is not immediately clear why the application services for these sites were not configured for a failover mode when the data center went down, particularly since the same data center had a power outage just 15 days earlier on June 14 - a power outage that affected many of the same sites.

Some Web service providers learned their lesson from the earlier outage: HootSuite was bumped offline when the Ashburn data center went down that day, but it managed to avoid being taken down again on June 29.

HootSuite’s resilience was likely due to the company’s strategy of working with Amazon as a cloud provider. A HootSuite spokesperson told ReadWriteWeb today that the company has multiple backups across different availability zones and data centers.

But not every Web-based company has such a forward-thinking policy, laments Jason Currill, CEO of Ospero, a London-based global hosting and infrastructure company that provides supplementary and backup cloud services.

Web Companies Need Backup Cloud Providers

“This is what happens when we put all our eggs in one basket,” Currill said. “And in the cloud, when one thing happens, it all comes down like a house of cards.”

Mixed metaphors aside, Currill is frustrated by the lack of attention many large-scale Web services are providing to their cloud strategies. Too many times, these Web service providers choose not to build any redundancies into their systems.

“They always cite cost” as a reason why, Currill explained, which he finds exasperating.

“Look at Instagram. They’re nine guys in a room that just got bought by Facebook for $1 billion," Currill said. "You really think those guys don’t have a spare nickel lying around for a DR [disaster recovery] plan?”

Currill is also scornful of Netflix, another site caught in the Ashburn outages last month. The company often cites the need to keep customer costs low, but Currill believes that if Netflix had a true competitor, it would lose a lot of customers when this kind of thing happened. Churn, he added, would change Netflix's minds about the importance of proper disaster recovery and failover planning.

Competition for Amazon Web Services

Ospero both works with and competes against Amazon, and not surprisingly, Currill is glad that Amazon is getting more competition in providing computing services in the cloud. “With Amazon services, no one’s exactly surprised that they keep going down like this,” Currill said. “Their model has always been stack it high and sell it cheap.”

That could change as Google (see "Google Compute Engine a Direct Challenge to Amazon Web Services") and Microsoft step up their own Infrastructure as a Service (IaaS) offerings. “I’m expecting quality from Google and even Microsoft down the road,” he said.

Customers using any cloud services, though, should take a very realistic look at their own disaster policies. Whether its a derecho or a bad line of code, servers will go down, and Web-based businesses had better be prepared.

Image courtesy of Shutterstock.

Not Like Hadoop

Red Hat JBoss Data Grid 6 is not like the big-data solutions found in the Hadoop ecosystem: as an in-memory, key-value store, Data Grid is much more optimized to handle the operations that Hadoop simply can’t: transactions like the kind found in e-commerce and financial trading systems.

Because so much read and write activity is needed in a single transaction, a batch processing system like Hadoop or even a relational database can’t keep up with the speed and scaling necessary to make huge systems found at banks, trading houses or mega-merchants like Amazon.

Countering the issues of speed and scalability is the requirement to adhere to the Principles of Distributed Computing formed by computer scientist Eric Brewer. Specifically, Brewer postulated the CAP Theorem, which holds that systems must be consistent with data, all requests must be available, and partition tolerance must be in place in case of system failure.

To get the speeds and uptime it needs, Amazon will use its non-relational Dynamo database to apply an “eventually consistent” approach to their database systems.

Dynamo is part of a class of non-relational databases known as distributed key-value store (DKVS) databases. DKVS is one of five classes that comprise the topology of the NoSQL landscape, each with a different architecture and approach to managing data.

DKVS databases, also known as eventually consistent key-value store databases, are specifically designed to deal with data spread out over a large number of servers. These systems use distributed hash tables for their key-value stores, and because they’re distributed, the database uses peer-to-peer relationships between servers, with no “master” control. Currently most of the databases in this class are Dynamo or Dynamo-based implementations of Dynamo, such as the open source Project Voldemort, Dynomite, and KAI databases.

Key-value store (KVS) databases are similar in architecture to DKVS. But instead of being distributed across servers, data is held on disk or in RAM.

The Big Data Ecosystem

This, then, is the sector of the big data ecosystem in which JBoss Data Grid lives. Based on the open source Infinispan platform, the Data Grid product sits as an addition to the JBoss Enterprise Application platform, Red Hat’s flagship suite of middleware services.

The use of Infinispan makes a lot of sense in this context, since from its inception Infinispan was not just meant to be another data grid storage system, but also one optimized for the cloud. Itself loosely based on JBoss Cache, Infinispan took the clustered-caching libraries of JBoss Cache and bulked it out to a full-fledged platform optimized for the cloud, according to Manik Surtani, Founder & Project Lead at JBoss Cache, Infinispan Data Grid, when he introduced Infinispan in 2009.

Three years later, Surtani’s vision is being realized. The Data Grid software can be accessed by other applications either through Representational State Transfer (REST), the standard memcache API, or the HotRod API. HotRod is another spin-off from Infinispan that facilitates the elasticity of Data Grid - the capability to scale up and down as needed.

Surtani described HotRod’s function as enabling two-way communication between client and server so that as the shape of the data grid changes in the cloud, client nodes will be be aware of their own standing within that ever-changing data grid, thus creating more efficient operations within automated cloud systems.

Becoming a Major Datacenter Player

This additional piece to the JBoss family of products will set Red Hat as a major datacenter player, giving customers a more complete stack to work with for their cloud-based operations, which is exactly where Red Hat wants to position the Enterprise Application Platform.

“JBoss Enterprise Application Platform 6 is really a major milestone for us,” said Craig Muzilla, Red Hat vice president and general manager of the middleware business unit, in a Web broadcast Wednesday. “It was designed from the ground up to be cloud-ready and to support cloud deployments.”

This fear, justified or not, is one way to explain the persistent efforts of many enterprise IT departments to forego public cloud-computing services in order to maintain their own private-cloud systems. It may also offer insight into the hype around datacenter tools such as Hadoop, Cassandra and Pig.

It’s Good to Be King - of the Datacenter

The overall goal? A Game of Thrones-style power play by IT departements intended to reinforce their own importance by keeping the datacenter local.

According to a March 2012 IDC study sponsored by Microsoft, cloud computing will create 13.8 million jobs worldwide in the IT sector by the year 2015. What the report also highlights is that while public cloud spending will go up, general IT spending will not grow as fast - and in some regions could actually be scaled back.

This makes sense. Corporations' budgets aren’t unlimited, and spending on one line item can bring down the amount available for other things. So, while it’s great that nearly 14 million cloud jobs are on the way, that doesn’t necessarily add up to bright prospects for corporate IT positions.

So despite the overall benefits of cloud computing, can we really blame IT workers - and even IT leaders - for dragging their heels on moving to the cloud? Or for doing anything else they can think of to remain relevant?

Hugging the Hardware

“People get nervous about jobs when they start thinking about their hardware going somewhere else,” blogged NTT Communications engineer Ladd Witmer. “There is an entire group of people in most organizations that do the systems administrative work, as well as the cabling and racking and stacking. There is also another group that manages the physical infrastructure and who would be affected by a movement to a public cloud provider.”

It’s not all doom and gloom, of course. Witmer goes on to write that “fewer people will be needed to maintain the basic infrastructure. Those people could, however, be redeployed to provide higher value to your organization.”

“Could” is not the same as “would,” though, and the change is still threatening enough for many IT workers to work hard to prevent. Better, they may reason, to spin change in their own way, safeguarding their own careers by implementing new solutions in their own datacenters.

Hadoop to the Rescue?

Perhaps chief among those kinds of solutions are Hadoop and its attendant ecosystem of tools that are used to tackle the problem of big data.

This is not to paint Hadoop as no more than a way for IT workers to hold onto their jobs. The needs Hadoop addresses are quite real. But there seems to be an almost frantic focus on Hadoop as the only big data tool worth considering, when capable tools including MongoDB, HPCC and 1010data are delivering their own scalable big data solutions.

And it’s not just big data: appliances, bring-your-own-device (BYOD) policies and social media are other areas where IT is struggling to decide whether to fight, admit defeat or try to take ownership and drive the discussion.

There are plenty of legitimate technology and business concerns around those issues, and it’s probably counter-productive to challenge IT workers' motivations. But it’s important to understand all the factors influencing the reactions to potentially disruptive technologies. That’s especially true when considering the technology is as powerful and transformative as cloud computing.

At the center of one of these dramas is a player that officially exited the enterprise computing market in November 2010: Apple. The reason: Apple makes a tablet that CxOs really want. Many may not actually know how it integrates with their networks, but unlike most any technology purchase to date, they're willing to invest in it now and figure out the solutions down the road.

Forrester's prediction, as articulated by a team led by Andrew Bartels of its Vendor Strategy Professionals group, is unusual but plausible: The rise of cloud computing in 2011, Bartels notes, led to a rise in server equipment sales. But the principal buyers were actually just a handful of customers who needed a broad infrastructure platform now. One such player was Rackspace. Another was China, as in "government of."

But those purchases are made - they're done. Meanwhile, Forrester survey results for Q2 2011 were the first indicator of trouble signs for enterprises smaller than the Chinese government, such as banks. Only about one-fourth said they really have an IaaS strategy, with many indicating they don't really know what an IaaS strategy is.

What has enterprise executives' attention locked up? Tablets, particularly the iPad. Until CxOs stop staring at iPads like cats with yarn dangling in front of their faces, Bartels' report asserts, the growth of cloud computing infrastructure within the enterprise will actually take a dip.

"In 2011, we estimate that Apple will sell $6 billion worth of Macs and an equal amount of iPads to the corporate market; in 2012, we project $9 billion in Macs and $10 billion in iPads; and by 2013, $12 billion in Macs and $16 billion in iPads," Forrester's report reads. "In contrast, global corporate spending on Wintel PCs and tablets will decline by 3% in 2012 and by 1% in 2013."

For the smaller businesses on the enterprise scale, the adoption of iPads has a long tail to it that brings in more Macs, particularly among members of the IT department themselves who evidently prefer working on Macs. Because of them, Forrester believes, Apple could double its worldwide sale of Macs (a majority of which are sold in the U.S.) in a two-year period, at the same time that spending on PCs levels off, and spending on Windows-based PCs declines. Note the forecast for "Wintel" PCs for 2013, which should be the year of Windows 8.

Spending on cloud infrastructure will resume its upward trend in 2013, Forrester projects, apparently when the iPad novelty wears off. Delivering software by way of services will not only continue to redefine the software market, but will also bring customers into that market as players themselves, providing services to their own customers online. Of course, this is what many analysts foresaw a year ago for 2012, but the iPad proved to be an unexpected event from an underappreciated source.

In context, Raftery is writing about a pair of reports from the Carbon Disclosure Project (CDP) and Verdantix. Raftery argues that "Cloud Computing – The IT Solution for the 21st Century" (PDF) and an addendum for France and the UK (PDF) are fundamentally flawed.

Why is that? In the case of the first study, the assumption is that reducing energy consumption will reduce carbon emissions. Raftery says that the flaw is assuming there's a direct relationship between energy usage and carbon emissions. "If I have a company whose energy retailer is selling me power generated primarily by nuclear or renewable sources for example, and I move my applications to a cloud provider whose power comes mostly from coal, then the move to cloud computing will increase, not decrease, my carbon emissions."

In the latest report, the assumption is that companies in France and the UK will move their applications to the cloud hosted in France or the UK. Where are most cloud hosting providers? Depends on which provider you're looking at. Quite a lot of it is in the U.S., which, as Raftery notes "has one of the most carbon intensive electrical grids in the world. France, on the other hand, with its high concentration of nuclear power (78%) has one of the least carbon intensive electricity grids in the world." The UK sits just above the world average, according to the same data.

Migrating a workload to the cloud can have a positive effect, of course, but it's hard to tell. Raftery suggests that cloud providers need to do a much better job of being transparent about the locations of the data centers, along with the carbon footprints. Without that information, you don't really know how green the cloud is after all.