The free learn-programming-online site now has 11 modules and 86 lessons in all things PHP, from the basics and the standard if/else logic flow to arrays, functions, and advanced object-oriented programming.

“The future of programming is making a tool for people to demonstrate their creativity,” Codecademy founder and CEO Zach Sims told me a few months ago. “In 2013, we want to extend Codecademy’s reach and help people move beyond beginners.”

PHP is a good start, as the language is simple enough to let beginners get their feet wet while powerful enough to build world-class websites like Facebook. And with over five million PHP developers on the planet, new programmers would be joining perhaps the largest cohort of developers around, with plenty of code help, sample, code, and forums to help them progress beyond the basics.

And with Andi Gutmans and Co. at Zend focusing on bringing the power of PHP to blended mobile applications, PHP skills are helpful not only on the server but also in cloud-connected native app development.

PHP has been one of the most-requested languages for Codecademy to add to its coursework, Sims told me. Just three months ago, Codecademy launched API training with YouTube, NPR, Bit.ly, and six other partners. Now students have one more language, besides the already existing Javascript, HTML/CSS, Python, and Ruby to use them with.

Codecademy’s introduction to its PHP courses:

PHP is the world’s most popular server-side scripting language. Interested in processing form data from your website, creating HTML on the fly, handling cookies, or talking to a database? PHP’s got you covered.

Courses are created by the community. We’re launching PHP as a work-in-progress and we want you to help create the future of PHP instruction for people all over the world.

photo credit: CyberHades via photopin cc

---

Big Data and Predictive/Real-time Analytics startups: Are you looking to jumpstart development & accelerate market traction? Sign up for the SAP Startup Focus program to receive technology, support, resources and community to help you develop new applications on SAP HANA, a cutting edge database platform. Get started here, and enter promo code “VB2013″ on the form.

---

Sept. 9 - 10, 2013
San Francisco, CA

Early Bird Tickets on Sale

Cloud-computing giant Rackspace has made a small but smart acquisition in ObjectRocket, a MongoDB database service.

Rackspace senior vice president of corporate development Pat Matthews said, “Customers have been asking for a powerful database,” and the acquisition is in line with their open cloud strategy. MongoDB was the best option from Rackspace’s perspective as it’s one of the fastest growing NoSQL open source databases, and many existing customers already use it.

According to a report from British technology research firm The 451 Group, NoSQL software revenue is expected to grow at an annual of 82 percent to reach £141 million ($214 million) by 2015.

Database technology floods the market, but Matthews said that MongoDB has experienced a “runaway success” with developers. It’s easy to get started and supports about a dozen programming language on the Web, including C, C++, C#, JavaScript, Node.js, and Scala. But it can be difficult to manage and scale, which is where ObjectRocket comes in.

In the future, Rackspace is likely to support other database technologies.

ObjectRocket co-founder Kenny Gorman has been recognized as a “MongoDB Master,” an award 10gen gives to only 35 MongoDB contributors around the world. (Check out a video interview with Gorman and Rackspace’s chief developer evangelist Robert Scoble below.)

The full team will relocate to Rackspace’s San Antonio headquarters (they previously worked remotely), and customers should see the benefits of an integration in a matter of weeks.

On the topic of Rackspace’s broader acquisition strategy, Matthews revealed that his team doesn’t plan to buy established companies. “We have a good sales and marketing engine and know how to acquire customers,” he said. Instead, they are on the lookout for small, under-the-radar teams with impressive technology. The company’s other recent acquisition, MailGun, was a six-person operation with just a seed round of funding.

Rackspace has not disclosed the terms of the deal.

---

Big Data and Predictive/Real-time Analytics startups: Are you looking to jumpstart development & accelerate market traction? Sign up for the SAP Startup Focus program to receive technology, support, resources and community to help you develop new applications on SAP HANA, a cutting edge database platform. Get started here, and enter promo code “VB2013″ on the form.

---

But it’s also costly.

And it’s relatively easy to have downtime. 40 percent of service interruptions are caused by power station failures. 25 percent are hardware failures, and 13 percent are software failures: mis-configuration or severe bugs.

Even when a service is up 99.5 percent of the time — which might sound pretty good to non-developers — it’s down almost 44 hours in a year. Three nines, or 99.9 percent uptime, is almost 9 hours of downtime, and four nines, 99.99 percent, is still almost an hour of downtime in a year.

Data and network services company MegaPath put together this infographic showing the causes and results of downtime:

photo credit: ex.libris via photopin cc

---

Big Data and Predictive/Real-time Analytics startups: Are you looking to jumpstart development & accelerate market traction? Sign up for the SAP Startup Focus program to receive technology, support, resources and community to help you develop new applications on SAP HANA, a cutting edge database platform. Get started here, and enter promo code “VB2013″ on the form.

---

Sept. 9 - 10, 2013
San Francisco, CA

Early Bird Tickets on Sale

10 messages per second to 1,000,000 live users on single, solitary server rack? That’s impressive by any standard, and that’s what Kaazing and Dell were able to achieve with living, dynamically updating webpages.

“Dell came to Kaazing … and asked if we could drive one message per second to one million users,” Kaazing executive vice president John Donnelly said in a statement. “We did just that and then some, achieving 10 messages per second to the same one million users.”

That’s on a single Dell PowerEdge R620 server rack using dual 3.3GHz quad core Xeon CPUs, 32GB of RAM, plus dual 10Gb Ethernet cards and four 1Gb Ethernet network ports.

Dell says that the system, which used TIBCO Enterprise Message Server, “can now support a higher number of users than ever before at speeds never before possible.”

In the old days, web browsers forced users to refresh entire web pages to update a single element are over. Modern web pages, however, enabled by modern web browsers, now allow dynamic updating of individual components. One such protocol is WebRTC (web real-time communication), but the Kaazing software uses the HTML5 WebSocket standard.

To achieve scalable live page element updating is a somewhat new challenge for web servers.

Live updating can create a great experience for end users, but it can also put a massive demand on web server software and hardware that were previously optimized for sending out complete pages every 15 to 50 seconds, but now are getting dozens of small update requests a second.

So 10 million messages per second is impressive. Also impressive is that the companies delivered them from server to client in an average 3.5 milliseconds each.

One little caveat: If you check out Dell’s online configurator for the PowerEdge R620, you’ll notice no 3.3GHz Xeon processors available. The top of the line orderable CPUS are 2.2GHz.

Update: Dell has updated its product ordering space and the new 3.3GHz processor is now available for order.

But when it comes to website availability and uptime under crushing Olympic-sized traffic, the race has already begun.

Compuware has been tracking the performance of the top 20 Olympic news sites in countries such as the U.S., UK, Germany, and China for several months. Now that the Olympics are on and the surge of traffic is hitting, some clear front runners have emerged.

“We’re measuring both uptime and availability,” Compuware product manager Lorenz Jakober told VentureBeat. “Then we compare it to the performance of the sites a month ago, which gives us a pretty good metric on what the differences are.”

And those Aussies are definitely feeling the heat. Websites in the basket of news agencies Compuware is tracking, such as ABCnews, 9News, news.com.au, are down between 10-20 percent compared to pre-Olympic traffic:

Source: Compuware

Australian Olympic sites: red, like blood, is a bad sign

For the U.S., measured properties would include sites like NBC.com, ESPN, and Fox Sports. And paradoxically, those sites are doing better than before the Olympics.

Source: Compuware

US Olympic news sites: green is good

“It’s funny actually,” Jacoba said. “Perhaps the Olympic fever hasn’t maybe hit us yet, with the time delay.”

Or perhaps server administrators have increased capacity prior to the Olympics. That’s exactly what the Bleacher Report was doing when I spoke to the company’s vice president of engineering a few weeks ago. That site is expecting to triple ordinary web traffic, and preemptively scaled up its database layer inside Amazon’s cloud. When it comes to serving web pages, the Bleacher Report can add 50 percent more capacity in just 30 minutes.

Perhaps the French should have done something similar:

Source: Compuware

Fo the French, it’s a little comme ci, comme ça

Image credits: Compuware, fstockfoto / Shutterstock.com

CEO Brendan McCorkle

CloudMine started out as a business-to-business big data service, but quickly pivoted into a instant, managed backend for app developers. The goal: enable app developers to focus on the front end of their applications … the user interface features that users actually touch, rather than the infrastructure and data management that enables them.

VentureBeat spoke to chief executive Brendan McCorkle yesterday.

“Any time you build an app there’s all this scaffolding that needs to go up: infrastructure, account management, data security and other basic services just so an app can run,” says McCorkle. “We build that scaffolding so you don’t have to.”

For example, a local business-finding app might need geolocation support, user registration, some definition of public and private data, and more. CloudMine enables all the backend components of this through a single API … which has the potential to significantly simplify development and shorten build time.

“Essentially, we can abstract away the backend. We can abstract away EC2,” McCorkle said, referring to Amazon’s Elastic Compute Cloud, a common cloud computing service for developers. “That enables you as an app developer to focus on your core competency and do your thing.”

For almost all apps, he argues, the server side infrastructure is not what makes a developer’s app unique. Connection to the cloud for data and computing are essential, in fact business critical, but they are not what the user sees. They are the underlying components that support the app, and CloudMine’s goal is to provide them quickly and easily.

In addition, the company promises benefits that extend beyond simplification and cost savings.

Because CloudMine abstracts the backend, app makers are able to host their data on multiple Amazon availability zones without any additional work, significantly improving the chances of their service staying up even if a part of Amazon’s cloud goes down. In fact, McCorkle told VentureBeat, this could even extend to mirroring on multiple clouds from additional providers such as Rackspace, further increasing reliability with zero additional developer overhead.

The company is seeing some traction in two very different spaces: enterprise and startups.

CloudMine has two Fortune 500 customers who are currently using the company’s backend to expose legacy data to mobile applications. And startup clients are using the backend infrastructure to rapidly prototype and deploy sophisticated mobile apps.

CloudMine began its young life at Startup Weekend Philly just nine months ago, and is currently running more than 1500 apps on the platform. Pricing is simple: $.05 per active user per month, or custom pricing for large data consumers.

The company received $20,000 in seed investment, and a $100k investment from Ben Franklin Technology Partners of Southeastern Pennsylvania. CloudMine is based in Philadelphia.

Image credit: ShutterStock

VentureBeat recently talked to Nginx co-founder Andre Alexeev about the company’s high-performance server software that is now powering over 25 percent of the world’s top websites.

Andrew Alexeev

“We’re like the man behind the curtain,” Alexeev says. “We power over 70 million sites overall — about 12 percent of the Internet, but still a lot of people have not heard of Nginx.”

Nginx, pronounced en-gin-ex,was recently chosen as one of the foundations of Netflix’s content delivery network, or CDN. CDNs help companies with massive data distribution needs serve clients all over the globe quickly and efficiently. Most companies choose a commercially available CDN, but at some size, it becomes more economical to build your own. That’s exactly what Netflix did — essentially following the YouTube model.

Serving content for Netflix is, obviously, not small potatoes. In June the company said it was serving nearly a billion hours of video each and every month. That translates into huge amounts of data, in some cases up to 32 percent of peak Internet bandwidth in North America alone. Not all of that is transferring immediately to Netflix’s own CDN; the company is taking a measured approach to the project. But the cost savings mean that it is the company’s preferred direction.

That’s just a fraction of what Nginx is doing.

The server technology powers the websites of some of the most well-known brands online. Think the busiest site on the planet: Facebook. And WordPress. Not to mention Hulu, Zappos, Zynga, Dropbox, and Yandex, the big Russian search engine. Why are all these giants of the web using Nginx?

“We’ve built Nginx to be much more efficient, much more scalable,” Aleexev told VentureBeat. “We’ve designed it to extract the most performance from hardware — it’s built to scale from 1,000 concurrent connections per server to 1,000,000 per server.”

The company is very young, exactly a year old today. But the Nginx open source software project has its roots in 2002. The server software was first released in 2004, and the core team has been iterating ever since. Nginx, the company, plans to make money the way many companies built around open source software do: consulting, services, implementation, strategy. And that’s pretty much what they did for Netflix.

“We helped them with the proof of concept and helped optimize the operating system they are using, which is FreeBSD,” Aleexev says. “We also optimized Nginx for the hardware.”

But Nginx also plans to offer commercial products, including extensions to the server software and a new product in the web acceleration space. To accomplish this, the company raised $3 million in venture capital last year from BV Capital, Runa Capital, and MSD. There’s still a long way to go to catch market leader Apache, currently running 64 percent of the world’s top servers. But Nginx’s focus is not just the most sites: it’s the busiest sites.

Server market share as of June – green line is Nginx

Image credit: ShutterStock

Tucked away in Douglas County, Georgia, next to a Six Flags theme park, lies one of Google’s data centers. The facility houses   unknown number of servers that deliver Google search, Maps, and Gmail. All these servers are powered by up to 250 kilowatts of electricity and can generate high levels of heat, which Google has worked hard to combat. In the Douglas County facility specifically, servers are kept cool with recycled, undrinkable water from the surrounding area.

Since 2008, the search-engine giant has been releasing these annual reports to tell us all how Google keeps its energy footprint low. Each report includes the power-usage effectiveness (PUE) of the collective Google data centers. Google defines PUE as “a ratio of the total power used to run a data center to the amount used to power the servers.”

Using Energy Star data, Google points out that other data centers use equal amounts of energy to power servers and the facilities that house them. In this case, those other data centers have a PUE is 2.0, meaning that one watt of energy goes to running the servers and one watt powers the data center’s lights and cooling systems. Google has shifted that ratio, claiming to have a 12 month average PUE of 1.14 in 2011, down from 1.16 in 2010. Not exactly 50 percent less energy than its competitors, but close.

Google employes a few tactics to keep its PUE low. It keeps its cold aisle temperatures around 80 degrees to save energy and uses outside air to cool its data centers, instead of powered cooling systems. Google has also built its own energy-saving servers.

You can read Google’s full data center energy report, along with reports from previous years, here.

Server room image via Shutterstock