And just who has built storage systems using the Backblaze specifications? Netflix is probably the most-famous adopter — it uses storage pods as part of its content-delivery network infrastructure — but others include Vanderbilt University, Crispin Porter + Bogusky, Rensselaer Polytechnic Institute and NASA’s Jet Propulsion Laboratory. Popular online photography service Shutterfly stores petabytes worth of users’ old photos on BackBlaze’s storage pod architecture.

Gleb Budman

Their uses are as diverse as their organizations are. There’s Netflix’s CDN and Shutterstock’s consumer cloud storage, while many are using pods as giant NAS devices that everyone can access. “Its more data than they ever thought could be possible for their company,” Backblaze Founder and CEO Gleb Budman told me. ”They just RAID them and they go.”

Oh, and he added, “I know of at least one individual person who built one of these for himself for his house.” It stores his media collection and helped his marriage. It appears some wives don’t appreciate sprawling hard-drive farms sucking up energy and taking up all the garage space.

Disrupting the storage industry 1 terabyte at a time

That the Backblaze design has caught on so broadly shouldn’t be surprising, Budman said. For years, storage vendors have been protecting their margins by loading even customers’ “write once, read very rarely” systems with enterprise-class features that often weren’t necessary. If you wanted an enterprise-class storage system, “you bought a NetApp,” he joked, and if you wanted just to house some non-critical data, “you bought a NetApp.”

So, in the world of high-volume storage, we’ve come to a place similar to the PC market decades ago when it was cheaper to just buy the parts and build your own than it was to buy a pre-assembled computer. “Dell basically killed the homegrown computer market because they really, really focused on optimizing costs,” Budman explained. “No one did that for storage equipment. [Storage vendors] said, ‘Hey, we’re selling a million-dollar design, why would we change that?’”

Backblaze’s data center full of storage pods.

Thanks in part to Backblaze, though, this system is changing. Ceaseless demands for parts led Protocase, the Canadian sheet-metal fabricator that makes Backblaze’s pod enclosures, to create a whole business — the aptly named 45 Drives — around selling pod parts or even wholly pre-assembled pods (second-generation ones start at $5,395 without the 45 hard drives they hold). Where it used to struggle to get business outside of Canada, Budman said, Protocase has sold Backblaze units to places as far away as China, Russia and Brazil.

Global electronics fabricator and supply-chain specialist Sanmina sells a modified version of the Backblaze pod design, as do a handful of value-added resellers and components companies around the world.

Interestingly, one place you won’t see Backblaze designs is in the other famous open source hardware effort — the Facebook-led Open Compute Project. Budman said he’s had conversations with the organization and some of its leaders and there has been interest in getting Backblaze involved, but that “for the most part what they want is help making their Open Compute system work.” He said he’d love to do it in theory, but there’s only so much time for a small company like Backblaze to spend on missions aside from improving its business.

Version 3.0: Now with 180TB and a lower cost

As for those third-generation storage pod designs, Open Compute Project, 45 Drives, guys with huge digital media collections and anyone else interested in building their own gear should have a lot to be excited about. Total capacity has been boosted to 180TB thanks to the prevalence of 4TB hard drives, and Backblaze has certified a few more types of hard drives because of the harsh lessons it learned about reliance on a single model during the hard drive shortage in 2010. The company has also replaced a bunch of the components it uses, everything from motherboards to memory to SATA cables.

Budman explains all the changes and the rationale behind them in a blog post published Wednesday morning, but the general theme is improved reliability and ease of management at a lower cost. All told, the new designs cost $1,942.59 — $37.41 less than the second-generation ones. Because of that recent shortage, though, hard drives still cost a little more than they did a few years ago.

Whatever comes of its efforts to be transparent about storage system design, Budman hopes it at least has a lasting effect on the availability of affordable storage. Organizations, he said, should “no longer have to make the decision between an expensive piece of equipment and not storing data.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Migrating media applications to the private cloud: best practices for businesses
• Strategies for the Future of Digital Content Storage
• The role of converged infrastructure in the data center

I wrote about the hackathon, while my colleagues Stacey Higginbotham and Derrick Harris reported on other news from the conference.

In today’s post, one coordinator of the hackathon, Zak Homuth, CEO of Upverter, provides details on the seven competitors, from a strip of temperature sensors for server racks to a rack that accommodates Open Compute servers and legacy servers.

The post also names the winning hack, a transmitter that plugs into an Open Compute server debug port and sends information to a server rack’s aggregation point, which can then send the information to the data center’s management framework. The device can also have an LED.

The four inventors of the device call it Project Cheesy Fingers, they told me last week. The coinage is a nod to the real life of a data center employee, who has cheesy fingers on one hand and a beer in the other one when he or she spots a technical problem. The device helps the employee determine if the problem is worth getting out of the chair.

One of the transmitter’s inventors, Prof. Andrew Cencini of Bennington College in Vermont, has written a post about the device.

The Open Compute Project Foundation will support the group’s efforts to patent their invention or build a prototype, Homuth wrote in his post.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Migrating media applications to the private cloud: best practices for businesses
• How direct-access solutions can speed up cloud adoption
• WAN design for the cloud age

Thus the hackathon at the Open Compute Summit on Wednesday. While most hackathons focus on coding, Facebook and Open Compute are hoping to use tools from companies such as UpVerter and GrabCAD to help make the collaborative problem-solving that occurs so easily around code, happen in hardware. These companies offer web-based collaboration software with UpVerter letting engineers share circuit designs and libraries and GrabCAD performing a similar service for mechanical designs.

After keynote speakers talked up the importance of collaborating to advance cloud-computing infrastructure at the event in Santa Clara, Calif., about 25 people formed seven groups and discussed ways to implement hardware concepts they had brainstormed in a matter of hours.

One group cooked up plans for a device with an LED that can plug into an Open Compute server and display debugging information. Another group threw around visions of a strip of thermometers that can go inside a server rack — useful for telling data center operators how hot their machines were running. And another drew up a prototype for a server rack that could accommodate the larger Open Compute servers and more traditional 19″ servers.

“There will be some very significant designs over the next six or eight hours,” said John Kenevey, the technical evangelist for Open Compute and a program manager at Facebook.

Similar events could play out at colleges and universities if Kenevey has his way.

Over the next year or so, Kenevey will flesh out the details of a higher-education Open Compute hackathon model — known for now as OCPU, short for Open Compute Project University — and approach schools about it.

The challenge is to make on-campus Open Compute hackathons appeal to students who often participate in hackathons. Perhaps students could receive course credits for teasing out Open Compute ideas, he said.

Such events wouldn’t be the first link between the Open Compute Foundation and academia. As my colleague Stacey Higginbotham wrote in November, an Open Compute Project contest is underway at Purdue, where students will try to design an Open Compute server chassis that’s biodegradable. Purdue students setting out on the project stopped by to witness the beginning of Wednesday’s hackathon which is set to last 12 hours.

This story was corrected on Thursday, Jan. 24, with the correct description of a device developed at the Open Compute hackathon.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• A near-term outlook for big data
• Migrating media applications to the private cloud: best practices for businesses
• Dissecting the data: 5 issues for our digital future

It should come as no surprise, then, that the companies serving the country’s 1.3 billion people with their social media, e-commerce and information-discovery needs are very, very big, too. Here are some statistics that demonstrate their scale.

Alibaba Group

Taobao, the eBay-like e-commerce line of business from Chinese internet giant Alibaba Group, does a lot of business. On a single day — Nov. 11, 2011 — the company did a whopping 19 billion Yuan (or approximately $3.05 billion) in sales. According to Alibaba Group CTO and Alibaba Cloud Computing President Wang Jian, the company site surpassed the 1 trillion Yuan (about $160 billion) mark for 2012 revenue at the end of November. Alipay, the company’s version of PayPal, handles about 3 billion Yuan (about $480 million) in transactions every day.

By comparison, eBay posted $3.4 billion in revenue for the entire third-quarter this year. Amazon.com, with which Taobao also competes (although Alibaba also has a business-to-consumer division called Tmall), closed its third quarter with $13.8 billion in revenue.

The women’s shoe department on Taobao

Of course, Taobao and Alipay are just two of Alibaba’s expansive portfolio of services, which includes a much-publicized (although recently reduced) partnership with Yahoo.

That type of business means Aliaba needs a lot of servers. In a single year not too long ago, Jian told me, the company bought more servers than it had in previous five years combined. If you charted Alibaba’s server count now versus five years ago, he added, the previous number would look like zero. How big is its database? Enough to store data for more than 800 million items for sale.

Baidu

The Chinese search giant is ranked fifth in the Alexa internet rankings (behind Facebook Google, YouTube and Yahoo), which is evidence of its popularity. All those users, I’m told, result in an annual server growth approximately equal to the previous three years combined. It has been reported that Baidu is planning possibly the world’s largest data center — spanning 120,000 square meters, costing $1.6 billion, housing 100,000 servers (totaling 700,000 CPUs and 3 million cores) and storing 4,000 petabytes of data.

Tencent

Somtimes compared with Facebook in the United States (although it’s actually quite different), Tencent boasted more than 717 million users for its popular QQ messaging service as of September 2011. That number has surely grown. The company says its highest-ever number of concurrent users was more than 176 million, although there are often tens of millions (if not more than 100 million people) using it at any given time. An individual with some knowledge of the company’s infrastructure told me Tencent adds about 100,000 servers per year.

Tencent usage at 5:51 local time on Jan. 10, 2012.

Weibo

Weibo, the Twitter-like platform from internet new-school internet company Sina had more than 400 million users as of April 2012. That’s about twice the number Twitter claims. And the Chinese use Weibo a lot, for everything from micro-blogging to self-publishing. It might actually be a more important tool in China than Twitter is in the United States, sources told me, because while the government can censor official news outlets, it can’t possibly control the stream of information coming off Weibo. And that will mean even more growth.

Not (yet) innovators

However, despite their sheer scale, Chinese internet companies are, by most accounts, less technologically inclined than their American counterparts. The biggest reason — one I heard time and time again — is that these companies tend to view themselves as traditional businesses rather than technology companies, and that employees often strive to work up the management ladder rather than remain career engineers. This inevitably affects R&D budgets, makes companies less willing to take risks and reduces the pool of employees that really, deeply understand complex systems.

As an example, one might look at the server situation within China’s big four internet companies. Alibaba’s Jian told me that although his company is running all white boxes in its data centers now, it had a lot of legacy IBM gear in its data centers five years ago. I heard the same thing about Baidu. Tencent, someone told me, had 10,000 webscale servers fail in six months last year and is considering a move back to traditional boxes.

However, maybe these companies are coming around on innovation beyond just buying more-efficient gear. Tencent, Baidu and Alibaba, for example, are all members of the Facebook-led Open Compute Project for designing webscale hardware. Tencent and Baidu actually created their own rack-design specification, called Project Scorpio, that is being merged into Open Compute’s Open Rack design in 2013. They still don’t build their own servers like Google and Facebook do, preferring instead to push their custom specs on server makers, but many innovative American companies, including eBay, do the same thing.

Power specs of Open Rack

One has to assume that a closer working relationship between engineers at American and Chinese internet companies will spur even more changes in the tech culture there. Although technical talent comes relatively cheap in China, perhaps they’ll realize that highly skilled, forward-thinking engineers (and data scientists, for that matter) are something worth hanging onto and rewarding with high salaries.

As Facebook VP Frank Frankovsky told PCWorld in July as the Open Rack designs were unveiled, “We compete with those guys, but on the infrastructure side, if we can make our infrastructure more efficient, it makes everyone that much better. Where we differentiate our business is in the service we provide to our end users.”

That differentiation comes from in large part from an incredible investment in research and technology. If they want to be considered thought leaders in their field — and if they want to expand significantly into cloud computing (as Alibaba and Sina clearly want to do) — China’s internet companies will have to start matching their immense scale with demonstrated technological prowess.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Flash analysis: the future of Yahoo
• The promise of SDNs in the enterprise
• Facebook’s IPO filing: ideas and implications

Smart companies trying to deliver services over the web realize that they’re not in the business of pleasing a CIO, but of pleasing consumers. And consumers don’t care what’s under the hood as long as the service works and their lives aren’t interrupted by a downed server.

That’s kind of the reason that cloud computing exists, and has become such a successful delivery model for IT resources. Users get generic server specs on virtual machines, but all the magic happens at the layers above. Smart developers can take advantage of features such as auto-scaling and failover, as well as myriad open source components and open(ish) APIs, to piece together applications that might not look pretty, but stay online and don’t cost a fortune to run.

The cloud underpins a lot of applications that many web users couldn’t live without, from Instragram to Netflix. They all required some architectural creativity along the way to grow into what they’ve become, and the cloud enabled that.

What’s good for the goose …

Life shouldn’t be any different just because a company decides to run its own servers rather than rent ephemeral boxes from Amazon Web Services. Especially for large-scale web applications or services, it might make good business sense to eschew the traditionally expensive world of off-the-shelf computing hardware and just build the cheapest-possible gear that gets the job done — gear that ends up looking a lot like those generic cloud computing resources.

Google wrote the book on this by designing its own servers, data centers and, it appears, networking switches, and the company has done alright for itself. Facebook, too, is lauded for its custom-built hardware and data centers. Both companies have gotten so good at designing gear and sourcing components that they don’t necessarily need buy much of anything from mainstream vendors in order to fill their data centers.

Facebook’s custom-built battery cabinet

But neither company would consider itself a hardware company (which is why Facebook was willing to open source its designs under the Open Compute Project banner), they just realize that a little hacking can go a long way toward delivering a better service. Economically, custom gear that does away with extraneous bells and whistles while adding performance where needed means a lower sticker price, a lower power bill and a better user experience. Whatever reliability is lost by removing fans, server cases and vendor software is made up for by smart software engineers who design systems that expect gear to fail, and to keep running when that happens.

Google is so confident in its software it promises zero planned downtime for Gmail and achieves higher than 99 percent uptime for the service overall. And ask yourself the last time you remember Facebook, with its 950 million users, crashing. In the enterprise IT world, it’s this type of infrastructural intelligence that’s driving the software-defined network movement, and VMware’s vision of software-defined data centers.

Backblaze blowback

This is why I was a little shocked to see some of the negative comments when I wrote recently about cloud-storage startup Backblaze’s efforts to deal with last year’s hard drive shortage by sourcing consumer-grade external hard drives from Costcos around the country. Given the choices — back off its unlimited storage for $5 promise, or find a way to procure capacity on the cheap — Backblaze almost certainly made the right choice. The software that runs the service expects hard drives to fail, and backing up data doesn’t require blazing fast data access. As long as consumer drives aren’t crashing by the hundreds, users don’t notice a thing.

What’s more, that epic feat of hackery wouldn’t have been possible if not for Backblaze’s even bigger contribution to cloud-service design — its $7,400 135TB storage pods. Because it designs and builds its own infrastructure, the company had the luxury of swapping out the most critical component without worrying about voiding a warranty or messing up something in a storage array whose blueprints it hadn’t seen. It’s possible that the then-bootstrapped Backblaze wouldn’t have made it out of the hard drive shortage alive, or at least without some upset customers, had it not been so in tune with its hardware needs.

BackBlaze’s storage pods in a data center

You can do it, too!

Given all this, it’s not insignificant that Facebook on Wednesday shared some tips to deploying its Open Compute servers in co-location facilities that might not be designed to handle custom rack designs. After all, unless you’re building your data centers like Facebook, Google or eBay, you have to play by your colo provider’s rules. Guidance from Facebook on actually deploying the servers under real-world circumstances makes Open Compute less good in theory and more good in practice.

Traditional big companies might not line up to deploy Open Compute gear as is — they’ll understandably wait until server makers such as Dell and HP productize the designs. But companies that want to be big like Facebook or Google (or even in sheer capacity a la Backblaze) ought to pay attention. Facebook’s server designs could be a great starting point from which to build your own specialized gear to make sure your application and your users’ experience are never at the mercy of fate or a vendor’s bottom line.

Feature image courtesy of Shutterstock user Jason Winter.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The promise of SDNs in the enterprise
• Migrating media applications to the private cloud: best practices for businesses
• Takeaways from the second quarter in cloud and data

This Open Compute Project specification focuses, as GigaOM has reported, on moving data center racks from 19-inch to 21-inch widths — the 19-inch requirement being a holdover (believe it or not) from railroad switching equipment.

According to a blog post on the project’s website:

The 537mm width (about 21 inches) of the chassis has a lot of practical engineering benefits, like improved airflow, greater energy efficiency, and better volumetric efficiency, as there is more space used for IT equipment instead of just air and metal. The rack itself is 600mm wide, which makes it the same as the overall width of a 19″ rack, so it fits into existing data centers worldwide.

The addition of more usable space in the rack, means that components can be arranged differently. Switches, for example, no longer need to go atop the power zone, for example.

Details on the specification are posted on Github.

The Open Compute Project was initiated by Facebooklast year as a way to promote what it has learned from its own energy-efficient data centers as a model for others. Since that time it has helped form a multi-vendor foundation.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Locating data centers in an energy-constrained world
• A near-term outlook for big data
• The capex connection: Why we pay for privacy on the Web

Not so for Backblaze. The San Mateo, CA., company, which offers a cloud-based data backup service, is in the market for a new data center and isn’t shy about discussing it. In fact, it’s posting a link to its RFP on its blog.

“Many companies have this knee-jerk reaction about disclosure. They worry that things they say will somehow be used against them,” Backblaze CEO Gleb Budman told me in a recent interview. Backblaze has been open about some other things, too. It shared information about how it builds the storage pods that run the company’s cloud-based data backup service.  It also talked about why it didn’t initially seek venture funding and also about why it changed its mind.

“There may be some things that should be proprietary for a very specific reason, but if we don’t see a reason not to share, we share,” Budman said

The RFP lays out exactly what the company is looking for in terms of initial physical, power, and network requirements and information on what it thinks it will need for expansion over the course of the next 3 to 5 years. For example, it expects to add 1 to 1.5 cabinets per month over a 36 month period. That’s because Backblaze is nothing if not growing.The company manages 40 petabytes of data now and adds about 2 petabytes per month.

Interestingly, while Apple, Google, Amazon remain incredibly tight-lipped about their data center build outs — our own Katie Fehrenbacher can attest to that after her data center road trip  – there’s also a countervailing trend pushing for more openness about data center gear. The biggest example is the Facebook initiated Open Compute Project, which pushes for standards-based, energy-efficient data center hardware. These companies may not want you in their data center, but they’re now willing to share some best practices about the hardware you should run in yours.

Backblaze’s 8-page RFP asks prospective data center providers what they offer in terms of physical security, and SAS 70 compliance and other regulatory certifications.

It’s an interesting document that could act as a template for other companies that are now in the market for a new data center.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• NewNet Q4: Platform mania and social commerce shakeout
• NewNet Q4: Platform mania and social commerce shakeout
• Migrating media applications to the private cloud: best practices for businesses

A year after adding water metering to its Prineville 1 data center, the company claims a WUE (water usage efficiency) rating of 0.22 liters per kilowatt-hour (L/kWh) for that building. WUE is one of several metrics formulated by The Green Grid to show just how efficient a data center really is. Since the goal is to save water, the ideal WUE value — the ratio of annual water usage to the amount of energy consumed by IT equipment — is zero.

While it is touted as an annual figure, Facebook will report its number quarterly, according to the Open Compute Project blog.

According to the post:

We think that 0.22 L/kWh is a great result, but it should be noted that the WUE concept is fairly new and, to our knowledge, no one else has publicly reported WUE yet. We hope that other companies will soon start measuring and reporting WUE so we can begin setting benchmarks for the metric and working together to find new ways to improve.

While many data centers use water to cool data center equipment, Prineville 1 features a mechanical system with a penthouse that “utilizes 100 percent outside air economization with a direct evaporative cooling and humidification (ECH) misting system. This design allows us to achieve a strong WUE,” according to the post.

The WUE rating does not factor in water used by people — for bathroom or office plumbing — but Facebook says it uses reclaimed water where possible (as well as waterless urinals) to minimize that usage. The company said WUE data for the second Prineville will come out next year when that building goes online and it will also produce WUE metrics for its Forest City, N.C. data center.

 Feature photo courtesy of Flickr user IntelFreePress

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Locating data centers in an energy-constrained world
• A near-term outlook for big data
• Migrating media applications to the private cloud: best practices for businesses

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Infrastructure Q1: IaaS Comes Down to Earth; Big Data Takes Flight
• Infrastructure Q2: Big data and PaaS gain more momentum
• Infrastructure Overview, Q2 2010

By way of background, Backblaze is a cloud storage provider focused solely on backing up lots of data for cheap (like $5 a month for unlimited capacity cheap). In order to do that, it had to build a storage system that could hold massive amounts of data without breaking the bank. As of last July, Backblaze’s architecture had evolved to a point where a 135TB pod cost less than $7,400 to build from scratch.

Understandably, the architecture generated a lot of interest from companies and organizations wanting to leverage it to soothe their own IT budgets, but none of them are Netflix. EMC’s (e emc) Pat Gelsinger said recently that the storage component of Facebook’s Open Compute Project, called Open Vault, isn’t yet ready for primetime because nobody is running — or would run — mission-critical workloads on it. That might be true of Open Vault today — the project just launched earlier this year — but it likely won’t be for long. If you consider a CDN that serves Netflix streaming video mission-critical, the criticism is already invalid for Backblaze’s designs as Netflix has adapted them.

Netflix’s 4U, 100TB server

It’s worth noting, too, that open source hardware isn’t the only piece of the stack threatening legacy storage vendors such as EMC. I’ve heard it suggested recently by someone experienced in building out large-scale cloud infrastructure that the Hadoop Distributed File System has the potential to become the default file system for large infrastructures once it works out some of the limitations around performance and availability. One of the biggest of those limitations — the NameNode– has been eliminated in the latest version of Apache Hadoop and is already integrated into Cloudera’s new CDH4 release.

Can storage deal with the open source disruption?

As with Open Compute’s effects on the server industry, though, open source storage doesn’t need to spell doom for legacy vendors if they’re willing to adapt. One reason is that, at least in the short term, there are still plenty of customers that don’t operate at Facebook or Netflix scale and can afford to pay a premium on smaller deployments that offer the features (and vendor support) those customers demand.

If the server shipments tell us anything, though, it’s that the rise of cloud computing and web giants will ultimately take a toll on the storage market, too. Fewer, but very large, customers will be responsible for a greater percentage of sales, and they won’t necessarily want all the bells and whistles that make enterprise storage products so expensive. And if VMware is correct, even mainstream enterprises will soon want to follow the examples of web giants like Google and Facebook by running relatively dumb hardware managed by really smart software.

If this scenario plays out, storage vendors will have to reassess how they deliver value and earn their money. That might mean adopting open source designs in their own gear while shifting their focus a lot more heavily toward software and services, or perhaps unlocking their storage-management software from the hardware and certifying it to run on open source gear.

Perhaps we’ll get some ideas for what the future storage and markets look like at our Structure conference June 20 and 21, where we’ll dive into the topic with Facebook’s Frank Frankovsy, Netflix’s Adrian Cockroft and VMware Steve Herrod. Whatever the case, it looks like something will have to give.

Feature image courtesy of Shutterstock user Zadorozhnyi Viktor.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Migrating media applications to the private cloud: best practices for businesses
• Infrastructure Q1: Cloud and big data woo enterprises
• A near-term outlook for big data