But after a major company-wide restructuring and the divestiture of several non-core businesses, CEO John Chambers (pictured here at D5) is finding that turning the massive Cisco ship around — something he seemed to have started two quarters ago, and which continued last quarter, isn’t coming easy.

There’s the global economy to worry about. All that messy complicated news coming out of Europe about sovereign debt and cuts in government spending around the world has a way of eating into technology budgets both at Cisco’s government customers and at its large enterprise customers.

Cisco’s guidance for the quarter ending in July was especially worrisome for investors, who promptly sent Cisco’s share price plummeting by more than 8 percent in after-hours trading. Cisco called for revenue to grow between 2 percent and 5 percent, which works out to sales in the range of $11.4 billion to $11.8 billion, well off the consensus forecast of $12 billion.

Guidance on earnings was equally disappointing. At 44 cents to 46 cents a share, the midpoint lags the consensus by two cents.

So what’s going on? I asked Chambers about it in a phone interview with AllThingD held after the conclusion of Cisco’s conference call with analysts.

AllThingsD: John, the markets clearly don’t like very much what they saw today. So, from a high level, what happened — good, bad and indifferent — with this quarter?

Chambers: The first thing from a high level is that we’re executing pretty well on our vision and strategy, and we did exactly what we said we would do. We guided for growth of 5 percent to 7 percent for the year and for the first nine months we’re at 7.5 percent [revenue]. We said profits faster than revenues, and we’re at 9.5 percent. Earnings per share increasing 13 percent year over year for the first nine months and gross margins down just 1 percent primarily on product mix. We’re winning versus our key competitors and winning at a pretty fast rate. When you’re number one or two in most product categories, holding your own in switching and making it very challenging for the Huawei’s of the world, the Junipers and Hewlett-Packards … Juniper and HP we’re pulling away from and we’ll see if we can maintain it. Huawei, for the first time we’re getting much better and competing against them and understanding their weaknesses. And if you look where we are in terms of the bigger picture, we’re in the right markets. We’re in the mobility market. We’re in video. We’re in the cloud market. We’re in the social networking segment. We’re pulling them all together, and our customers are buying the architecture in a pretty good amount. Even in service providers, where most people thought they would never move toward having preferred vendors, we’re seeing something close to that at some service providers and at many of them they are starting to think about going all-Cisco.

So on things we can control and influence I think we’re in pretty good shape. In terms of the market, I’d like to add another couple of points [of growth] in service providers, another couple of points from commercial customers. The public sector is at 3 percent. I’d take that for the year, but we think it’s going to be flat, give or take a couple points.

The issue is the enterprise. And there the problem is not that they don’t have the money or that they don’t understand that it’s important to get productivity. It’s that they’re uncertain. When they are uncertain, that’s because of economic issues primarily because of Europe. And uncertainty on government policy. Then you see people deciding not to invest. And that affects not only capital spending but jobs.

So I think the market understood what we’re saying and I think most people would give us credit for being a very good indicator of what the point in time change is. But this is not necessarily a given for what is going to happen in the second half of the year. I’m just trying to be as transparent as I know how.

The July quarter is usually your seasonally strongest. Given that your guidance was relatively weak compared to the consensus, what are we to make of the quarter coming up? Is it a secular weakness or mostly the economy? Are your competitors just taking it on the chin worse that you are?

Let’s look at Juniper. It’s down 6 percent a year and routing down 9 percent. Huawei is growing 11 percent a year, but its service provider segment is growing only 3 percent. HP’s networking business is back to the levels in their switching business to what they were when HP first bought 3Com. There’s an explosion in the data center business, it’s to the point that companies who have been there a long time like IBM or HP, we’re growing 67 percent and their servers are flat or slightly down. So the results speak for themselves in terms of what we’re doing right in some areas. But we’re learning to tie things together in a way that saves customers money, saves them time to market and allows them to achieve their business goals quicker. That is the game we’re playing for. The major thing we’re after is getting the enterprises spending again. Customers — almost uniformly — are saying that my business is okay, not great, they expect it will go up gradually, and that they’re probably going to spend more in the second half of the year than they did in the first. But immediately as a follow-up to that, they all say that’s true only if they’re not surprised by something from the economy. That’s the kind of uncertainty we’re seeing, and in talking with my peers in the industry who are in similar markets, they can finish my sentences. The question is whether it’s temporary or is it a blip? We just don’t know yet.

So given the restructuring we’ve been talking about for the last year or so, is Cisco the right size? Your overall headcount is down more than 8,000 from a year ago, but it’s up by more than 1,300 since the last quarter. Are you at the right size or are there more changes coming?

Out of the 1,353 people we added, the vast majority were either advanced services or engineers. We needed more engineers. The additions were around either building products or converting services. In terms of our organization structure, we re-did Cisco. We’ve learned from what we did well in the past, and you wouldn’t see the turnaround as quickly as we did if the structure weren’t so strong. But we needed to restructure how our customers buy, and how we build products. We needed to be nimbler and simpler in how we get decisions done. And that is a journey. In the past we tended to get a market transition, good or bad, and take off on a good one or address a bad one, and we would end up gaining market share almost always coming out of these. We’ll see if we do it again this time. But we weren’t constantly reinventing ourselves to avoid hitting the next wall or the next inflection point. That is what we’re trying to do. This is a continuous journey. While we were four or five inches around the waist, I think there’s still more work to be done in our middle levels. I think you’ll see us address that in the next year or two. Does that mean we’re going to adjust the market given that the market may have slowed? I’m not sure it has yet, we’ll know in a couple of quarters which way it’s going. The answer is, not in a major way. It’s too early to say which way this market is going. We’re not going to over-react or under-react.

You just made a major acquisition with NDS, about $5 billion. You still have a lot of cash on the balance sheet. What’s your stance on acquisitions?

Ongoing at Cisco we will do innovation through internal development, including internal start-ups, through strategic partnership, and acquisitions and intergrating all of the above. NDS is one of multiple moves that we will make, not just in the video space for us, but it was also a major cloud play for us and a major social media move if we do this right. It plays right into the sweet spot of our service providers and content providers. Our ideal target has not changed: 100 engineers with a product that is just about to come to market, where our customers say that if they were owned by Cisco they’d buy a lot of it. The $5 billion price is higher than what we’ve traditionally paid, but it’s on the order of Stratacom and Tandberg, for which we paid about $3 billion each. But our ideal target is smaller, and you’ll see us continue to be selectively active in the market.

You’re said to be heavily focused on gross margins. One point that came up on UCS: You say it’s growing like crazy, off a low base, but one of the analysts pointed out this week that it has the overall effect of bringing down the gross margin a bit. How are you addressing that?

That’s true on specifics. UCS by itself, even with a premium versus our peers, is going to be below our gross margin of 65 percent. So, by definition, as you add more of those it has a major effect on gross margin. When you combine UCS with our Nexus 2000 and 5000 switches the blended version gets the margin higher, though still not as high as the overall gross margin. Our challenge on gross margin, and the reason why we’re going to focus aggressively on each gross margin area this year, is that it’s more of a product mix issue than it is an issue of pressure on gross margins on any specific product. In terms of the base for UCS, it’s getting close to a $2.5 billion run rate and probably closer to $3 billion by now. So the base is getting larger, and in North America our market share is close to 20 percent and globally our best guess is 14 percent as best as we can tell. So that’s pretty good execution.

At this point, as he did last time we talked, Chambers asked me what song I’d pick to musically illustrate Cisco’s quarter, sticking with a tradition started a few quarters back and continued last quarter. I told him I wanted it to be a surprise, but that I think he’d like it.

This quarter, I dedicate to Cisco Ringo Starr’s “It Don’t Come Easy.” The hard work of transformation done, Cisco is finding that, despite being leaner and meaner, it has still got some way to go and finds itself in a tough market. In the video below, Ringo performs with fellow Beatle George Harrison at the 1971 Concert for Bangladesh. As everyone at Cisco knows, nothing worth having comes easy.

The consensus of Wall Street analysts has Cisco reporting $11.6 billion in sales, and 47 cents per share of earnings.

The big question, writes analyst Sanjiv Wadhwani of Stifel Nicolaus in a note to clients on May 7, will be around margins. Last quarter, Cisco gave guidance for gross margins — a key measure of profitability — with a narrow range of 61.5 percent to 62 percent, while operating margins were guided to the range of 27 percent to 28 percent. Wadhwani thinks that guidance may stand up as pressure on Cisco’s supply chain from the Thailand flooding, favorable pricing on switching products, and a less-aggressive posture from Hewlett-Packard’s networking arm are all providing a little breeze at Cisco’s back.

Yet one product in Cisco’s stable may, in success, be hurting margins overall: Cisco’s Unified Computing and Servers line (UCS) tends to carry a lower gross margin, Wadhwani writes, and so may eat into its overall gross margin. The product line — which combines computing, storage and networking into a single product offered to corporate and service-provider data centers — had 10,000 customers worldwide last quarter, and was showing “positive momentum” in Wadhwani’s checks. “Overall, we believe that there is intense focus on margins internally, which should allow the company to report an in line margin quarter,” Wadhwani wrote.

Cisco has been operating “with more confidence and aggressiveness with its refreshed product line, making it tougher for competitors,” writes Shaw Wu of Sterne Agee in a note to clients issued May 7, and has been a lot of the reason that Juniper and Alcatel-Lucent have missed expectations recently. He expects Cisco to give guidance for the quarter ending in July that’s more or less in line with consensus expectations. He also sees Cisco benefiting from Apple’s next iPhone: “We believe this could mark the fourth quarter in a row where Cisco does not guide down expectations further building investor confidence. We see Cisco benefiting in the second half of 2012 from the continued build-out of 4G LTE wireless infrastructure ahead of the iPhone 5 refresh likely in the September-October time frame.”

I’ll be covering Cisco’s earnings announcement later today. Having dedicated songs to CEO John Chambers two quarters in a row now, I’m going to have to scramble to see what song fits today’s results, because I just know he’s going to ask.

A sister site to the company’s massive data center in Maiden, N.C., this facility will be significantly smaller — at least initially. Currently, Apple is building a 10,000 square-foot data center on the 160 acres it recently purchased in Prineville, Ore. But that’s clearly just the beginning.

As part of its deal with Prineville and Crook County, Apple was granted a 15-year property-tax break. In return, it is to invest $250 million in the data center, and hire at least 35 employees to run it.

“We’re looking forward to joining the community of Prineville with our new data center,” Apple spokeswoman Kristin Huguet told AllThingsD. “In addition to the hundreds of employees at our retail stores, we will be hiring dozens of people and bringing hundreds of construction jobs to the area.”

Once finished, the Prineville facility is expected to be even more environmentally friendly than the one in Maiden, running on 100 percent renewable energy.

Not very, according to Greenpeace, which describes it as among “the dirtiest thing[s] on the internet.” In a report released today, the environmental organization slammed Apple as among cloud computing’s worst eco-offenders (along with Amazon and Microsoft), claiming the company relies too heavily on coal to power its data centers and has fallen short on its efforts to source clean energy. Coal-based energy makes up 55 percent of Apple’s data center power, Greenpeace claims.

“Apple right now is falling behind companies like Google and Facebook, who are taking a leadership role on this issue,” said Greenpeace spokesman Dave Pomerantz. “It’s a shame that a company that built its reputation on thinking differently is now behind the curve.”

But according to Apple, it’s not the one behind the curve here. And, more to the point, it says Greenpeace has gotten its numbers wrong.

According to Apple, its Maiden, North Carolina, data center requires just 20 megawatts when it’s running full bore, and well over half of that energy will come from renewable sources as soon as the 171-acre solar array it is building nearby is finished.

But according to Greenpeace, the Maiden facility requires 100 MW of power running at full capacity, of which renewable energy powers 10 percent and coal 55.1 percent.

Odd, considering Microsoft’s Quincy, Washington, data center, which is exactly the same size as Apple’s, only requires 27 MW, and its Chicago facility, which is 200,000 square feet larger, requires 60 MW.

How is it that a fully operational 700,000 square foot data center demands significantly less power than a not-yet-finished 500,000 square foot one?

And how is it that Apple’s Maiden data center is running on 55.1 percent coal, when publicly available records show that 46 percent of the power Duke Energy supplies it with is coal-fired?

That’s not clear, and Greenpeace initially didn’t have much of an explanation. The organization referred me to the EPA’s 2010 eGrid data for North Carolina on which its report is based. But that draws from data collected in 2007. As for why Microsoft’s fully operational 700,000 square foot data center requires 40 MW less energy to run than Apple’s not-yet-fully operational 500,000 square foot data center, Greenpeace referred my question to Microsoft.

Apple, on the other hand, was more than happy to talk up its power plans.

“Our data center in North Carolina will draw about 20 megawatts at full capacity, and we are on track to supply more than 60 percent of that power on-site from renewable sources, including a solar farm and fuel cell installation which will each be the largest of their kind in the country,” Apple spokesperson Kristen Huguet told AllThingsD. “We believe this industry-leading project will make Maiden the greenest data center ever built, and it will be joined next year by our new facility in Oregon running on 100 percent renewable energy.”

In response to that response, Greenpeace said that Apple’s information just doesn’t jibe with the factors of investment and facility size used in making its estimates.

UPDATE: Good analysis of Greenpeace’s dubious math by Rich Miller over at Data Center Knowledge.

Well, today we heard from another bank, and its opinions about Intel’s Romley chip and what it means for data center spending couldn’t be more different. Chris Whitmore, an analyst with Deutsche Bank Market Research, published a note to clients today, arguing that Romley will indeed spur a new round of spending in corporate data centers, and that it will have an equally strong secondary effect on the fortunes of enterprise storage companies, specifically EMC and NetApp.

One of the things that Romley will encourage, Whitmore writes, is a growth in the density of virtual machines running in each server. (Remember that, more often than not, a physical server is virtualized or subdivided into many virtual servers, allowing each machine to act like several machines.) More virtual machines allows you to consolidate your physical machines and add more in the same footprint if you want, which in turn means more computing work getting done overall. Whitmore estimates that, in general, data centers will boost their workloads by 20 to 25 percent by the end of next year.

Roughly 26 percent of Romley chip purchases will be used in these virtualized environments, Whitmore estimates. And that tends to spur demand for storage to support the virtual machines. In fact, the growth of terabytes worth of storage products shipped mirrors closely the unit growth of servers. (See the graphic, below, which I screen-grabbed from the report; click to see it bigger.) In short, it’s good news for NetApp and EMC. Whitmore says both are taking share from other vendors, including IBM, Hewlett-Packard and Dell, with sales growing at north of 20 percent a year — a growth rate that’s higher than that of the overall market, which grew 14 percent last year. He rates shares of both EMC and NetApp a “buy,” with price targets of $35 and $60, respectively.

Great news for EMC and NetApp, but what does it mean for Intel? Whitmore says to expect a mixed bag. Companies wanting to boost their use of virtual machines will be buyers. Companies that aren’t into virtualization so much, maybe not. “We believe our estimate of x86 servers shipped into virtual environments growing from 21 percent in 2011 to 26 percent in 2013 could prove conservative,” Whitmore writes. “As a result, although we expect Romley to have a relatively muted impact on overall server unit demand, we do expect it to drive another leg of virtual machine growth.”

Wyse used to pop up on my radar screen back in the late 1990s when there was a sudden craze in what was then called “thin clients.” The basic idea behind a thin client is to put a keyboard and a display on an employee’s desk, and have all the computing horsepower for that individual workstation — all the data and software that it uses — reside on a remote server. The advantage is that thin clients are cheaper to operate, easier to maintain and upgrade, and also easier to secure and control.

With cloud computing being very fashionable right now, it’s no surprise to me that Wyse has adapted how it describes its mainline specialty. Another phrase that gets passed around quite a bit is “virtual desktops.”

For its part, Dell is on a long-term campaign to diversify its business away from the personal computer business it revolutionized in the 1990s, and to move into more profitable areas of enterprise IT sales and services. It’s also embracing cloud computing in a big way.

Since Wyse is private and apparently not huge, Dell isn’t disclosing how much it is paying. Wyse is big enough to have sold more than 20 million thin-client systems worldwide; it claims that 200 million people touch its products every day. It sold a million units last year, and has a strong presence in the health care and financial services sectors, both of which have a lot of appeal for Dell generally.

Update: So here’s an informed estimate concerning how much Dell paid. Brian Marshall of ISI just issued a note to clients pegging the price at between $350 million and $400 million. He reckons that’s about level with Wyse’s sales of about $375 million over the prior year. Hewlett-Packard paid about $220 million for Wyse rival Neoware in 2007 at a time when its annual sales were in the $75 million range, he says.

Here’s Dell’s announcement:

Dell Announces Intent to Acquire Wyse Technology
-Extends Dell’s desktop virtualization capabilities and drives attachment of enterprise solutions, including servers, networking, storage and services
-Strengthens Dell’s strategy to offer customers innovative, end-to-end IT solutions from the edge to the core to the cloud

Dell today announced it has signed a definitive agreement to acquire Wyse Technology, the global leader in cloud client computing, to significantly extend its desktop virtualization offerings. The addition of Wyse will expand Dell’s desktop virtualization capabilities and provide new solutions and services opportunities for the full range of Dell’s enterprise offerings.

In some environments, a virtual desktop allows enterprises to more efficiently and securely manage their users and end point devices. With this acquisition, Dell expands its enterprise solutions portfolio and offers customers an ever broadening array of tailored solutions to meet their needs.

“The total market for desktop virtualization solutions should continue to see strong growth globally, with the larger revenue and margin opportunities coming from the datacenter infrastructure, cloud and services offerings that are tied to thin client and desktop virtualization technology sales,” said Matt Eastwood, Group Vice President, Enterprise Platform Research, IDC. “Thin client and desktop virtualization solutions typically drive high attach rates to data center solutions, including servers, networking, storage and services. The end-to-end datacenter infrastructure stack for these solutions is expected to exceed $15 billion by 2015.”

Wyse Technology Leadership
Wyse, ranked No.1 in thin client unit shipment volume on 4Q 2011[1], has shipped more than 20 million units worldwide and has over 200 million people interacting with its products each day. The company has more than 180 patents, both issued and pending, covering its solutions, software and differentiated intellectual property.

The Wyse solutions portfolio includes industry-leading thin client solutions with advanced management, desktop virtualization and cloud software.

Cloud clients: Wyse offers a wide selection of secure, reliable, cost-effective thin and zero clients designed to easily integrate into any virtualized or web-based infrastructure, while meeting the budget and performance requirements for any application.
Cloud software: Technology powering a new world of cloud connected smart devices.
Management software – Secure, easy and scalable remote device management for the extended enterprise.
Virtualization software – The best user experience with Microsoft, Citrix and VMware virtual desktop infrastructures.
Mobility software – Secure mobile connectivity to your personal, private or public cloud for mobile devices.
Services: Wyse offers a wide range of cloud computing services to complement its cloud clients and software solutions. Services are available for specific or on-going engagements.

Wyse Technology, founded in 1981, is headquartered in San Jose, Calif. and serves customers in more than 50 countries around the world. Wyse has more than 3,000 resellers and shipped more than 1 million units in 2011.

The transaction was approved by the board of directors of each company and is expected to be accretive to Dell non-GAAP earnings in the second half of its Fiscal Year 2013. Additional terms of the transaction were not disclosed. The transaction remains subject to customary conditions and is expected to close in the second quarter of Dell’s FY13.

Quotes
“Desktop virtualization can help organizations streamline IT management, improve productivity and security, and increase cost efficiency for discrete workloads or usage scenarios,” said Jeff Clarke, president, End User Computing Solutions at Dell. “The Wyse Technology desktop virtualization capability complements Dell’s strongest-ever device and computing solutions portfolio, and strengthens our position in offering customers among the broadest set of computing choices from the edge to the core to the cloud.”

“The combination of Wyse and Dell provides us with tremendous growth opportunities for our core desktop virtualization business, helps us expand into new and fast-growing market segments including mobility and cloud computing, and provides us with reach and scale we did not previously have,” said Tarkan Maner, president and CEO, Wyse Technology. “We believe that taking this step with Dell is a very natural progression for our business and offers our customers many great advantages not available to them today.”

An analyst call with Jeff Clarke, president, Dell End User Computing; Dave Johnson, senior vice president, Dell Corporate Strategy; and Tarkan Maner, president and CEO, Wyse Technology; will be webcast live today at 8:45 a.m. Central Time and archived at www.dell.com/investor.

About Wyse Technology
Wyse Technology is the global leader in Cloud Client Computing. The Wyse portfolio includes industry-leading thin, zero and cloud PC client solutions with advanced management, desktop virtualization and cloud software supporting desktops, laptops and next generation mobile devices. Wyse has shipped more than 20 million units and has over 200 million people interacting with their products each day, enabling the leading private, public, hybrid and government cloud implementations worldwide. Wyse works with industry-leading IT vendors, including Cisco®, Citrix®, IBM®, Microsoft, and VMware® as well as globally-recognized distribution and service providers. Wyse is headquartered in San Jose, California, U.S.A., with offices worldwide.

About Dell
Dell Inc. (NASDAQ: DELL) listens to customers and delivers worldwide innovative technology, business solutions and services they trust and value. For more information, visit www.dell.com.

The data center, which is slated for completion in early 2013, will be the third of its kind in Asia, joining two others that will be built in Hong Kong and Taiwan. Google has pledged an investment of $300 million in its Hong Kong data center and more than $100 million in Taiwan.

Read the rest of this post on the original site »

HP will answer some of those questions today, with a big announcement of a new enterprise hardware, software and services platform at a company event in Vienna. HP calls this its Next Generation Information Platform: IDOL 10. Practically speaking, the suite includes Autonomy, with its software smarts around finding meaning in unstructured data as varied as TV interview transcripts or chains of email communications; and the analytical muscle of Vertica, a company HP acquired in February.

The way HP sees it — and, to be fair, it’s not the first company to make this kind of argument — the ratio of data that businesses are creating to what they actually use productively is pretty big. Only 15 percent of that information is neatly organized into the rows and columns of a traditional relational database, HP argues, leaving a lot more information — fully 85 percent — that would be useful if you could only capture it, determine its meaning, and analyze it: Video, audio, email, texts, social media, meeting notes. Add to that the explosion of other real-world information gathered from sensors and other measuring devices, and it gets even more complex. It’s a concept that HP is calling “information optimization.”

Mike Lynch, the former CEO of Autonomy, is expected to speak at today’s announcement event in Vienna. And if you happen to be there, you’ll probably also hear from Jan Zadak, HP’s executive vice president for global sales. I talked to Zadak last week, and he filled me in on the news.

It’s all about capturing that “lost” 50 percent to 85 percent of information and making it useful to businesses, Zadak told me. “What we’re trying to do is address the notion of how to enable enterprises around the world to harness the potential of data,” he said. “Traditionally, what you would do is run your relational databases and put all your information in columns and rows, and then you would crunch your data. That is what you would typically do to address what we call ‘structured data’ — everything we can pull into a structural database. But 85 percent of information is unstructured — images, charts, emails, tweets. All types of conversation.”

HP has researched the issue a bit, and reckons that more than half of the information produced at your average business remains unconnected, undiscovered and unused. It’s sort of like that old myth that people use only a small percentage of the capacity of their brains — except it’s probably true. Most business decisions are made with incomplete information, and who wouldn’t want to try to to address that shortcoming?

There’s a lot to the announcement from a product perspective. I won’t go into everything, but here are some highlights. There’s an HP social media solution, intended to help companies mine useful information out of feeds on Twitter and Facebook and elsewhere, to improve customer experiences and make sure the brand isn’t suffering because of some over-eager or ticked-off commenter.

And there’s a backup system, sold under HP’s StoreOnce brand, that came directly from work done by HP Labs. The idea is to store data once, rather than make unnecessary, redundant copies that take up precious storage space.

The announcements also contain some new dashboarding and performance monitoring tools that are aimed at helping IT execs better monitor what their systems are doing and whether or not they’re performing as efficiently as they should be.

Back to Zadak for a moment. As head of global sales, he’s the one who has been charged by CEO Meg Whitman with providing the “single face of HP” to enterprise clients. To that end, I asked him what his priorities are for 2012: Apparently that means spending a lot of time developing the sales force. He took over HP’s Sales University when Thomas Hogan left the company in April, and so far “several thousand” HP employees have been through it. Expect more investments around supporting the sales force generally. Given that so many business segments saw declines in sales in the most recent quarter, that doesn’t seem like a bad idea.

The Charlotte Observer reports that Apple has been issued the permits necessary to prepare the 171-acre parcel of vacant land adjacent to the data center for a new construction project. The name of that undertaking: “Project Dolphin Solar Farm A Expanded,” a riff on the data center’s original code name.

Details beyond that are slim, but sources close to Apple confirm that the Catawba County solar project is indeed under way.

Which makes perfect sense, really. A solar farm would be in keeping with Apple’s commitment to renewable energy. Apple’s manufacturing site in Cork, Ireland, is entirely wind-powered, and its Elk Grove, Calif., and Austin, Texas, facilities are now powered by 100 percent renewable energy resources, according to the company’s 2011 Facilities Report Environmental Update.

The Project Dolphin Solar Farm would advance this commitment even further by establishing an on-site resource for renewable energy generation and setting the bar a bit higher for an industry whose pollution-limiting efforts extend no further than purchasing renewable energy credits.

Built from the ground up by Google, these three facilities will be the first in the Asia-Pacific region to be fully owned by the company. Google expects them to be operational in one to two years, barring any major hiccups.

The investment comes as demand for Internet services in the region soars and competition between local players heats up.

“More people are coming online every day in Asia than in any other part of the world, so locating data centers here is an important next stage of Google’s investment in the region,” Google said in a statement.

Along with other Internet giants, Google has spent large amounts on building server capacity, to allow users remote access to stored files and programs through so-called cloud computing. In the past two years, it has converted an old paper mill close to the small town of Hamina, on the Baltic Sea coast of Finland, into a massive data center.

The mill was built to produce paper for magazines and newspapers, but the paper industry has seen demand for its products decrease as competition from online media has left many newspapers struggling. Paper manufacturer Stora Enso Oyj closed production at the mill in 2008, and Google saw the opportunity to fill the structure with the technology that contributed to the demise of papermaking in Hamina. A main attraction for the U.S.-based company was the cool Finnish climate.

Read the rest of this pst on the original site »

Next we saw Violin Memory — which makes flash-based storage arrays that are intended to make enterprise applications run faster — land $40 million in venture capital funding.

Now we see a third player entering the “flash madness” narrative. Pure Storage is coming out of stealth today, announcing its plans to sell flash-based storage arrays. It is also announcing that it has landed a $30 million C-round led by Redpoint Ventures, with Samsung Venture Investment joining. (Yes, that would be the venture capital arm of the South Korean electronics giant that happens to be the world’s biggest manufacturer of flash memory.) Greylock Partners and Sutter Hill Ventures also participated. The latest round brings Pure’s total funding raised to date to $55 million.

So what is Pure Storage all about? I met up with CEO Scott Dietzen last week and got the download.

The fundamental problem with enterprise storage is that hard drives just can’t keep up with everything else that’s gotten faster in the data center. Flash memory is fundamentally faster, it uses less energy and it takes up less space. We all know this.

The problem with flash is that it has always tended to be more expensive than hard drives. Today, you can buy a one terabyte hard drive for $100 or less. But just try getting that same amount in flash memory and see if the price isn’t, well, a lot higher.

The same principles apply in the data center. CIOs would love to convert to flash-based systems, as long as they’re reliable and affordable and work with the applications and other hardware they already have.

Pure Storage is essentially promising to deliver just that, Dietzen says. The company’s first product is an all-flash storage array that is 10 times faster and 10 times smaller than hard-disk-based systems. It’s called the Pure Storage FlashArray, and it is being aimed at mainstream enterprises in a manner that’s easy to deploy.

Pure’s founders are John Colgrove — one of the founding engineers at Veritas, now part of Symantec — and John Hayes, a founding engineer at Bix, which was ultimately swallowed up by Yahoo. Dietzen hails from Yahoo as well, by way of its acquisition of Zimbra, where he was CTO.

An early key hire was Michael Cornwell, who was lead technologist for flash at Sun Microsystems (now part of Oracle). Cornwell also worked at Apple, where he was Manager of Storage Engineering for the iPod division, and oversaw that product’s transition to — you guessed it — flash memory. Remember the first iPod nano? That was his baby.

Another key name: Greylock venture partner Frank Slootman, the former CEO of Data Domain, is on Pure’s board.

So what’s so special about a storage array built on flash memory? “Disks get slower every year,” Dietzen says. “Intel says processors have gotten 175 times faster over the last 15 years.” Disks just keep getting more data packed onto them, which doesn’t really make them any faster. The mechanical arm inside the disk that grabs data from the platter really can’t go much faster. “Disks today are comparably slower than tape was 15 years ago,” he says.

This creates a problem. Storage needs are going up, but hard drives are slowing data centers down, preventing them from reaching their full potential. It’s only because of cost — about $5 per gigabyte — that hard drives are still appealing. Enterprise-grade flash, on the other hand, tends to cost $40 to $100 per gigabyte, and because flash is historically less reliable, you have to buy double what you really need.

Pure’s play is to get over the cost hurdle. Dietzen says the company can get the cost down to $5 per gigabyte and less.

How does it do that? By reducing the amount of data you actually store. What happens in enterprise environments is that various bits of data get copied and recopied, over and over. Imagine a big filing cabinet with 50 copies of each document scattered around in different folders, when you really only need one. Suddenly the size of that file cabinet need not be so big. The same applies in data storage: Why bother having 10 copies of the same block of data, when one or two will do?

Using a technique known as deduplication, a system can eliminate all those unneeded copies and thus streamline the whole operation. Deduplication, combined with compression, was the primary principle behind Slootman’s Data Domain, which is now part of EMC.

But deduplication is expensive on hard drives, and really doesn’t make sense. Because the mechanical arm in a hard drive is always searching around for where its next needed block of data is to be found, if you employ deduplication, you end up with a bunch of reference signs telling the arm where to go, Dietzen says. The end result is that the disk has to spin more, not less. Flash memory chips don’t have that problem. “We make that process fast, because there’s no performance hit to the deduping process,” he says.

On top of that, Pure has created some algorithms that make the process a lot more granular than on hard-disk-based systems, by working with smaller disk-sector sizes. How small? He wouldn’t say exactly.

Unlike other storage companies — like, say, EMC — Pure’s array, Dietzen says, is built from the ground up for running flash. “The disk-centric companies are slotting flash into places where disks used to be, but they’re not changing the software to take advantage of the flash, to protect the flash from uneven wear and other things.”

A few early companies have tried the hardware, among them the law firm of Fenwick & West, whose CIO Matt Kesner is quoted in Pure’s press release as saying that the data used for various workloads was reduced from 50 to 90 percent.

One key thing that’s going on in the data center these days is virtualization — running several virtual computers within one single physical computer. When you run a lot of virtual machines, you have a lot of data that, like the paper in that big file cabinet, is essentially the same. Dietzen says that Pure’s flash array is able to eliminate a lot of that data. “Even if those virtual machines are a mix of Windows and Linux, there are a lot of commonalities between them,” he says. It’s not uncommon to see the data footprint for virtual machines reduced by a factor of 15 or 20 to one.

And that has caused some interesting reactions among early customers trying out the array. “Some people try it and are shocked when they put 15 terabytes on it and see there’s only one terabyte and think we’ve lost a lot of their data,” Dietzen says. “It’s a little scary at first, but then they run all their workloads and see all the data is there.”

Having closed on Friday at $14.97, the price of Cisco shares is nearing its lowest point in about five years (Cisco hit $14.18 in March of 2009). CEO John Chambers has blamed toughening competition in its main networking business, lower profit margins resulting from a product transition and a drop in government spending for many of its troubles. A significant restructuring is coming some time before the end of the summer that will combine offering retirement packages to eligible employees and laying off others. It’s also possible that other parts of Cisco’s business may be sold, spun off or shut down.

Chambers’ plan is to trim Cisco’s operating costs by $1 billion a year. Meanwhile, the video business that Chambers constantly talks about is starting to get interesting and competitive, and other products, like its blade servers, are starting to show some traction. And by the way, the Internet isn’t exactly getting smaller, you know.

Whatever it is, it better happen soon, Wu says, because investors are getting impatient. “In sum, we believe Cisco is fixable and not structurally flawed, but admit we need to see more dramatic steps be taken,” he writes.

While some investors have been calling for Chambers to step down, Wu isn’t with them. “From an investor standpoint, most believe that Cisco will be very difficult to turn around and that a management change is needed. While we do not believe that John Chambers needs to go, as we believe he has proven to be one of the greatest managers and visionaries of the modern era, we do believe he needs to make bigger moves than what has been done so far.” One suggestion? Boost the dividend to 3 percent from its current 1.6 percent.

Also? Cisco may have to take some “bitter medicine” on the price of its switches and other networking gear. Many Cisco customers and its channel resellers told Wu that Cisco’s prices are too high when compared to competitors, and that it may be pricing itself out of the market. “Many believe that Cisco still deserves a premium, but 50 to 100 percent seems a bit excessive,” he writes. “Because of this, we believe Cisco will likely need to take the bitter medicine of lower gross margin for longer-term good.” He cut his assumptions on Cisco’s gross margin accordingly from about 62 percent to a little higher than 55 percent.

That said, most of the the bad news at Cisco is priced in, making its depressed price a fair buying opportunity, he says. “We believe the Cisco story is getting better, and we’d rather be a buyer at these depressed levels than wait for obvious evidence of improvement. By then it may be too late.”

Wu still rates Cisco shares a buy, with a price target of $25, which is a drop from his previous target of $29. He also lowered his fiscal year 2012 estimates on Cisco’s revenue and per-share earnings to $45.9 billion and $1.50 per share, from $46.5 billion and $1.80 a share. The new target price represents a multiple of 12.5 times Cisco’s projected calendar year earnings of $1.62.

Wu also thinks Cisco will start setting more realistic expectations going forward, and back away from its projecting a long-term annual growth rate of 12 to 17 percent, which, he says, “no one believed anyway.”

Given all that, Cisco’s close to turning the corner, he says, though it won’t happen right away. “We realize that Cisco may take a few quarters in fixing itself, but we believe management will make the right moves in restoring investor creditability.”

(Image borrowed from the 1938 Warner Bros. classic animated short, Porky In Wackyland.)

To coincide with Earth Day, which apparently was today, Google decided to give a little peek inside its data center in Moncks Corner, South Carolina. Google would like you to know that its data centers are very energy efficient, as you’ll learn toward the end of this segment.

Google probably also wants to take advantage of the fact that it scored higher that Apple on the Greenpeace Dirty Data report because its facilities rely less on coal-generated power than those of Apple, Facebook, Hewlett-Packard, IBM or Twitter. But the environmental stuff is saved for the end.

And if this isn’t enough data center fun, you might have missed my tour of the Lucasfilm data center last month, the one that when I last saw it, was actively cranking on the bits behind all the special effects for the summer blockbuster movies. Hey, Memorial Day is about five weeks away!

The bulk of the video focuses more on how Google protects data belonging to its customers with multiple layers of physical security (hint: don’t even think of trying to sneak in), and destroying hard drives when they wear out. After watching it twice, what I really want to know is this: Where can I get myself one of those crushers?

(Fort Knox Image via Wikimedia Commons)

Mark Zuckerberg up first: calls this the Open Compute Project.

It’s easy to think that the operations is just kind of this cost part of the business, but really products and operations fundamentally are linked, he says.

Zuckerberg is pandering to the infrastructure crowd in attendance by talking up how essential data centers are. Adding capacity improves the Facebook experience by enabling better “type-ahead” features when users search, and helps speed up the site. And these challenges are the same for partners too.

What the mass manufacturers put out wasn’t what we needed for social apps, so we made it more custom, Zuckerberg says. Now Facebook wants to share its improvements with everyone else in the industry so they can improve too. “We’re not the only ones who need the kind of hardware that we’re building out.”

Jonathan Heiliger, VP Technical Operations, up next. He points to a rack of servers behind them, says “they’re real Facebook servers, or rather, real Open Compute servers.”

Facebook designed and built its own data center in Prineville, Oregon over the last year and a half, and tailored the servers there to work on its software, Heilinger said. The goal was to be as efficient as possible. The industry average power effective usage is 1.5, average for leased data centers is 1.4-1.6, but Facebook has got it down to 1.07. Across the table from me, a hardware guy mutters, “That’s really good.”

The data center design team is only three people, Heilinger said, because it was built on top of other people’s work, so now the company wants to give back by sharing its specifications and design documents.

The servers Facebook built are 38 percent more energy efficient and cost 24 percent less, according to Heilinger.

Another Facebook data center guy is now digging into the specifics, but I already have a statement from Greenpeace in the inbox about how Facebook’s announcement isn’t good enough.

“It’s commendable that Facebook is working to increase the energy efficiency of its business, and specifically its data centers–an area of neglect for many years. But as the global warming footprint of the IT industry, and Facebook specifically, continues to grow significantly, a focus on energy efficiency alone will only slow the speeding train of unsustainable emissions growth. Efficiency is simply not enough.

“If Facebook wants to be a truly green company, it needs to reduce its greenhouse gas emissions. The way to do that is decouple its growth from its emissions footprint by using clean, renewable energy to power its business instead of dirty coal and dangerous nuclear power.”

Some of the data center design innovation includes the voltage of the electrical distribution system, localized power supply and a ductless evaporative cooling system (no air conditioning at all).

Now Amir Michael is describing and demoing Facebook’s redesigned servers and racks themselves. Facebook motherboards don’t usually need to be expanded, so that plastic and extra size was removed, for example. And servers are mounted on shelves rather than traditional rails.

He tells an endearing story about choosing to use blue LEDs to light the data center rather than green ones even though they were more expensive, concluding that the end result looked “so sweet.”

Now Michael and Heiliger are narrating a video describing their accomplishments and their commitment to openness, complete with uplifting music and camera shots that lovingly pan and linger over racks of servers.

“It’s time to start treating data centers like Fight Club, and really demystify the software and hardware,” said Heiliger.

Om Malik of GigaOM (and my old boss, hey Om!) is brought up to moderate a panel of experts from companies like Zynga, Intel and Dell to talk about how the Open Compute Project affects the industry.

“We commend Facebook for taking this step,” says Zynga infrastructure head Allen Leinwand, who adds he is “definitely considering” using it.

Here’s a link to the YouTube post of that video I described.

And here‘s a site dedicated to the Open Compute Project.

As the testimonials continue at the event, here’s one sent via email: Forrester Analyst Rich Fichera comments,

“The results speak for themselves – a high volume data center with a claimed PUE of 1.07, certainly one of the most efficient large data centers in the world,” writes Fichera. “What’s more valuable is Facebook’s decision to publish its server, rack and power specifications as part of the ‘Open Compute Project,’ making these specifications available to users and vendors in the hopes of creating an ecosystem around these stripped down cost optimized servers and associated infrastructure.”

Back at the event, Rackspace Chairman Graham Weston says, “We’re seeing the emergence of a new stack: data center, server and software working together,” says .

Weston said Rackspace will incorporate Facebook’s innovations in future data centers, in order to benefit from the dramatic decrease in energy costs.

Facebook also sent, via email, a response to Greenpeace’s statement:

From the beginning, we realized that the biggest impact we could have is focusing on efficiency, both for our own operations and the world. The Open Compute Project has the potential to save many times the energy that Facebook will ever use. We sent information on OCP to Greenpeace earlier today and we’ve offered to answer their questions. We hope Greenpeace “friends” the Open Compute Project and we encourage others to do the same.

Jason Waxman of Dell tries to put the announcement in context, “It’s putting something viable on the map that gives a clear direction on efficiency,” he says. Forrest Norrod of Dell adds that the demystification of data centers is “seminal.”

OK, everyone’s looking down at their phones at this point. Time to wrap it up, Facebook!

Please see the disclosure about Facebook in my ethics statement.

NetworkEffect will be at Facebook headquarters in Palo Alto, Calif. on Thursday morning for a press conference that promises “a behind-the-scenes look at the latest technology powering Facebook.”

Asked for detail, a Facebook spokesperson said the event would cover “Facebook’s hardware and software technology.”

With that as a hint, there are some infrastructure and operations deadlines on the horizon for the company: first, the planned opening of its Oregon data center this spring. And second, the Earth Day deadline Greenpeace has proactively declared for Facebook to stop depending on coal to power its service.

Facebook, in fact, told a local TV news reporter the Oregon data center has been built to be especially green, with recycled and local materials, and a system where one out of every 1.15 watts in the system actually powering computer infrastructure, rather than the typical one out of 1.8.

Will that be tomorrow’s main news topic? We’ll have to see. Tune in to NetworkEffect around 10 a.m. PT for coverage, or watch the press conference on Facebook’s live channel.

Please see the disclosure about Facebook in my ethics statement.

The order is said to coincide with the forthcoming release of a new product that Isilon is expected to announce next week.

So huge an order for data storage would coincide with the construction of Apple’s huge data center in Maiden, N.C., and that’s expected to be the hub for a new version of iTunes that relies more on storing media in the cloud and less on using its customers local hard drives.

If you have trouble getting your head around the petabyte, the fine folks at another EMC unit, the backup service Mozy (soon to be a unit of VMWare) produced this fascinating graphic. As they tell it, one petabyte is enough to store more than 13.3 years worth of HD video, meaning 12 petabytes would be enough to store nearly 160 years worth.

The scale of the storage infrastructure, if true, would amount to another potentially intriguing clue to the environment Apple is using inside its data center. Previously it had disclosed in job ads on its Web site that its hardware there will include a mix of systems running Mac OS X, IBM’s AIX, Oracle’s Sun/Solaris, and some Red Hat Linux-based machines.

PREVIOUSLY:

• Apple’s Area 51: The Truth Is Out There
• Apple’s N.C. Data Center Intended for iTunes, MobileMe
• Apple Owns Another 70 Acres Near NC Data Center
• Was Apple Planning on Doubling Its North Carolina Data Center All Along?
• Apple Reaching for the Cloud With MacBook Air and N.C. Data Center
• That’s Apple’s New Data Center? Where’s the Giant Glass Cube?

As part of my recent visit to ILM, where first I interviewed CIO Kevin Clark and then toured the ILM data center, I got a look behind the scenes of “Rango,” the tale of a pet chameleon who is unexpectedly stranded in the desert and then finds his way to a tiny Wild West-like town called Dirt.

Dirt doesn’t exist. There’s no movie set to visit to plan shots as you might do with a set for a live-action movie. So the folks at ILM created a virtual version of the town–and the interiors of all its buildings–to help Verbinski and his team map out how they wanted to shoot each scene.

The mock-up of the town of Dirt was created inside an empty studio at ILM, with numerous cameras pointing into the center from the walls and ceilings, and as you’ll see in the video below, a director can “see” it using a digital tablet that presents a scene as it would appear if you were a character walking around in it. ILM’s Colin Benoit, who was layout supervisor on “Rango,” and Michael Sanders, digital supervisor on the film, were my tour guides. The video ends with a clip of the movie, where you see how the technology comes into use. Enjoy.

]]> http://allthingsd.com/20110328/a-virtual-tour-of-the-town-of-dirt-from-the-animated-film-rango/feed/ 0 http://allthingsd.com/20110325/lucasfilms-data-center-and-an-encounter-with-the-real-death-star-video/ http://allthingsd.com/20110325/lucasfilms-data-center-and-an-encounter-with-the-real-death-star-video/#comments Fri, 25 Mar 2011 18:00:44 +0000 Arik Hesseldahl http://newenterprise.allthingsd.com/?p=4362 Wednesday I introduced you to Kevin Clark, master of all things data at Lucasfilm and Industrial Light and Magic. Today, as promised, I’m taking you into the data center Clark commands.

Shortly after my chat with Clark, ILM’s publicist Greg Grusby ushered me into the room, where, as you’ll hear, the roar of air conditioning units cooling all the machines makes it hard to hear much else.

The room is 10,500 square feet and contains the systems running all the things that you’d find in pretty much any other corporate data center for things like email and Web service. StarWars.com is hosted in the room, for example. But I was more interested in the machines used to produce the crazy cool visual effects. As Clark mentioned in our chat, ILM is using almost a full petabyte of storage–or nearly 1 million gigabytes–and for that it relies upon NetApp appliances.

I lingered over an older rack of machines nicknamed the Death Star. These are the old customized Racksaver servers, each containing a single AMD Athlon processor and 2 gigabytes of memory, that were used to render imagery for “Attack of the Clones” and “Revenge of the Sith.” They’re also now considered so old that it’s not cost-effective to run them any longer.

Then we saw the new machines: The IBM racks, which to me actually look like they belong on the Death Star, are a mix of LS22 servers with AMD processors and newer HS22‘s with Intel processors. They average 32GB of memory per blade and currently cranking away on several movies. If you’re among those eagerly awaiting the release of films like “Cowboys and Aliens,” “Pirates of the Caribbean 4,” “Super 8,” “Transformers 3,” “Battleship” and “The Avengers,” every blink of the lights on the Bladecenter brings them one step closer to completion.

We’re not done with the tours. Next up, I’ll take you on a walking tour of the Town of Dirt seen in the animated feature “Rango.” And before we’re done with all this, I’ll tell you what ILM considers its “secret weapon,” at least from a computing perspective. Enjoy the video.

(Image via Wookiepedia)

]]> http://allthingsd.com/20110325/lucasfilms-data-center-and-an-encounter-with-the-real-death-star-video/feed/ 1 http://allthingsd.com/20110323/apple-data-center-theories/ http://allthingsd.com/20110323/apple-data-center-theories/#comments Wed, 23 Mar 2011 11:50:13 +0000 John Paczkowski http://digitaldaily.allthingsd.com/?p=59016 Scheduled to go live sometime this spring, Apple’s 505,000-square-foot North Carolina data center is, according to COO Tim Cook, intended to support iTunes and MobileMe. But we don’t yet know in what capacity, and Cook’s remark, which is at once unambiguous and utterly cryptic, leaves plenty of room for speculation. And theories about the potential capabilities of this new facility abound.

In a research note this week, Bernstein analyst Toni Sacconaghi reviewed a few of the more plausible ones, which run the gamut from the long-rumored iTunes streaming service to the back end for a natural language voice interface and navigation service for its iOS devices.

The first: an easy way of scaling the company’s iAd mobile advertising program. With its installed base of iOS devices likely to hit 200 million by the end of fiscal 2011, iAds could put quite a strain on Apple’s ad serving capabilities. Says Sacconaghi, “If iAd gets traction while serving interactive, multimedia ads then Apple’s underlying advertising platform will need to be significantly larger, and at a scale comparable to Google’s or Microsoft’s ad platforms.”

The second is a no-brainer and seems fairly likely to pan out given recent chatter about Apple’s MobileMe service going free come April: an overhauled verision of MobileMe that provides improved cloud-based synchronization of data and media, along with meaningful storage capacity.

The low-cost iTunes subscription service we’ve been hearing about for years now is the third. Again, this seems a completely plausible use for Apple’s new data center, and as Sacconaghi notes, the time may finally be right for the company to launch it. “In our meeting with Apple executives last month, VP of Internet Services Eddy Cue suggested that the reason that music subscription services had failed to receive traction with consumers was because they were too expensive, highlighting prevailing rates of up to $15/month,” he wrote. “We note that an annual subscription to Napster now costs $8/month, while the paid version – unlimited, higher-quality, ad-free – of music streaming from Pandora costs $3/month.” Add to that Apple’s 2010 acquisition of streaming music outfit Lala and the traction Spotify and Pandora have been gaining in the market recently and this seems a likely scenario as well.

Fourth on Sacconaghi’s list, an aggressively priced video streaming service. Given the popularity of Netflix’s iOS app among iPhone and iPad users, it might make sense for Apple to offer its own competing video subscription service. Or, it could simply acquire Netflix. It’s not like Apple doesn’t have the money to do it–even if Netflix’s market cap is north of $11 billion.

And finally there’s that voice interface and navigation service I mentioned earlier. This one might seem a stretch, but don’t dismiss it out of hand. Last April, Apple acquired Siri, developer of a virtual personal assistant supported by speech recognition, natural language processing and semantic Web search. And in 2009 it purchased PlaceBase, a mapping outfit that specialized in enhancing maps with private and public data sets. Put those two acquisitions together with a massive data center and Sacconaghi’s theory looks at least plausible.

“Apple could offer its own navigation service comparable to Google’s free and very popular voice-based navigation system on Android,” he writes. “The iOS device user-base could then potentially periodically upload anonymous information on routes travelled and speeds encountered, perhaps even in real-time, which would allow Apple to report back traffic conditions to its user-base.”

Sounds like a killer feature for the iPhone 5, doesn’t it?

PREVIOUSLY:

• Apple’s N.C. Data Center Intended for iTunes, MobileMe
• Apple Owns Another 70 Acres Near NC Data Center
• Was Apple Planning on Doubling Its North Carolina Data Center All Along?
• Apple Reaching for the Cloud With MacBook Air and N.C. Data Center
• That’s Apple’s New Data Center? Where’s the Giant Glass Cube?

]]> http://allthingsd.com/20110323/apple-data-center-theories/feed/ 16 http://allthingsd.com/20110321/twitter-brags-of-successful-data-center-migration/ http://allthingsd.com/20110321/twitter-brags-of-successful-data-center-migration/#comments Mon, 21 Mar 2011 22:30:56 +0000 Liz Gannes http://networkeffect.allthingsd.com/?p=4481 Twitter, after years of fighting the Fail Whale, has recently made major changes under the hood to ensure its messaging service will stay online. Its engineering and operations teams coordinated a massive effort starting last September to migrate Twitter not once, but twice, shifting 20 TB of tweets and live traffic-serving from a first data center to a second testing center to its third “final nesting ground.”

Over the last couple months in particular, I’ve heard murmurs through the grapevine of Twitter engineers boasting as shifts happened with no disruption to the service.

Twitter VP of Engineering Michael Abbott publicly bragged about the migration today in a blog post, but declined to disclose where the data center musical chairs actually occurred, or why multiple moves were necessary.

Since 2008, the company–which coincidentally is celebrating the five-year anniversary of the first tweet today–2008 had been using NTT America Enterprise Hosting Services, which has a data center in San Jose, Calif., and later expanded to Santa Clara. In July 2010, Twitter said it would move into a custom-built data center in Salt Lake City before the end of that year.

But that didn’t happen, and in December, Twitter reportedly leased data center space in Sacramento.

In a rare sentence that didn’t include a bird metaphor (flocking! nesting! migration!), Abbott described the implications of the data center move in his blog post: “This move gives us the capacity to deliver Tweets with greater reliability and speed, and creates more runway to focus on the most interesting operations and engineering problems.”

]]> http://allthingsd.com/20110321/twitter-brags-of-successful-data-center-migration/feed/ 0 http://allthingsd.com/20110321/peter-levine-veritas-veteran-and-data-center-guru-joins-andreesen-horowitz/ http://allthingsd.com/20110321/peter-levine-veritas-veteran-and-data-center-guru-joins-andreesen-horowitz/#comments Mon, 21 Mar 2011 19:32:01 +0000 Arik Hesseldahl http://newenterprise.allthingsd.com/?p=4190 Venture capital firm Andreessen-Horowitz said today that it has appointed Peter Levine, a veteran of the enterprise software company Veritas that’s now a part of Symantec, and the former CEO of XenSource, now part of Citrix, as its first venture partner.

Levine is the third partner to join AH in recent months. In January it named HP and Opsware veteran Mark Cranney as a partner for market development. And in March it added IronPort’s Scott Weiss.

“Adding Peter makes us smarter at the firm on a certain class of products where he is much more experienced and goes much more in depth than we do, in areas like virtualization and storage,” AH co-founder Ben Horowitz told me. A key area of expertise is one that Levine developed specifically at Veritas, he said, that of working with manufacturers of infrastructure products. “Veritas was probably the most successful company in the history of enterprise software at the OEM model except for Microsoft,” Horowitz said. “It’s a very complicated thing to do–and a very complicated thing to do correctly–so he brings a specialized skill set to the table.”

Horowitz also said Levine will help AH expand its reach and find deals in places where it hasn’t had a presence before, places like the Massachusetts Institute of Technology where Levine is a lecturer. One example: Bromium, a stealth startup that AH says it is investing in. While Horowitz didn’t disclose the amount the firm is investing, he did describe Bromium as a “security plus virtualization company.”

“It’s the kind of deal we wouldn’t have known about without working with Peter,” Horowitz told me.

“We’re at the second generation of what you can do with virtualization,” Levine told me. “Companies like Citrix and XenSource did a lot of the hard rock-breaking to get chipset support from companies like Intel to support virtualization, and once they did that there was an opportunity to take virtualization to the next level. Bromium is a company that takes advantage of all that.”

Levine is continuing in his role as a vice president of Strategic Development at Citrix and will continue teaching a class on Technology Sales at MIT’s Sloan School of Management. Previously, he was senior vice president and general manager of the Data Center and Cloud Division at Citrix, having joined that company in 2007 by way of its $500 million acquisition of XenSource, a provider of open-source virtualization sofware, where he was CEO. XenSource’s customers included Microsoft, Symantec, HP, NEC and Dell.

This will be Levine’s second go in the venture capital ring. He spent three years as a general partner at the Mayfield Fund and in that capacity served on the board of Consera Software, which was purchased by HP. He sat on the advisory board of VMWare and was an investor in Actona, which was ultimately acquired by Cisco Systems.

Levine first rose to prominence as an early employee of Veritas Software, and during his 11-year stint there helped to grow it to 5,000+ and more than $1.5B in annual revenue. His last job at Veritas was executive VP, where he was responsible for worldwide marketing, OEM sales, business development and several product divisions. Before that, he was a software engineer at MIT and worked on Project Athena, an early-1980s research project to build a campus-wide distributed computing network that turned out to be a forerunner of the kind of corporate networks we now use every day.

]]> http://allthingsd.com/20110321/peter-levine-veritas-veteran-and-data-center-guru-joins-andreesen-horowitz/feed/ 0 http://allthingsd.com/20110318/seven-questions-for-salesforce-coms-parker-harris/ http://allthingsd.com/20110318/seven-questions-for-salesforce-coms-parker-harris/#comments Fri, 18 Mar 2011 11:30:47 +0000 Arik Hesseldahl http://newenterprise.allthingsd.com/?p=4125 Last week I took advantage of the fact that I was in San Francisco for Hewlett-Packard’s big summit meeting, and stopped by the offices of Salesforce.com. There I met up with Parker Harris, executive vice president for technology, and one of the company’s four founders.

There’s so much going on at Salesforce it’s hard to keep track of it all. We talked a bit about Chatter.com and the result of the company’s efforts to promote it during the Super Bowl, and also about the state of mobile devices and where its priorities are. But I started with a question about Japan.

NewEnterprise: Everyone is talking about what’s been happening in Japan. You have a data center under construction there. Has there been any effect on your plans?

Harris: No. The data center is in Tokyo so it’s outside of the area directly impacted by the earthquake. We chose the location not only for the earthquake-proof nature of the building, but also for access to diesel generators, which have proven pretty important given the power situation. There’s been no interruption at all in what we’re doing.

So what are your pain points, what are you dealing with this year? Judging by your growth I’m guessing the list is long.

The big one is around trust, reliability, availability and scalability as we grow. I would say it’s not the biggest pain point that we have because we’ve been focused on it for so long. We did have a period several years ago when we had a lot of issues. I think a lot of major services go through that: eBay, Google and Twitter are all examples. I think that’s because none of them are the same. They all grow organically as the customers and technology grow. Chatter is a big focus now.

Speaking of Chatter, how did the Super Bowl ad for Chatter.com work out?

The Super Bowl was an interesting challenge because we had to make sure we could handle the load of Super Bowl traffic. We have an interesting relationship with Will.I.Am. He’s a friend of Marc’s. They started in this odd place where he wants to get into technology and wants to expand his brand. And Marc started talking to him about collaboration. And it was kind of a crazy idea. It was kind of a consumer play with the Super Bowl. Chatter.com is kind of a pro-sumer product where we want individuals to use it. We didn’t really think people sitting on the couch drinking beer would use it right away. But we knew it would attract some attention, but the after effect of discussion around the ad, the YouTube video of the making of the ad, and all the talk around it had a good effect. It got me ready to make sure I had a Web site that could handle a lot of traffic. We partnered with Akamai to cache a lot of the static content. We did a lot of testing of the sign-up process during spikes and peak loads.

Did you the see the spike you hoped for?

We saw a huge spike in traffic to the Web site and traffic to the sign-up page through the following week. A lot of it was from phones, from people sitting on the couch. This is part of our transformation to what we call Cloud 2.0 that we’ve been talking about so much. Historically our Web sites didn’t work on mobile devices that well. Our app didn’t work that well. Chatter.com was a case where we did the mobile version first.

What are you working on right now?

With Chatter it’s about adoption, and how do we get people outside your company collaborating with you. We kind of do that now, but there’s stuff we need to do in the product to make it more usable. That would be a big next step for Chatter. In the Service Cloud we want to reinvent the low end. We’re taking Chatter as an influence, and we look at cool little companies like Zendesk, which does a nice job at the low end of the market. We want to reinvent the portal experience with a Chatter influence on the Service Cloud. On the Sales Cloud we want to focus a lot on the sales rep’s experience, and I think mobile is a big factor there. Phones are a big deal, but tablets are an even bigger deal. So we’re doing a lot of design work and experimentation around the tablet experience. And how much do we re-think our experience on the tablet.

I’ve heard some people say that if you want to invest in the future of the iPad and think Apple stock is too expensive, then Salesforce is a good bet because you’re doing so much on that device. Do you think that’s fair?

We’re paying a lot of attention to the iPad. But we’re expanding that to a tablet focus. We definitely think Google’s Android will get a lot of adoption. It’s a fragmented market still but they’ll get there. I think the iPad is still winning in the enterprise. We don’t want to underestimate Android.

What do you think of HP’s WebOS? Léo Apotheker had a lot to say about that yesterday. Do you have any interest there?

We’re not keeping an eye on that right now. We’re looking at Research In Motion’s Playbook and seeing where that goes mainly because we have a ton of people who use the Blackberry in our customer base. It’s still the best email device. I gave mine up for an iPhone, but for cranking through email it’s still better. And because of our close relationship with RIM we’re going to see if there is something we can do with the Playbook. The mobile space is a hard place to make bets. So we’re working hard on our HTML5 strategy so that we can have something that will work that’s cloud based with other devices.

]]> http://allthingsd.com/20110318/seven-questions-for-salesforce-coms-parker-harris/feed/ 1