The following parts is when I felt the guys pain.

I was curious and I wanted to find the scattered bits of my online life before dumping everything on my laptop onto the Web.

So I decided to go on a scavenger hunt into the cloud.

…

Before I started the search, when I thought about cloud computing, this is the image that came to mind: a giant cartoon cloud just slurped information off of my computer like magic. My files just floated in the sky until I wanted them back.

Video: What is cloud computing?

The cloud doesn't work like that. It's made up of a massive and growing network of data centers, which are huge warehouses full of computers. They store and process information from all around the world, largely in secret.

Then he finally connected with Rich Miller who helped him a bit.  Except he realized he was kind of a clueless.

I found it shocking that the gut of the cloud, an image I found so soft and quaint, was actually comprised of an enormous and ever-growing network of machines.

But apparently lots of people already knew this.

"All the clouds live in data centers," Rich Miller, editor of a prominent cloud-computing blog called Data Center Knowledge, told me. "There's always hardware involved, and bricks and mortar. ... It's not a fluffy cloud. It's living in someone's building."

Awesome. So all I needed to do to find my family photos and the rest of my data was to call up the data center where it lives and go there, right?

Wrong. I quickly learned tours of the cloud aren't easy to come by.

I want a tour of a data center.  Let’s call Google.

Google, which has most of my sensitive data, like e-mail, calendars, to-do lists and documents, declined an interview request for this story. A spokeswoman said the company doesn't give tours either. Go figure.

Excited he gets a tour of an IBM data center.

Dismayed, I started turning to companies who don't have my data, just hoping to get a sense of how this system worked. IBM offered to give me a tour, maybe because, like me, it's trying to break into the cloud world.

and , finds a PUE of over 2.0.  Hopefully the IBM rep didn’t try to explain PUE to him.

Inside, I found rows of black, refrigerator-sized computer towers, 4,000 of them in all. They buzzed and whirred so loudly that I had to lean in to hear my tour guides. In front of the towers, grates and pinholes in the floor pump out frigid air to keep the machines from overheating. The computers breathe this air in and then exhale air as hot as a hair dryer's.

I'm told the cooling bill here costs more than running the computers.

So, he asks more questions and gets more confused.

As I walked around the center, IBM employees did their best to explain this hyper-complicated system. They rattled off machine types, specs and technical details faster than I could write them down.

I confess that I left the IBM tour not feeling much better about the safety of my data. Not that there seemed to be anything wrong with their cloud computing center. The IBM staff was friendly and helpful. The machines looked nice.

Maybe he is a snipe hunt.

A snipe hunt, a form of wild-goose chase that is also known as a fool's errand, is a type of practical joke that involves experienced people making fun of credulous newcomers by giving them an impossible or imaginary ta

So maybe he is asking the wrong question.

But the more I mulled over my failing scavenger hunt, the more I thought that maybe I was asking the wrong question. Perhaps it doesn't matter where my data is, just that there's some way for me to get a sense of how well it's managed.

And, then he realizes maybe he doesn’t own the data.

"Terms of service" agreements offer some details on free services. But, after reading several, it's still unclear to me who owns my data, if I can ever delete it from some sites and what would happen if any of these companies goes bankrupt. In response to an e-mail question about what would happen to Facebook data if the site closed, a company spokeswoman wrote, "The business is doing well and continuing to grow."

Who do you trust?  He talks to Microsoft’s Brian Hall.  Note: I was interested in what Brian had to say as I worked with Brian on Windows XP.

Still, without information, it's hard to know who to trust.

That makes it easy to fall back on flimsy methods of comparison, like going with a brand you already know. I'm sure this is how I ended up with so much data on Google's servers. It's a huge company. Billions use their search. Tens of millions save files with Gmail. They've got to know what's up, right?

That's exactly what the big cloud companies hope you will think. Microsoft's general manager of Windows Live, Brian Hall, told me brand recognition is the best way for people to compare services.

"Consumers, they don't really care if there are 9,000 data centers or two data centers as long as they have confidence that we're going to protect their data and they'll have access to it when they want to have access to it," he said. (In case you're wondering, Hall said Microsoft has "between 10 and 100 data centers" worldwide. Really specific.)

After all this, his conclusion is good.

The most important thing I realized on this search, though, was rather basic:

The cloud is not some fluffy ball of magic, it's an energy-sucking and fallible machine.

One I'll be more cautious before trusting.

But when you mention "companies like Google" - are there really many companies like Google? I don't think so - not today.
Best,
Philip

I’ve actually had a few skype conversations with Philip and last year at Data Center Dynamics London I met Philip. So, I know he is a regular reader.

I agree there are not many companies like Google.  Here are a few things I think Google does that fit the model of an information factory.

1. Urs Hoelzle as executive and influential in the company running data centers understands the role of Google’s information factories.  Once I asked Urs why he doesn’t shut down idle servers, his response was he would rather think how does he use the servers while they are idle.  And, Urs can think this way given his position and influence in Google.  What Google knows that few do is turning on and off servers, is not reliable enough for a lights-off type of operation. Desktops, laptops, mobile devices, and phones all have this problem as well, but people are pushing the buttons and can try again when turning on fails.
2. Their focus on PUE accuracy and reporting demonstrates their thinking in process control and statistical accuracy.
3. Google knows the shell of a building is cheap, and 85% of the costs in data centers is in the power and cooling infrastructure.
4. The cost of electricity is greater the the cost of a server over a typical 3 year lifespan.
5. The vendors – data center, server, network are just as silo’d as companies IT organizations and don’t drive for overall system efficiencies.  So, Google designs their own systems, and uses the vendors as subcontractors to their designs.  It may not be totally accurate analogy, but Boeing designs the plane and subcontracts out pieces and components.  There are some pieces that are off the shelf, like engines (for servers processors), but many time parts don’t perform as advertised when integrated.

I could go on, but these are just a few ideas that demonstrates Google runs their data centers as computers, see this paper.  The information factory metaphor communicates the scale, power, and complexity.

As Philip says there are not many like Google.  Which means they have inefficient information factories that are a drain on the companies revenue.  And, in this economy cost reduction is a priority.  Do you cut costs by making the system more efficient?  No.  You cut costs by limiting headcount, budgets and capital expenditures which ironically many times will increase your costs long term as you grow and decrease your overall performance per watt.  Right now many companies don’t need the performance so you are removing capacity from the system to cut costs.  Makes sense, but I bet the company executives did not consciously decide to reduce the capacity of their information factory.  How can you not think reducing costs capital and operating expenses reduces capacity?

In this economy Google may reduce the rate of their expansion, but overall their information factory capacity is growing and performance per watt is improving.

Philip asked a good question.  “are there really many companies like Google”? 

No, but there will be more.  Google can do this because the company itself is an information factory.  And, the future successes in internet services will be those who have the most efficient information factories that can produce information at the lowest costs.

What I propose is a more accurate description of what data centers are in this economy.  The Data Center is an information factory, a building that makes information suitable for use with information machinery – servers, storage, and networking hardware. Information is the raw material input into the factory. Software running on the hardware processes information increasing the value. Like any other manufacturing process electricity is used to power and cool the machinery.  How much power is used to run these information factories, in 2006 1.5% of the US electricity production was in data centers, doubling 2000 consumption, growing at a 12% annual rate.

The above is an image Google uses to illustrate its green Information Factory (aka data center).

So the choices to green the data center are now how do you green your information factory.  Making factories energy efficient is a concept many are familiar with.  Applied to the information factory how do you consume less energy and/or greener energy while increasing the value of information? Making power delivery more efficient applies to all parts of the data center. Cooling systems is a whole topic specialists who can figure out the most efficient way to remove the heat from the IT equipment.  More efficient servers are another choice. And of course there is virtualization.  Not too long ago, for every watt of power supplied to a server, there was another watt used by the power and cooling systems.  Now companies like Google consume only 0.21 watts for power and cooling for every watt used by their information factory hardware which by the way consume less power than what is commonly used by the industry.

Where do you start? Most companies start where they have budget to spend. Huh? Sound silly. Well that is what happens in most companies as the IT organizations within a company are in silos of separation. Imagine if you wanted improve a car’s MPG and approached the problem based on which department had the budget available to make changes to the car.What is needed is an information engineer whose job it is to figure out how to improve the performance per watt in the whole system and prioritize the areas to address.

Companies like Google, Microsoft, Yahoo, Amazon, and eBay have addressed this problem by creating groups who have responsibilities to engineer their information factories.

Is your company running centers of data or information factories? The ones who think like information factories are driving to new levels of performance per watt.  Breaking down silos, to get groups to work together. You can Green the data center by looking at how much energy gets consumed by your information factories to create higher value information. Another choice is where data centers get their power from and the carbon impact.  Using 1.5% of the US electricity consumption data centers have the opportunity to locate near places with renewable energy and is commonly discussed by Google, Microsoft, and Yahoo.

The growth of the readership is amazingly linear.

There is a lot of I’ve learned over the past 2 years, and this blog is part of my system to document what is going in the industry.  It is now so ingrained in how I work that blogging is a natural part of the day, and explains why I am now up to this being my 988 post on the blog.  Hitting 1,000 posts by Nov 26, 2009 my exact 2 year anniversary is not going to be a problem.

And, the international reach is beyond my expectations. Below is the top 100 countries, rated in order of traffic hits to the blog over the last month.

Thanks for visiting my blog, and thanks to my new and old friends who continue to supply me with new ideas to write about.

-Dave Ohara

dave at greenm3 dot com

The authors have deleted this blog. The content is no longer available.

You can create your own free blog on WordPress.com.

There were hopes the TechHermit blog would continue.

TechHermit Returns with New Authors, Speculates the end of Microsoft’s Data Center Program

DataCenterKnowledge spreads the word TechHermit’s blog continues.

Tech Hermit Blog Returns

September 22nd, 2009 : Rich Miller

Back in July I noted the passing of Shane McGew, who wrote about the data center industry at his Tech Hermit blog. So I was surprised to find new posts at the Tech Hermit blog this week.

Here’s the story: “Today we are announcing that through detailed negotiations with the McGew family a group of avid readers have purchased the rights to the Tech Hermit brand and will continue to post under this heading and keep the same edgy feedback that we came to love with Shane. We hope to earn the same level of trust and respect in time.”

Shane was always pretty plugged into goings-on in data center operations at Microsoft, a trend that continues with the new team (whose members remain anonymous). A post today notes the departure of another Microsoft data center executive, Joel Stone, who is headed to Global Switch. Stone’s departure follows the exit of Global Foundation Services corporate VP Debra Chrapaty, who is off to Cisco.

Many of your recognize Christian.  Daniel is not as well known, and brains behind the 4th generation Microsoft data center.

But Microsoft has indicated how the next generation of data center will improve upon the Chicago design.

Moving to containers allows Microsoft to bring in computing capacity as needed, but still requires the company to build the physical building, power and cooling systems well ahead of time. The company's next generation of data center will allow those things to be built in a modular fashion as well.

Daniel had an interview with PCworld that gives you some ideas of his thinking.

"The idea of modular, portable data centers is key to the industry's future," said Daniel Costello, Microsoft director of data center research, in a presentation at GigaOM's Structure 08 conference in San Francisco. "That's why I'm here to talk about data centers, not just for Microsoft but for our customers as well."

Buying these boxes from server vendors can be more energy-efficient and cost-effective than building a new, traditional data center, he said, and Microsoft sees them as more than just a way to add extra computing capacity at short notice. "We see them as a primary packaging unit," he said.

Using shipping containers is part of an effort by Microsoft to radically rethink its data centers, as it tries to add more computing capacity in a way that is cost effective and power efficient. "At Microsoft, we're questioning every component in the data center, up to and including the roof," Costello said. That includes "eliminating concrete from our data center bills."

"The definition of a datacenter has changed. It's not just bricks and mortar any more, and moving forward, we think it can be a lot more energy efficient," he said.

But vendors building portable data centers today are filling them with equipment that was designed for traditional data centers. "Moving forward, we need to design systems specifically for this form factor. If we look at the containers, that form factor will change over time as well."

Microsoft has approached every major server vendor about providing it with equipment, Costello said. He said he thinks "all major vendors" will offer portable data centers within the next two years. Vendors offering them today include Sun Microsystems, Verari Systems, Rackable Systems and American Power Conversion.

The cost benefits come partly from economies of scale. Shipping 2,000 servers in a container is more cost-effective than shipping and installing separate racks, and portable data centers don't require raised floors or as much wiring.

They can offer a better "power unit efficiency" ratio than do traditional data centers, he said. PUE is a measure of a data center's power efficiency. If a server demands 500 watts and the PUE of a data center is 3.0, the power from the grid needed to run the server will be 1500 watts, according to a definition from the Green Grid industry consortium.

"We've seen PUE at a peak of 1.3" in modular data centers, Costello said, compared with between 1.6 and 2.0 for a traditional data center.

The containers can accommodate 1,000 watts per square foot, allowing companies to power a lot more servers in a given area, he said. Many companies are unable to add more equipment to their data centers because power supplies and cooling equipment are at maximum capacity. The portable data centers are an alternative to building new facilities or extending old ones.

Daniel mentions some of the downsides of containers.

There are some drawbacks and plenty of questions to be answered, he said. Some of the cons include a higher cost of failure if the power to a container is cut off, as well as new risks in terms of regulatory compliance. In addition, portable data centers offered today can't accommodate servers from multiple vendors, he said.

Bottom Line: Because containers encapsulate a large amount of information capability with power and cooling infrastructure, the decision to use containers requires a multi-discipline view of the benefits and impacts of using Containers.  Google and Microsoft have done this exercise, and any one who is thinking about containers should consider both of these companies have figured out where it makes sense.

It is one of Mike’s long posts, so I’ll give you excerpts, and give you a Part 1 to get started, and work on a Part 2 later.

I will do my best to try and balance this view across four different axis the Technology, Real Estate, Financial and Operational Considerations.  A sort of ‘ Executives View’  of this technology. I do this because containers as a technology can not and should not be looked at from a technology perspective alone.  To do so is complete folly and you are asking for some very costly problems down the road if you ignore the other factors.  Many love to focus on the interesting technology characteristics or the benefits in efficiency that this technology can bring to bare for an organization but to implement this technology (like any technology really) you need to have a holistic view of the problem you are really trying to solve.

Mike contrasts Moore’s law vs. the glacial pace of innovation in the data center.

Isn’t it interesting then that places where all of this wondrous growth and technological wizardry has manifested itself, the data center or computer room, or data hall has been moving along at a near pseudo-evolutionary standstill.  In fact if one truly looks at the technologies present in most modern data center design they would ultimately find small differences from the very first special purpose data room built by IBM over 40 years ago.

Mike goes on to discuss modularization in the data center.  The NSA is listening to advice like this and have made modularization a requirement of their $1.5 bil data center.

Data Centers themselves have a corollary to the beginning of the industrial revolution.   In fact I am positive that Moore’s observations would hold true and civilization transitioned from an agricultural based economy to that of an industrialized one.   In fact one might say that the current modularization approach to data centers is really just the industrialization of the data center itself.

In the past, each and every data center was built lovingly by hand by a team of master craftsmen and data center artisans.  Each is a one of a kind tool built to solve a set of problems.  Think of the eco-system that has developed around building these modern day castles.  Architects, Engineering firms, construction firms, specialized mechanical industries, and a host of others that all come together to create each and every masterpiece.    So to, did those who built plows, and hammers, clocks and sextants, and the tools of the previous era specialize in making each item, one by one.   That is, of course, until the industrial revolution.

Data Centers are not buildings they are information factories.

The data center modularization movement is not limited to containers and there is some incredibly ingenious stuff happening in this space out there today outside of containers, but one can easily see the industrial benefits of mass producing such technology.  This approach simply creates more value, reduces cost and complexity, makes technology cheaper and simplifies the whole.  No longer are companies limited to working with the arcane forces of data center design and construction, many of these components are being pre-packaged, pre-manufactured and becoming more aggregated.  Reducing the complexity of the past. 

And why shouldn’t it?   Data Centers live at the intersection of Information and Real Estate.   They are more like machines than buildings but share common elements of both buildings and technology.   All one has to do is look at it from a financial perspective to see how true this is.   In terms of construction, the cost of data centers break down to the following simple format.  Roughly 85% of the total costs to build the facility is made up of the components, labor, and technology to deal with the distribution or cooling of the electrical consumption.

Data Centers are mostly built out of sync with the pace of technology change.

From the perspective of the technology drivers behind this change is the fact that most existing data centers are not designed or instrumented to handle the demands of the changing technology requirements occurring within the data center today.

Data Center managers are being faced with increasingly varied redundancy and resiliency requirements within the footprints that they manage.   They continue to support environments that heavily rely upon the infrastructure to provide robust reliability to ensure that key applications do not fail.  But applications are changing.  Increasingly there are applications that do not require the same level of infrastructure to be deployed because either the application is built in such a way that it is more geo-diverse or server-diverse. Perhaps the internal business units have deployed some test servers or lab / R&D environments that do not need this level of infrastructure. With the amount of RFPs out there demanding more diversity from software and application developers to solve the redundancy issue in software rather than large capital spend requirements on behalf of the enterprise, this is a trend likely to continue for some time.  Regardless the reason the variability challenge that data center managers are facing are greater than ever before.

Mike brings up the problem that occurs as people build costly custom buildings.

From a business / finance perspective companies are faced with some interesting challenges as well.  The first is that the global inventory for data center space (from a leasing or purchase perspective) is sparse at best.    This is resulting from a glut of capacity after the dotcom era and the resulting land grab that occurred after 9/11 and the Finance industry chewing up much of the good inventory.    Additive to this is the fact that there is a real reluctance to build these costly facilities speculatively.   This is a combination of how the market was burned in the dotcom days, and the general lack of availability and access to large sums of capital.  Both of these factors are driving data center space to be a tight resource.

In my opinion the biggest problem across every company I have encountered is that of capacity planning.  Most organizations cannot accurately reflect how much data center capacity they will need in next year let alone 3 or 5 years from now.   Its a challenge that I have invested a lot of time trying to solve and its just not that easy.   But this lack of predictability exacerbates the problems for most companies.  By the time they realize they are running out of capacity or need additional capacity it becomes a time to market problem.   Given the inventory challenge I mentioned above this can position a company in a very uncomfortable place.   Especially if you take the all-in industry average of building a traditional data center yourself in a timeline somewhere between 106 and 152 weeks. 

to be continued…

NSA To Build $1.5 Billion Cybersecurity Data Center

The massive complex, comprising up to 1.5 million square feet of building space, will provide intelligence and warnings related to cybersecurity threats across government.

By J. Nicholas Hoover
InformationWeek
October 29, 2009 01:07 PM

The National Security Agency, whose job it is to protect national security systems, will soon break ground on a data center in Utah that's budgeted to cost $1.5 billion.

The NSA is building the facility to provide intelligence and warnings related to cybersecurity threats, cybersecurity support to defense and civilian agency networks, and technical assistance to the Department of Homeland Security, according to a transcript of remarks by Glenn Gaffney, deputy director of national intelligence for collection, who is responsible for oversight of cyber intelligence activities in the Office of the Director of National Intelligence.

The budget document specifies green features.

Facility design goal will be to the highest LEED standard attainable within available resources and will include: sustainable site characteristics, water and energy efficiency, materials and resources criteria, and indoor environmental quality.

And even the generators have exhaust scrubbers.

Generators will include Selective Catalytic Reduction (SCR) pollution control equipment, chemical storage tanks and feed
system.

Endowing The Future

The internationally recognized University of Missouri Research Reactor (MURR), a
10-megawatt facility, is the most powerful among the dozens of research reactors located on our nation’s university campuses.

Even worldwide, few facilities can compare.

Those at MURR treat it like the unique national resource that it is, employing the facility as a research source – providing products and services that will save or extend people’s lives.

The superior level of the science at MURR helps put the products and services it offers a step ahead, further fueling the depth of its research.

What kind of reactor is it?  Here is description and a comparison to a local nuclear plant.

Operations

Power Level: 10MW_th
Power Density, Core: 303 kW/liter, with peaking factor greater than 3
Primary Coolant Operating Temperature, Outlet (T_h): 136° F
Primary Coolant Operating Pressure: 80 psia
Core, Fuel Type: Open pool PWR, HEU aluminide fuel
LEU Conversion Feasibility Study: Currently underway
Reflector: Beryllium and graphite
Flux Trap, Peak: 6 x 10¹⁴ n/(cm²sec)
License Status: 20-year renewal request submitted August 2006 (to extend to October 2026)

General

The University of Missouri Research Reactor Center has an impeccable 40+ year record of safe operation. This safety record is a combination of stringent NRC-directed safety regulations, high-quality technical and operations staff, and a philosophy of proactive, preventive maintenance. MURR operates 6.5 days per week; 52 weeks per year.

MURR’s reliability record is the envy of the industry.

Missourians are no strangers to nuclear energy, relying heavily on Ameren UE’s nearby Callaway plant to cool or heat their businesses and homes and feed their equipment and appliances. But MURR, as a multifunctional research reactor, differs significantly from such a power reactor. As the table below shows, MURR is considerably smaller. While a power reactor such as Callaway needs a source of electricity to cool the reactor core, MURR’s pool is capable of absorbing all the heat from the reactor core without the aid of forced convection.

But the facility is not just for power.

Products & Services

With its reliable and growing Products and Services division, the people at MURR are making possible custom-made quality nuclear services to a global community – offering a full line of off-shelf products, analyses, research and commercial isotopes and radiochemicals, and analytical and neutron irradiation services to its customers.

• Radioisotopes/Radiochemicals
  MURR’s quality systems, bulk and research grade radioisotopes and radiochemicals and 16 MeV cyclotron makes it possible to deliver tailored products to fit customer needs.
• Irradiation Services
  Our reactor design and full-power weekly operating cycle make possible an extensive array of irradiation services.
• Contract Manufacturing
  MURR’s swift and accurate development and production capabilities help meet on-time needs for its customers and colleagues.
• Analytical Services
  The folks at MURR can’t solve all the mysteries of life, but with Neutron Activation Analysis and an array of other techniques, they can come close.
• Business Incubator Facilities & Economic Development
  From concept to commerce MURR, with help from the University of Missouri, is the place to be if you’re a small business in need of incubator facilities.

Maybe next time I am in Columbia, I can get a tour of the facility.

And, there are other parts of the University who take a green engineering approach as well.

Engineering goes green

The college’s favorite color connects to more than just St. Patrick

• Story by Chris Blose
• Illustration by Josh Nichols
• Published: March 13, 2008

Around mid-March, a strange thing happens at the College of Engineering: People start wearing a lot of green. The occasional figure shows up dressed as St. Patrick, complete with flowing beard and green robes. The dome of Jesse Hall glows green at night. All of these things come in celebration of Engineers' Week, or E-week.

That’s just one week, though. All year long, students and scholars in engineering take the “green” connection to another level through environmentally friendly research and actions — from electric cars and alternative energy to recycled materials and energy audits.

In honor of E-Week at Mizzou, Mizzou Wire presents a few of the many examples of the university’s green engineering. Let us know if you have more examples.

Intel seeks new 'microserver' standard

by Stephen Shankland

• Font size
• Print
• E-mail
• Share
• 2 comments

SAN FRANCISCO--In September, Intel introduced its back-to-the-future idea of tiny "microservers." Now the company wants to make the design into a standard others can use, too.

The chipmaker will offer its design specification to the Server System Infrastructure Forum by the end of the year, said Jason Waxman, general manager of Intel's high-density computing group. If the group's board votes its approval for the specification, group members may use the designs royalty-free, he said in a meeting with reporters here.

"Before the end of the year, it will happen," Waxman said.

An Intel 'microserver'

(Credit: Stephen Shankland/CNET)

The computer industry is in constant tension between proprietary designs and standards that anyone may use. The former can mean tidy profits for companies, as long as the technology is widely adopted, but the latter can spur broader adoption. Intel's primary business, selling processors, benefits more from the latter when it comes to cultivating a new server market segment.

What is inside the Intel’s offering?

What's inside?
The diminutive server consists of a single quad-core processor and four memory banks. Intel showed 16 microservers housed in an 8.75-inch-tall chassis that supplies them all with power, cooling, and a network connection to the outside world. Along the bottom of the chassis is a bay with 16 "sleds" that each has a trio of 2.5-inch hard drives that directly connect to each microserver.

The present microserver uses a 1.86GHz quad-core processor, the "Lynnfield" model of Intel's new "Nehalem" generation. Its top power consumption is 45 watts, but early in 2010, Intel will release a dual-core "Clarkdale" model that consumes only 30 watts when running flat-out.

That's at the top end, though. Intel's goal is for the entire microserver--which also includes memory and supporting chips--to idle at just 25 watts of power.

An interesting part not discussed is how much is the microserver?