Amazon.com Data Center

I've been searching for information on amazon.com data centers, and there isn't a whole lot out there.  Found one Turner Construction project with reference to one facility.

Amazon.com Data Center
Seattle, WA
This phased fast-tracked demolition, renovation and structural upgrade of approximately 10,000 sq. ft. basement and approximately 18,000 sq. ft. of first floor in existing downtown five-story office building. Electrical support for data center included construction of Seattle City Light electrical transformer vault, N+1 redundant power and cooling, and structural reinforcement of the existing structure for seismic upgrades. In addition open workspace and offices were carved out of the area. Intensive just-in-time pre-construction sessions with the design team, city utilities and Amazon.com staff has made it possible to design, obtain permit and substantially complete in record time - five months from formal Notice to Proceed to an operational data center

There is the UC Berkeley/Stanford paper.

Despite significant efforts in the field of Autonomic Computing,
system operators will still play a critical role in administering
Internet services for many years to come. However,
very little is know about how system operators work,
what tools they use and how we can make them more efficient.
In this paper we study the practices of operators in a
large-scale Internet service Amazon.com and propose a new
set of tools for operators. The first tool lets the operators
explore the health of system components and dependencies
between them; the other monitors the actions of operators
and automatically suggests solutions to recurring problems.

I think I need to talk to some other data center friends who work in Seattle to find more about amazon.com.

Read more

HP Labs announces 3 Green IT projects - Sustainable Data Center, Replacing Copper with Light, and Modeling/Measuring Energy & Resource Use

HP Labs has a press release on their 3 new Green IT initiatives.  This is part of the battle between HP and IBM for green data center solutions in the enterprise. Much of what is mentioned in this press release are the same type of things I've seen the IBM guys talk about.

HP today announced new research initiatives from HP Labs, the company’s central research arm, aimed at developing new technologies and business models that leave a lighter carbon footprint.

Initially, HP Labs will focus its research in sustainability on three major projects, including: an industry-first initiative to reduce the carbon footprint of data centers by 75 percent; groundbreaking research to replace copper wiring in servers with laser light beams; and tools for measuring and managing the amount of energy used to develop products.

Sustainability is one of five major research themes of the newly redesigned HP Labs, which recently refocused its efforts to address the most complex challenges facing technology customers in the next decade.

“HP’s long-standing commitment to the environment is second to none in the technology industry. Today, HP Labs extends our dedication with these important research initiatives that will advance the state of the art in sustainable IT,” said Prith Banerjee, senior vice president, research, and director, HP Labs. “HP Labs will lead the industry in developing the technology that could dramatically reduce energy consumption and the carbon footprint of entire industries.”

It is good to see the efforts in optical networking to replace copper is a priority.

Photonic interconnections make it possible to fit dozens, and eventually hundreds, of processors on server system chips. In addition, the optical connections are 20 times more efficient than what is on the market today and will save companies multiple gigawatts of power annually, thus driving down IT costs. The photonic interconnects, which range in distance from 100 meters to 100 nanometers, also enable more flexible system configurations that can be quickly redeployed based on business needs.

Read more

Amazon and Google, Winning the Battle

And, the news continues about amazon and google, claiming these 2 will be the winners as the US Internet matures. Reuters reports on a new study.

SAN FRANCISCO (Reuters) - An Internet analyst for a major Wall Street firm argues in a new report that Google Inc and Amazon.com Inc will be long-term winners, while Yahoo and IAC InterActiveCorp fall by the wayside and eBay Inc becomes a merger target.

Sanford C. Bernstein analyst Jeffrey Lindsay argues in a 310-page report entitled "U.S. Internet: The End of the Beginning" to be published on Tuesday that Google and Amazon are best placed to withstand the current economic downturn.

"We expect two players to continue to perform strongly, Google and Amazon," Lindsay writes. "Both Google and Amazon.com are still racking up annual growth rates in the 30-40 percent range, with only a relatively modest slowdown in sight."

Read more

Bill Gates talks about Microsoft's Web Services in Big, Big, Data Centers

Bill Gates keynote transcript at TechEd has the following information about Microsoft's web services running hundreds of thousands of servers. In the future millions.

Now to make these services reliable, to make it easy to call them, to provide the kind of security that you want, there are a lot of new developments that have had to take place, things like identity federation. Things like the protocols that you see in the WS* standards, and that we've made easy to get to through the Communication Framework libraries. We're taking everything we do at the server level, and saying that we will have a service that mirrors that exactly. The simplest one of those is to say, okay, I can run Exchange on premise, or I can connect up to it as a service. But even at the BizTalk level, we'll have BizTalk Services. For SQL, we'll have SQL Server Data Services, and so you can connect up, build the database. It will be hosted in our cloud with the big, big data center, and geo-distributed automatically. This is kind of fascinating because it's getting us to think about data centers at a scale that never existed before. Literally today we have, in our data center, many hundreds of thousands of servers, and in the future we'll have many millions of those servers.

When you think about the design of how you bring the power in, how you deal with the heating, what sort of sensors do you have, what kind of design do you want for the motherboard, you can be very radical, in fact, come up with some huge improvements as you design for this type of scale. And so going from a single server all the way up to this mega data center that Microsoft, and only a few others will have, it gives you an option to run pieces of your software at that level.

You'll have hybrids that will be very straightforward. If you want to use it just for an overload condition, or disaster recovery, but the software advances to make it so when you write your software you don't have to care where those things are located, those are already coming into play. So the services way of thinking about things is very important, and will cause a lot of change.

I think Mike Manos has been in a few BillG reviews. Bill mentions data center issues - power in, how you deal with heating, sensors, and you can be very radical.

Read more

James Hamilton's Comparison Google Application Engine vs. Amazon Web Services

James Hamilton attended the Google IO conference and posts his comparison of Google Application Engine (GAE) vs. Amazon Web Services (AWS).

Last week at Google IO, pricing was announced for Google Application Engine. Actually it was blogged the night before at: http://googleappengine.blogspot.com/2008/05/announcing-open-signups-expected.html.

The prices are close to identical with Amazon AWS although GAE differs substantially from the AWS offerings.  The former offers a easy to use Python  execution environment whereas Amazon offers the infinitely flexible run-this-virtual-machine model. Clearly the Amazon model costs more to provide so, by that measure, AWS pricing is somewhat better:

Google Application Engine Pricing:

· $0.10 - $0.12 per CPU core-hour

· $0.15 - $0.18 per GB-month of storage

· $0.11 - $0.13 per GB outgoing bandwidth

· $0.09 - $0.11 per GB incoming bandwidth

· From: http://googleappengine.blogspot.com/2008/05/announcing-open-signups-expected.html

Compared with AWS Pricing:

· $0.10 - $0.80 per VM hour (depending upon resources allocated)

· $0.15 per GB-month of storage

· $0.100 - $0.170 per GB outgoing bandwidth

· $0.100 per GB incoming bandwidth

· From: http://www.amazon.com/S3-AWS-home-page-Money/b/ref=sc_fe_l_2?ie=UTF8&node=16427261&no=3435361&me=A36L942TSJ2AJA and http://www.amazon.com/EC2-AWS-Service-Pricing/b/ref=sc_fe_l_2?ie=UTF8&node=201590011&no=3435361&me=A36L942TSJ2AJA.

There are some important differences that make the pricing comparison somewhat biased in a couple of ways. Two important differences: 1) as mentioned above, Amazon gives an entire virtual machine so EC2 is much more flexible than GAE both in that it can run arbitrary applications in arbitrary languages and that it supports all execution models whereas GAE only supports HTTP request/response.  Another key difference is the storage subsystem.  In the numbers above, we’re comparing the Amazon blob store (S3) with the more structured storage model offered by GAE.  The more comparable AWS SimpleDB pricing is considerably higher than the GAE storage pricing. SimpleDB charges $1.50 GB/month in addition to machine usage and network transmission costs.  GAE is offering much more affordable semi-structured storage and the GAE storage model actually supports data types rather than having to force everything to character format.

And, James makes a good point about the Amazon as a retailer vs. google.  Amazon understands as any good retailer, Wal-mart, Costco - the pennies added up, and are important given your low margins.

GAE is still free to start with under 5M page views/month and up to ½ GB storage for free.  Obviously this helps developers get started without strings and that’s a good thing. But, more importantly, it avoids Google from having to go to the expense of billing very small values.  In a weird sort of way, I’m more impressed with AWS billing $0.04 on some accounts in that it shows there billing system is incredibly lean. Scaling down billing is hard, hard, hard.

Read more