What is the PUE of your cloud data center? Google's is 1.10, Microsoft's is 1.13 - 1.2, Amazon is ?

It is a pretty safe assumption that a Cloud Data Center has a low PUE.  The Cloud business is so competitive that the cost to run the power and cooling systems takes a direct hit out of profit margins so almost everyone should be driving more efficient systems.

How efficient are the cloud companies?  

Google is easy to figure out as they quote PUE quarterly here.

GigaOm's Stacey Higginbotham had a post on efficient data centers quoting PUE.

Microsoft gave Stacey a bunch of data, but not an exact number.

Microsoft sent me a bunch of information on its PUE figures for its newest data centers which range from 1.13 to 1.2. It doesn’t disclose the PUE for all of its data centers, however.

For Amazon, there is no clear answer.  Note: James Hamilton does not claim the PUE is representative of Amazon.  Given Amazon will let temperatures rise in warehouses for workers, it is hard to believe they wouldn't do the same for voiceless servers.

Amazon’s data center guru James Hamilton published a presentation on Amazon last year that assumed a PUE of 1.45 for the online retailer’s data centers.

Are VCs part of what is wrong with DCIM? Features that miss the concept of START

I don't know about you, but whenever I see a DCIM dashboard.  I am confused on where to start.  One of my data center friends has gone through the exercise of looking at about 15 different DCIM solutions.  We agree that one stands out as it is easier to use than the rest.  Why?  You know where to start and the first steps to use the product.  Windows 95 focused on usability and highlighted "START".

Windows 95 introduced a redesigned shell based around a desktop metaphor; unlike Windows 3.1 (where it was used to display running applications), the desktop was re-purposed to hold shortcuts to applications, files, and folders. Running applications were now displayed as buttons on a taskbar across the bottom of the screen, which also contains a notification area used to display icons for background applications, a volume control, and the current time. The Start menu, invoked by clicking the "Start" button also contained on the taskbar, was introduced as an additional means of launching applications or opening documents—while maintaining the program groups used by its predecessor, Program Manager, it now displayed applications within cascading sub-menus. The previous File Managerprogram was also replaced by Windows Explorer.

So, what is wrong with DCIM and why is the "START" not the focus of the user experience.  One of the tips shared to my data center friend from one of the vendors is the VC community is pushing for more features to compete against the other DCIM products.  But, no users are asking for these features.  Yeh, who is screaming for more features.

Focusing on "START" was key to Microsoft's success.  If Microsoft had tried to demonstrate and market all Windows 95 features, then users would probably be confused and not know where to start.

When you buy your DCIM solution do you know where to start?

Who drives your data center? Finance or Operations

Most of you are hard core data center folks.  Operating a data center can be a pain or put you in the "zone" so you can focus on bigger issues to run the data center.  A good analogy is like driving a car.  How many of you would like to drive a data center that is designed by a bunch of finance guys?  This issue is illustrated by Bob Lutz in "Car Guys vs. Bean Counters."  BusinessWeek reviews the book.

In Car Guys, Lutz argues that Detroit’s steady decline can be blamed on the fact that there aren’t enough Bob Lutzes anymore. After legendary designer and car-guy’s-car-guy Bill Mitchell retired as GM’s design chief in 1977, Lutz writes, the balance of power—at the company, in particular, and in Detroit, in general—began shifting from the car guys to the number crunchers. As a consequence, product planners determined which customers to target with a new sedan or wagon; engineers fretted over inexpensive assembly; and managers fretted about cheap mass production. Only at the end were designers summoned to wrap a steel body around a nearly completed vehicle.

The results, Lutz laments, were the not-so-fondly remembered Cadillac Cimarron, GMC Envoy XUV, Pontiac Aztek, and others.

How many of you have walked into a data center and you can tell the data center was driven primarily by number crunchers who didn't have a clue what the electrical, mechanical, or operational issues are.  They have a budget.  Hit it.

What the finance guys miss is you can't reduce outages by saying you X nine's of availability.  One of the top things that is going to affect your outage is operations.  And Operations is going to be #1 when it comes to mean time to repair to reduce the outage time.

Now there are some technical staff who are like a Tim Allen character who like the "sweetest" tools that give them a jolt of adrenaline, ups their testosterone level, and a new thing they can brag about driving envy in their friends.  

There is a balance of designing, and being fiscally responsible.  The days of the over designed data centers are in the past and are being phased out as too expensive.

Part of the problem though for an Operations guy vs. a Finance guy is almost always the Finance guy is better at the company politics.

Lutz: This is the downside of being a creative person who does not play the political game too well. If I had,for instance, been a little bit more circumspect in my dealings with Lee Iacocca and perhaps had held my mouth, I might well have been his successor at Chrysler Corporation.

Ryssdal: Seems kind of an easy answer: You shot your mouth too much.

Lutz: I tend to be a person, when I don't know something I say, "I don't know, I'll have to look it up." I think boards like a CEO who is totally buttoned up, has all the figures. People with my personality generally don't make CEO.

How many Data Center Operations Guys are good at politics?

Can you see the Changes in the Data Center industry? We all suffer from Change Blindness

I was sitting down with a data center executive for lunch yesterday and we chatted about a bunch of different things.  Much of what we shared was observations and perspectives on what is going on in the data center industry.

One of the things we laughed about is the number of people, a very large number of the people who based their perception of what is going based on in the data center industry on what they read.  Whenever we read something that doesn't make sense, we can check with friends, or just know the news doesn't make jive with other data we have.

Part of what contributes to being able to recognize the truth from fiction is whether you understand how we all suffer from Change Blindness.  Change Blindness is many times worse inside a company where people don't get out much, and compare their observations with others.

Change Blindness

• Posted 01.27.11
• NOVA scienceNOW

Psychologists who study the fascinating phenomenon of change blindness know that merely looking at something is not the same as actively paying attention to it. As the demonstration in this video shows, people can be blind to significant changes in a visual scene that are obvious to someone who expects that these changes are going to happen.

LAUNCH VIDEORunning Time: 00:51

 

The data center growth that doesn't get reported, the big players are growing much faster

I read this Network Computing article that was reporting on an Uptime Study.

Data Center Study: The Big Get Bigger

Posted by Kurt Marko
August 14, 2013

Business success goes to those that can strategically and efficiently wield technology, and in this data saturated and hyper-connected age, that requires data centers. The latest Uptime Institute Data Center Industry Survey demonstrates that scale matters and operating data centers and computer rooms, which was never a task for amateurs, is increasingly the realm of those that make data centers their core business.

Uptime's survey, with responses from 1,000 data center facilities operators, IT managers and senior executives from around the globe, shows data center operators are expecting healthy budgets, with nearly a third in the U.S. and Europe seeing increases of 10% or more. Most of the bump is driven by third-party operators, which the Uptime Institute defines as "companies that provide computing capacity as a service in any form: Software as a Service, cloud computing, multi-tenant colocation, or wholesale data center providers."

The really big players Google, Microsoft, Amazon.com, Facebook, and Apple are not going to fill out out an Uptime Institute Survey to report their growth.  I define the BIG as those who run 100,000 servers plus.

Netcraft reports on AWS's growth of 10% in 4 months.

Surveys are good for those who don't know.  For those who know they would laugh at growth of the big players estimated in the 10% + range.