Is vendor lock-in a problem to solve or should you manage your purchases better?

CIO magazine has a post on the Open Compute Project and how it takes aim at vendor lock-in.

Facebook Data Center Project Takes Aim At Vendor Lock-in

Facebook's Open Compute Project aims to reduce 'gratuitous differentiation' among IT vendors

By Joab Jackson
Thu, October 27, 201

IDG News Service — By launching the Open Compute Project as a stand-alone foundation, Facebook is hoping to further standardize the market for data center equipment, thereby reducing costs and vendor lock-in, the new foundation's board members said at a press conference Thursday.

"What has been missing here has been standardization at the systems level," said Andreas Bechtolsheim, an OCP board member who co-founded Sun Microsystems and later was the first major investor in Google. He railed against "gratuitous differentiation" on the part of server vendors, who each have their own unique chassis or some other component that prevents users from intermingling the equipment with gear from other vendors.

But are vendors really the problem or is it how companies make their purchases the problem? The smart data center guys are looking at the hardware purchases and data center designs as a complete system and they need engineering talent in the company to design the systems and choose the right hardware.

You could blame the vendors, but they are doing what any company would do figuring out how to maximize their profits, and differentiate their products vs. others.

Are vendors reacting to the way customers buy or are vendors defining how customers buy?  Facebook is drawing attention to what happens if hardware designs are open sourced.  But as Kevin Heslin asked last week at the Open Compute Project, how does the foundation determine what is open sourced and what is not?

Is Open Source Hardware the answer or is it how companies purchase?  Think about it.  Buying Open Compute hardware changes the purchasing process.

Open Compute Project as a lower cost model, Rackspace and Facebook

Rackspace and Facebook are on the board of the Open Compute project.  SeekingAlpha has post on the open source process both of these companies embrace.

Rackspace Doubles Down On Open Source Process

Open source has been vital in creating the cloud. The open source process allowed many companies to use what began as Google's (GOOG) MapReduce, evolving it into things like Hadoop (originally a Yahoo (YHOO) project), entire cloud stacks like Red Hat's (RHT) OpenShift, and services like Amazon's (AMZN) EC2 in relatively short order.

The author drills to the bottom line of why Open Compute?  lower costs.

Strip away its adherence to the open source ethos, after all, and Rackspace is just a hosting vendor, no different from hundreds of other small hosting companies like Westhost or SoftLayer. At some point this competition will come down to costs, as usage of the resource scales exponentially. Low cost infrastructure will win.

Right now costs are not the key issue in the cloud. Right monetization is the issue. That's why every vendor – from Oracle (ORCL) to Dell (DELL) to Apple (AAPL) – can call what they do “cloud.” But as usage increases 10 times, then 100 times and 1,000 times, what seem like small differences in costs now will become magnified into big differences in profitability.

One way to look at Open Compute Project is it is an effort to build a lower cost data center ecosystem.

Open source, and the Open Compute project, put Rackspace at the forefront of these changes. They are key to its long term survival.

Google vs. Facebook European data center locations

One way to look at Google vs. Facebook European data center locations is to plot on a map the sites.

A - Google Dublin

B- Google Hamina, Finland

C - Google Belgium

D - Facebook Lulea, Sweden

Google has three major data center sites.  In Asia, Google has three future sites - Singapore, Hong Kong, and Taiwan.  Recognize a pattern?  Three is a good number, says GreenM3 :-)

Keep in mind given Google has built its three data centers it most likely looked at sites all over Europe before it picked its three locations.  Think about being in three sites as an option to building one mega site.  I watched an established company follow the advice of the so called data center experts justifying a single site as their first data center.  For less than 10 MW, it can make a lot of sense to have three 3 1/3 MW sites to enable geo redundancy of services.

Happy Halloween

Halloween is a great time to be with the kids.  After a week in NYC, I am ready for a day for candy, trick or treating, and following the kids as they run from house to house.

Happy Halloween!

Here are my kids 7 years ago, and they are just as happy.

What I am working on next? Product Life Cycle, Supply Chain, Asset Management

I ran into Kevin Heslin and Rich Miller at the Open Compute Project hosted by Facebook.  I figured I would run into Kevin and Rich given the NYC was in their backyard.  Luckily i had been planning a trip with some data center executives to the NYC area for months and it just so happened we were planning on being in NYC at the same time as the Open Compute Project which is also very convenient as many of us had gone to the Open Compute Project in Palo Alto.

Many people know me through my blogging role as I have now reached 4 years writing on the green data center topic.  Prior to that I spent 14 years at Microsoft working on mostly Windows from Win3.1, NT3.1, NT3.5, NT4.0, Windows 2000, and Windows XP on a whole bunch of different OS features.  TrueType font technology is where I spent the most time on a technology cumulatively between Microsoft and Apple.  Fonts is where I made some great friends I still work with today.

At Apple I re-engineered the distribution system to handle triple the capacity with a 1/3 of the labor by analyzing batches of orders to process them as a group instead of individually.  Some ideas stick and can be re-used over and over. After distribution logistics I moved on to OEM supply chain for Apple products, with the best being the Mac II, and even though it was a market flop it was fun traveling all over to get the parts for the Mac Portable.  I cut my teeth on operating systems for the Mac II and Portable, and moved to the OS group to work on KanjiTalk, then work on System 7.

Kevin Heslin asked me what I am busy working on.  I told him I am working on supply chain and asset management systems.  Kevin said, "I thought you were working on Green IT."  There are huge ways to Green IT addressing the supply chain in the data center ecosystem.  Analyzing the overall hardware flowing through the product life cycle will allow you to see things you can't see when staring at servers in a rack which pretty don't move for 3-4 years.  PUE and green technologies can reduce the carbon impact required in the power and cooling systems.  But, there is probably at least a 3X improvement potential in how the IT systems are put together in the data center

After 4 years of blogging on green data centers it is easy for me to write on the topic, spending on average an hour a day and I can write 3 posts.

There are other ideas i am working on like what does a location service look like in a supply chain system.  This stuff takes hundreds of hours over months to figure out, talking to some really smart people.

4 years ago I started working on green data center ideas, and it has been a great experience.  Now, I am spending more time on greening the IT systems, and solving the problems in different way with systems, knowledge, and data analytics.  The data center is a critical part of the solution, and I will still spend time on data centers, but I am more excited to think of new information systems for the data center.

I don't know if this answers the question Kevin asked, "what I am busy working on?"  And, most likely in 6 months there will be something new to study.  But, for now I am thinking about Supply Chain Management, Asset Management and Inventory Control.