Long Term Customer Satisfaction in a Data Center who are you focused on?

Here is a question.  Who are you focused on if you want to achieve long term customer satisfaction of a Data Center build or lease?

Most would focus on the decision makers of the initial project.  But, too many times the people who start the project are not the ones who live with the decision made.  And worse case the team making the initial data center choices are optimizing for their budget and internal visibility vs. the long term cost, operations, and availability of the data center.  Any problems in operations can easily be diverted by saying that the operations team is at fault, and the design was perfect.

I always watch out for those who make it seem like their designs are perfect and don't have issues.  Any good design has trade-offs.  And, some of those trade-offs may not be the ones you may make.  A high availability data center will have higher costs to build and operate.  An energy efficient design may have higher inlet temperatures which makes it hard for legacy systems to be accommodated.  There is no perfect car.  Especially for everyone.  There are no perfect data centers.  People are most proud of their acquisition within the first months and they talk about how it is the best data center as if they are Donald Trump showing his latest building.  After a year the novelty wears off.  

Except…  There are a set of people that will show off their data center years after it was commissioned.

Who?  The operations team who take pride in their work.  Those who had an active role during construction and have a loud voice in operations are way more likely to be proud of their data centers.  These are the people who will tell their peers about the vendors used, procedures, best practices, and the issues they have run into.  

If people spend more time focusing on the data center operations team then there is a good chance you'll increase customer satisfaction.

In the data center industry the big are getting bigger.  The small are folding their operations into the cloud.  The middle is silent as they get squeezed in markets, margin, and find it hard to compete. In this shift, the role of data center operations will grow.

Acquire God like Powers to debug your Server Code - Takipi

One of my developer friends just deployed Takipi to debug server code.

God Mode in Production Code

Takipi tells you why code crashes and threads freeze. Server debugging doesn't have to suck.

Here is a video of the product.

A different way for DCIM, drop the M - Data Center Infrastructure

It has been interesting watching the DCIM market develop.  I designed a DCIM solution 4 years ago and what I thought was obvious I find few people would do.

Given the popularity of my post on DCIM not living up to expectations. I'll write a bit more.

One good rule is to follow what is articulated and followed by Google's demi god of infrastructure Jeff Dean who reports to THE god of google's infrastructure Urs Hoelzle.

One of Jeff Dean's demi god powers is.

If Dean has a superhuman power, then, it’s not the ability to do things perfectly in an instant. It’s the power to prioritize and optimize and deal in orders of magnitude. Put another way, it’s the power to recognize an opportunity to do something pretty well in far less time than it would take to do it perfectly. In Silicon Valley, that’s much cooler than shooting cowboys with an Uzi.

DCIM. Data Center Infrastructure Management.  Sometimes I think the solution would be better solved if the M - Management was dropped.  What is needed in Data Center Infrastructure.

What are the real problems in the Infrastructure that need to be solved.  Jeff Dean's #1 rule is the necessity of solving a problem.

I think one of the things that have caused us to build infrastructure as we were often doing things out of necessity, so we would be running into problems where we needed some infrastructure that would solve that problem in a way that could make it so that it can scale to deal with larger amounts of data or larger amounts of requests volumes and all of these kinds of things. There’s nothing like necessity of needing to do something to cause you to come up with abstractions that help you break through the forms. So map reduce was born out of needing to scale our indexing system.

I think I have a new bar joke "What is the necessity of management in DCIM?"  DCIM solutions would be so much better if the decision on what DCIM is deployed was made by Operations Team and not the Management Team.

Would you rather operate an air cooled rack or an oil immersion rack?

I regularly have people mention how immersion cooling is interesting.  But, almost all these people are outsiders, not people who work in data centers.   Poll question how many of you want to spend the day with mineral oil soaked components?  I don't see any hands up.

DCK just had a post on CGG's use of Green Revolution Cooling.

This solution must have looked good on paper to someone for CGG.  What do the operations guys think?

When you let operations have a voice what looks good on paper will many times have issues.

Arstechnica just posted a tour of Facebook's HW group.

Could you imagine 100,000 servers immersed in oil at Facebook.

The operating costs for an immersion system are an unknown and there is no way it will deliver a 50% reduction in a Facebook data center.

A skeptical view of Zynga's new CEO, Om Malik explains

Om Malik posts on GigaOm on his views of Zynga's move to add a new CEO to solve its problems.  Om explains his logic and starts off questioning the PR spin.

SUMMARY:

Zynga has hired Don Mattrick as its new CEO, replacing founder Mark Pincus. The stock has jumped almost 20 percent in two days. Everyone believes that things will get better — that is, everyone except me. Let me explain why I don’t buy the PR spin.

Om throws some zingers like the new CEO doesn't know mobile.

The glowing press releases not withstanding, Zynga isn’t any step closer to solving its number one challenge — mobile. The company’s business tactics, which worked dramatically well on the Facebook platform, don’t work as well on mobile. And frankly, what does Mattrick really know about mobile? For God’s sake, the guy ran Xbox for six years.

Zynga was built on the Facebook ecosystem which limits its ability to grow beyond.

From the very beginning Zynga has been a company optimized for short-term gains. It used the well-established pattern of super fast release, iterate, re-release to grow its games. That pattern of developing and releasing games works well on the web. Of course, that means sub-par creativity and leads to the short shelf life of a game. Of course, this resulted in a business logic — user acquisition channels, game development methods and technology stack — were optimized for one platform: Facebook. The company rewarded teams that build Facebook hits.

We'll see what Zynga looks like in a year, but that is an eternity in the game business.

It is so pervasive and embedded that unless Mattrick undertakes a company wide apheresis he is destined to fail. And with Pincus still as the chairman and chief product officer, you know nothing really has changed.