Data Center Pulse Blogs


Center Pulse collaboration session of 2012 in conjunction with the Green Grid Technical Forum

It's that time of year again, Data Center Pulse will be holding it's annual Collaboration Session in conjunction with The Green Grid's Technical Forum at the Double Tree Hotel in San Jose, CA.

Data Center Pulse will be going over the Data Center Top 10 and presenting the findings to The Green Grid. We will also discuss some or all the following topics:

A new metric for "Useful work" to measure the performance and efficiency of your IT systems

The Green Grid Project Mercury Case Study

Impact of Cloud Computing on your Data Center Planning & Strategy

Other: Topics to be determined by Data Center Pulse members

The session is open and free to Data Center Pulse Members. You can submit your request to participate by sending an email to membership@datacenterpulse.org

If you aren't a member, but qualify (your role has a primary and direct responsibility for your company's data center). You still have time to sign up through LinkedIn groups. ( Search Groups: Data Center Pulse).

We look forward to sharing ideas and concerns with you on the 5th.

More Information on The Green Grid Event:

The Green Grid Forum 2012: Come and watch the Sparks fly!

Data Center 2025; how to save $11M by monitoring PUE; new metrics!; productivity of data centers. http://www.thegreengrid.org/events/TGGforum-2012.aspx

Data Center Pulse is meeting in the same location the day before. Members of DCP get a 10% discount off attendance, and if you're not a member of The Green Grid, you can "sneak in" by signing an NDA.

Come see the biggest event in business computing efficiency of the year! Mar 5 (DCP meeting), Mar 6-7 (The Green Grid Forum) http://www.thegreengrid.org/events/TGGforum-2012.aspx for more info.

 

Three Years Later

DCP was formed on September 8, 2009. It seems like only yesterday when we started this! Quite a bit has happened since that date. We have over 2200 core members in 66 countries representing over 1000 companies in almost every industry. Our original charter has remained the same; Influence the industry through the insight of the consumer - the data center owner/operator. Join Mark Thiele and I as we reflect back on the three years and address how we plan to get back to basics with DCP in 2012.

 

Is Public Cloud Computing Green – Or at least Greener than Traditional IT?

Unfortunately, there isn't a simple answer to the "Is Public Cloud Greener" question as the only real answer is "it depends".  At the core of the question is that assumption that because you're theoretically using fewer physical machines more effectively, that you are thereby greener or more efficient.  However, even if we stay with that assumption, when you dig into the details of what being "greener" really means, you'll see that there's still no easy answer.

In order for cloud to be greener, it has to accomplish two things at a very high level:

  • Use less power, people, and resource than it's none cloud equivalent to do the same work
  • And, use power that is at least as clean, if not cleaner than the power you were using

While many of us make the assumption that the first bullet is true, we really have no way of knowing whether the second one is, unless the provider offers this information in some verifiable and transparent way.

The above discussion and the detail that supports it are covered very well in a recent blog/presentation done by a friend and fellow Data Center Pulse member Tom Raftery.  I highly recommend spending a few minutes with Tom's blog and presentation if you'd like to increase your understanding of the nuances of setting expectations with SPs, and creating or managing sustainable and green IT solutions.

http://greenmonk.net/is-cloud-computing-green/

Great job Tom, keep up the good work.

 

Why Enterprises Will Force Down the Cost of Virtualization

Force it's such a strong word, but it definitely applies if my reasoning has any basis in reality.

The point of this blog is to illustrate that regardless of the intrinsic value of any given tool in the IT tool chest, once it's implemented it becomes a cost. Now, please don't read this the wrong way, I'm not an advocate of the thinking that IT is merely a place that helps us cut the cost of IT. What I am saying is that once you've effectively implemented a new technology solution, managing the on-going costs becomes a factor in your survival.

In this case the target area is IT server infrastructure costs:

Servers

O/S

Power, Space, & Cooling

Virtualization

Management tools & Security

People

Taxes

Etc

Historically virtualization has been assumed to be a great technology to be used to reduce your overhead costs in the following areas:

Servers

O/S (?)

Power, Space & Cooling

People

Taxes

Notice that the following cost areas are specifically excluded

Virtualization

Management tools & Security

The above two items were left out of the "reduce your overhead costs" bullets because you didn't have virtualization and your management and security costs likely went up or at best stayed the same.  Now you probably asking, "so why is that important?"  It's important because, whether you actually reduced your total cost of ownership or have kept it the same (virtualization benefits), you now have a new cost basis.  Only now, a much larger percentage of your cost basis is the cost of the virtualization, where it used to be the servers.  Let's look at the issue another way using a simple example (Just a rough approximation):

Table 1: Infrastructure cost distribution as a percentage prior to virtualization

Servers

Management & Security

O/S & Tools

People

Power Space & Cooling

30%

10%

10%

20%

30%

 

Table 2: Infrastructure cost distribution as a percentage after virtualization

Servers

Virtualization

Mgmt & Security

O/S & Tools

People

Power Space & Cooling

10%

30%

15%

10%

15%

20%

 

The above tables are only meant to be a means to show that the costs that make up your infrastructure have shifted. The shift in costs is important if for no other reason than the simple fact that the areas with the biggest percentage of cost are also the areas that get targeted the most for savings. As some might say "a target of opportunity". 

How virtualization becoming a target will force the issue

In closure, I'm simply attempting to illustrate the fact that no one in the business cares what we in IT did yesterday to reduce costs. What the business only does care about is what we're doing today. If you reduced your server costs last year finance doesn't see a line item that says "IT Guy Mark Thiele saved us 100K last year by eliminating fifty servers, so this year we leave him alone". What the CIO & finance person do see is what it costs to run IT today.  Unfortunately, that puts virtualization square in the cross hairs as a target of opportunity.

If you believe any of what I've written here, then you're likely to see why I believe customers will become more and more likely to introduce new tools into their environment. New tools invariably means multiple hypervisors, which in turn will force the dominate vendors in the space to review their pricing practices. 

Authors note:  The above aside, there are other reasons to have more than one hypervisor in your environment, as certain functions and workloads have still proven to work more effectively with specific hypervisors.

 

A Bus To The Future

 

I'll get straight to the point. I have a simple request directed to the personal and professional contacts I have established over the last 23 years. I need your help to raise $25,000 to buy a school bus for Dalit children in Patna, India. You can make your tax deductible donation through this PayPal Link:

 

Huh? Why do you need to buy a school bus for an area literally on the other side of the planet? Don't you just build and operate data centers for eBay? Yes, my day job is to build and operate the foundation that enables 100 million people to buy and sell products all over the planet...but there is something else that is equally as important. Just like Data Centers are the foundation that allow on-line businesses like eBay to grow, compete, and thrive - Education is the foundation that allows CHILDREN to grow, compete, and thrive. If children do not have access to education, then they are always at a disadvantage.

Let me explain my connection here. Earlier this year, thieves in Northern India stole a bus from the Emmanuel School in Patna. I'm sure the first question coming to your mind is why couldn't they just use their insurance to get it replaced? Sure, that sounds logical, but that's not how it works in poverty stricken areas like Patna. It can take more than nine months to get a claim paid. The payment would also not pay for a new bus. The reason this theft is personal to me is that three years ago a group of my family, friends and business contacts bought the school bus that was stolen. That simple gift enabled hundreds of extremely poor kids from remote villages to go to school. My wonderfully giving group of friends helped those kids have a shot at the future. Hundreds of Dalit children used that bus to get to the only school that would take them. Sounds like an exaggerated story, but it is not. I met many of these kids personally when I visited Patna with my mom. I know it has made a big difference in their lives. You can see the original website we published here. (please do not use the links there to donate, just use PayPal button above)

Normally, I keep my personal and professional lives separate, but I believe this is story will resonate with many in my network. We're not just going to replace the stolen bus, we're going to buy a larger one to get even more kids to school!  If you want to be a part of this, just click the PayPal donate button and give what ever you can. You can also keep up to date on the progress through our new website, Just Let Me Learn.

 

It is tax deductible and I will be including your name (and your company if applicable) in the new plaque we will give to the school leaders when we purchase the bus (you can also opt out of the plaque if you wish). Remember, that most companies match charitable donations by employees, so keep that in mind if you are donating personally. For corporations, please contact me directly (dean.nelson@mac.com) if you need a differnet payment method to donate. My sincerest thanks in advance!

 

For those who want more background first, read on...

In 2007, I took my mom with me on a business trip to India. Her name is Sandy and she lives in the mountains of Colorado. She had never traveled outside the US before, so India was going to be an eye-opener, a BIG eye-opener! Interestingly enough, before we left, she found a number of missionaries in India over the Internet. Yes, she found people she had never met before in another country and coordinated to come visit them in their homes while I was working! That's just how she is...a tough (and stubborn) mountain woman who takes care of herself (and I love her for it). Needless to say, I explained that flying around India isn't like flying around the US. First, a 67 year old grandmother, who knows zilch about the language, culture, customs, transportation, food or any other aspect of the country doesn't just travel around alone like she's taking a trip to Starbucks. Secondly, she's my mom! I would not let her travel alone. Thankfully, we worked out a l schedule that worked for us to travel everywhere together. 

To make a long story short, it was life changing for my mom...and for me. The people of India, specifically the Dalit children and those helping them, profoundly touched us. You can see more on that in the original website I put up here. When we returned, I felt compelled to help with one immediate need. That need was to raise funds to buy a bus to get kids to the school in Patna. These kids lived in extreme poverty and had no way to get to school because they lived so far away and their parents couldn't afford to get them there. The bus we purchased gave them that opportunity. I was really pleased with the outpouring of support I received from my friends and business connections. We raised the money in less that 8 weeks. We bought the school bus and helped hundreds of kids in multiple grades get to school for more than three years. I was so thankful to all those who donated to make this happen!

But as you read earlier, the bus was stolen. It was a crushing blow to a school struggling to make ends meet. The school had to scrape up money to rent a bus to keep giving these kids the chance. A chance to raise themselves out of poverty through the gift of knowledge, the gift of an education. An education is something that many of us take for granted every day.

I look at my family, and how blessed we are. I remember back to the faces of the kids we met and the smiles they had being at that school! I want to make sure they, along with every future class, continues to have that opportunity.

Below is the PayPal donation link through the non-profit organization my mom and I setup for our work in India.

 

I sincerely appreciate any help you can provide! Please forward this blog to others you think would be interested in participating.

Happy Holidays,

Dean

 

Quicksilver Winner Announced

Today I am pleased to announced that the team of AHA Consulting Engineers & Winterstreet Architects have been selected to design Phase II of the eBay Data Center in Utah - Project Quicksilver.This was by far the toughest choice to date in this public RFP process. The five finalist teams that presented to us all had strong credentials, competent project teams and very creative designs. We took our time internally scrutinizing all of the elements of our score sheet to make our decision. As we have said in the past, we selected the team that best fit our needs at this moment in time. I am very confident that my peers will receive great benefits from seeing all five design approaches at the DCP readout session in Q1 of 2012 (stay tuned for updates). In my personal opinion, the public RFP design process has been extremely successful in driving creativity and innovation from a gamut of talented companies and individuals in the industry. The big ideas and design stretching requirements have given birth to entirely new ways to approach data center design in alignment with IT workload efficiencies. The industry is being influenced by companies like this gaining access to the challenges customers like eBay are facing in delivering value for their business. It is not just a data center facility that needs a low PUE. It is an engine connecting a 100 million users to do more than 60 billion dollars of transactions a year. IT is our business and the engine that powers it must be agile and tuned for performance and efficiency at every layer of the stack.

Below is a video introducing the project leaders and more on the process and high level design approach.

 

 

Please watch for updates on the Project Quicksilver project page.

For more information email: modular@ebay.com

Why in Spite of the Numbers Private Cloud Will be Prominent for Years to Come

 

Why in Spite of the Numbers Private Cloud Will be Prominent for Years to Come

Public cloud is the way forward, it's cheaper, has better scale, probably better security than most in house IT solutions and it allows your IT team to focus on business benefit higher up the stack. So, why am I saying Private Cloud will be here for some time to come?

Assessment/Predictions:

1.      Inappropriate Business Risk Evaluation:

a.      When it comes to protecting the company jewels, risks are to be avoided at all costs. Unfortunately, few organizations (biz or IT) have the maturity to effectively measure actual risk value. Considering the business has a very low appetite for risk to their IT systems, IT will be able to play on that risk for years to come, thereby avoiding any changes that involve moving critical IT systems/applications to Public Cloud. Lastly, there will be further missteps with the Public Cloud providers that will give internal IT all the ammunition they need to justify staying in-house.

2.      Market opportunity:

a.      Gartner predicts that the Cloud Market will be nearly $150 billion by 2014. While there are many groups making predictions (UBS, Gartner, Forrester, IDC, etc), generally speaking the prognosis is bullish to say the least. Why is the size of the market important? It's important because it drives behavior by the companies trying to maintain or grow their slice of the cloud revenue pie. Does anyone really believe that IBM, HP, Dell, NetApp, Cisco, EMC, VMware, Citrix & others are just going to lie down and let 4-5 "Amazons" be the world's clouds? They can't afford to ignore this change any more than a firefighter can ignore wind direction when fighting a forest fire. The vendors who have traditionally supported enterprise IT will do everything in their power to demonstrate the viability of private cloud. They will make it easier to setup and more efficient to own, all while heavily marketing the risks of going Public.

3.      IT isn't anywhere near being commodity yet:

a.      There has been much discussion about the "commoditization" of IT as a result of cloud. Specifically, that there won't be any need for internally built infrastructure anymore. That the future of IT will be to buy platform services from external providers and that those providers will be the ones worried about how to manage the physical assets. To some degree, this prediction is real, but I strongly believe it misses two things:

                                                    i.     The human equation: As humans still run IT, decisions about the future of IT organizations and IT roles, will be driven by "base" human needs and concerns, not purely by what might appear logical. As a leader, I learned very early that to ignore the human equation is to fail.

                                                   ii.     IT at all levels still has considerable development left in it. Even today we don't have everyone using white box servers, why is that? It's because the vendors are still offering differentiation that appeals to the buyer. I've said this before, but it bears repeating, "IT is like a painter's pallet. The painter may have the same brush, and the same colors, but will still create a new and unique painting every time".

4.      IT Organizations will respond:

a.      As IT organizations come to grips with the change Cloud brings to IT infrastructure, application ownership and organization design, they will start their own internal efforts to protect what they do. What's the best way to protect your job, innovate and improve, demonstrate how being internal and close to the business can still provide "perceived" value, regardless of what Amazon, HP, or Terramark are selling in the Cloud?

5.      10 Years to get 50% of IT into the public cloud:

a.      Over time IT organizations will change their job descriptions and ability to contribute at new levels. These role changes combined with public cloud providers becoming institutionalized will allow that move.  There will also be a slow retirement of traditional IT infrastructure specialists over the next 10 years and the majority of IT talent coming out of college will want to focus on other aspects of IT (I.e., application development, delivery and business representation). The above mentioned changes will naturally increase the percentage of IT work being done in public clouds.

6.      Within the next 10 years several large cloud providers of today will be gone or dying:

a.      This may seem like an obvious prediction, but it's not. My prediction is more related to cloud providers being caught in technology "change" traps that cost them their margins, as opposed to just being unable to execute an effective cloud offering.

I fully realize that there is controversy around this topic, but also believe that discussion and debate are the best ways to learn. Through sharing our concerns in an open forum more of us are likely to accept and internalize new ideas. In a similar vein to a blog I wrote a while back about Amazon I referenced the fact that yelling at us about what the right solution or option is isn't likely to make us change our behavior.  I look forward to your comments and tweets on this subject.

 

Decoding Data Center Modularity

The word du jour at least in the data center space is "Modularity". The only word used more often and loosely in the IT space is "Cloud". Even though the two words "Modularity" and "Cloud" are hyped, it doesn't mean there aren't real opportunities in both areas of technology. The trick is in understanding how the terms should be used, and where and how they should be applied. Come join me on Brightalk December 8th at 9:00 AM Pacific time A BrightTALK Channel

Quicksilver Finalists Selected

In technology, competition truly is the mother of innovation.

The first round of the Project Quicksilver competition is complete! Through the months of August and September we received 68 requests to participate in eBay's second public Data Center RFP codenamed Project Quicksilver. 61 companies qualified, and 20 submitted design proposals by the October 7th deadline. Today we have selected the 5 finalists based on a comprehensive and balanced scoring system that rated each companies submission based on the design concept, team capabilities, overall operational efficiency, sustainability and cost.

 

 

Our congratulations to the finalists (listed in alphabetical order):

  1. Advanced Design Consultants, Inc
  2. Team: Deerns & Gensler
  3. KlingStubbins
  4. M+W Group
  5. Team: Winterstreet Architects & AHA Consulting Engineers

Each of the finalists will be receiving a design fee for their efforts in reaching the finalists round. They will also have the opportunity to present their designs to the Data Center Pulse end user community in early 2012. DCP members, stay tuned for details on that upcoming event.

Once again, I am very impressed with the excitement, creativity, and thoroughness of the proposals we received. I want to personally thank all 61 of companies that embraced this challenge. Similar to Project Mercury, we received many comments from the participants expressing their apprecition on being able to participate without the constraints and traditional boundaries that are normally placed upon a Data Center design project. We also received a deluge of emails from others who are observing or would like to be incldued in all levels of the project. Throughout the process, participants were able to gain insight into how eBay, the worlds largest on-line marketplace, is addressing their uprecedneted growth and uber-efficiency requirements. The hundreds of design charrettes that have resulted from Project Mercury learnings coupled with the public RFP process through Data Center Pulse is truly breaking the mold on how the industry approaches Data Center design, consutruction, and operation. We want to thank everyone who has contributed their passion and sweat into making this possible.

In the video below, Mike Lewis, Director of Mission Critical Engineering, James Monahan, Partner at CDCDG and I discuss the Public RFP process, scoring and selection of the finalists for Project Quicksilver.

We will be selecting the winner of the Project Quicksilver design in November. Stay tuned to Project Quicksilver updates through the project page. For more information on Quicksilver and Mercury, please email modular@ebay.com.

 

Deep thoughts on Cloud SLA’s

OK, they aren't really deep thoughts so much as they are observations on the SLA assumptions made between provider and customer.

Earlier today there was a great back and forth on Twitter about SLA's and Cloud. Following are some of the associated Tweets:

@GeorgeDWatt
When SLAs are broken cash compensation may trump credits. It hurts more & if service is really bad may simplify exit. #CloudViews

@mthiele10
So much talk about Cloud SLA's. The real issue is how the provider measures outages vs. how the customer does. Penalties are immaterial.

@vambenepe
@mthiele10 Not being able to compensate for "actual loss" doesn't mean give up on defining any kind of penalty. We're beyond "eye for eye".

@GeorgeDWatt
MRT @mthiele10: ...Cloud SLA's. The real issue is how provider measures outages vs. how customer does. Penalties are immaterial. #CloudViews

The above Tweets and many others that occurred over a 30 minute period got my writing juices flowing, which is where this blog was born.

Assumption 1: We need SLA's, but maybe not for the commonly held belief of why

Cloud SLA's are really an opportunity for the customer to learn more about how the provider can protect them from risk. The process of building and agreeing to the SLA is also a necessary step for the customer to take in evaluating their own assumptions on the quality of existing internal architectural design. In other words, the customer must fully grasp and internalize the actual capabilities and design characteristics of their application and physical infrastructure prior to moving anything to a cloud provider. It's only through this self-evaluation will the customer determine if what the cloud provider is offering will meet the business requirements.

Key thought: Assumptions between IT and the business when the application was in-house will be put through a much more vigorous test when it's supported on a public cloud offering. In other words, mistakes that you might have been able to apologize your way through when the application is internal won't be so easy to ignore when they are more "public".

Assumption 2: Penalties are necessary as a way to measure success or failure

I believe my interpretation of "Penalties" is slightly different from many in the Cloud buyers community, but I could be wrong. It's also possible that I haven't read the right blogs.

As suggested in the above "assumption 2" statement, penalties should be used more as a measure of success in the relationship. The fact that you've moved a business critical function into a public cloud means that you need to create a partnership with your provider that mirrors any expectation of communication and responsiveness you would have as a buyer of an internal service. It's essential to realize that communication is the first and easiest process to break. If you were having trouble communicating with your internal teams and customers, moving an application into the public cloud won't simplify the issue, it will magnify it.

It may seem counterintuitive, but you really must fully understand how to manage and support your application effectively before you give it to someone else.  If you hand if off thinking "they'll take care of it", you're in big trouble already. The most common mistakes made in outsourcing are "doing it for the wrong reasons", and "not fixing the service before handing it off".

Assumption 3: A Cloud SLA should be considered a "work-in-progress"

Your SLA should be a mirror of your current needs for any number of common metrics:

  • Performance
    • By region, by application, by process, by instance, etc.
  • Scale needs
    • Up, down, how fast, lifecycle management
  • Cost
    • When, for what, who
  • Failover or recovery requirements
    • RTO, RPO, data locations
  • Merger & Acquisition efforts
    • Integration, new markets, new regions
  • Enterprise growth plans
  • Etc., etc.

Seeing as how all of the above metrics are likely to change as your company changes, you need to stay close to your new cloud provider partner to avoid prioritization miscues. So again, the SLA must be treated as a work in progress and a living document. If you haven't gone over your SLA with your customers and your provider partner in three months or longer, you're falling down on the job.

Key Points:

Your SLA should be a tool for helping you define requirements against "real world" needs and capabilities, while taking into account what you actually have versus what you're asking for.

Communication, communication, communication, it is the key to having a successful partnership, whether it's cloud or some other business arrangement. Build in regular and meaningful opportunities to communicate or find yourself surprised when expectations aren't met. This isn't something you leave up to someone else to do. If the provider isn't offering it, then demand it, if they can't support it, then you're with the wrong provider.

Your SLA with any provider, but especially a CSP, must be considered both a "work-in-progress" and a "living document".  As technology, the economy, and your business change, so should the SLA.