Data Center Pulse Blogs
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.
Definition of Modularity:
- Designed with standardized units or dimensions, as for easy assembly and repair or flexible arrangement and use: modular furniture; (borrowed from http://www.thefreedictionary.com/modularity)
Interestingly most people ascribe the definition above to the building of modular data centers. Sadly, this definition doesn't do a modern data center any justice, nor does it buy the data center owner what they really need.
In simple terms many of us assume a modular data center is one that can be built incrementally with standardized blocks of capacity. Building data centers is a CapEx intensive endeavor, so anything you can do to reduce your exposure is a good thing. However, if the only thing you consider when building your data center is whether you have enough "space", then you're missing several critical areas of opportunity and cost management.
Data Centers are first and foremost warehouses, highly technical and very expensive warehouses, but warehouses all the same. In a typical warehouse for furniture you might not have anything more critical from a functional design characteristic than how big the doors are and how high the ceiling is. Once you've got those critical considerations accommodated then it's just a bunch of common use space. What if the warehouse were a multipurpose facility that had different security and temperature requirements? Now try to build more space on this facility with a single variety building block. You'll start to see that since your capacity for refrigerated space was very small, but your high security area was large you end up with a mismatch of need, which equals wasted space and wasted cash.
With today's "modular" data centers you're buying chunks of data center capacity that are all built to a specific standard
Example: Tier III and 5kW per cabinet (or 300 Watts per square foot)
A standard building block of capacity is great if you always need that exact standard. What happens when you need a portion of space that can efficiently handle 25kW per cabinet or maybe you need 50 cabinets of Tier IV. Now you're beginning to see where the efficiency assumptions in many of today's modular data centers begin to let you down. But the issue is really more complex than my simple analogy above, creating a modular data center means that you have direct management of the expansion or contraction of any of the primary service capabilities of your facility. These service capabilities include, power density, cooling, space, and physical durability among others. How could you possibly future proof your facility for a 15 - 20 year lifespan if the day it's built it has the same capabilities it will have when it's retired. Do you really want to own another large data center that's "full", but has lots of empty cabinet and floor space?
So when you're shopping for data center space, do your homework and make sure you give the often used but rarely supported modularity message a real test before you buy.
Earlier this week we pre-announced that Ebay will be launching another public Modular Data Center RFP through Data Center Pulse. This is the second round of the public RFP process. Project Mercury, which was the result of the first public RFP, will finish commissioning by the end of this month and will be fully operational by October. Today we formally announce project Quicksilver. Quicksilver is Liquid Metal Mercury that moves and changes very quickly. Besides the obvious play on the Mercury name, we picked this name because it represents the capability we are looking for in data center portfolio. How can we create a generic, flexible data center infrastructure that can move and change with our business needs?
As you will see, we are taking the public RFP to the next level. The video below gives more insight into the project by describing the Scope, Requirements, Process and Schedule.
You can watch the project page for all of the updates as we go through this journey. My team, partners and suppliers have done incredible things in Project Mercury. I look forward to the next phase in our evolution as we execute Project Quicksilver.
Interested parties, should email email@example.com for more information.
Let the new battle begin!
It seems like forever since I have had a chance to blog! Needless to say, we've been absolutely swamped with business growth and pushing innovation as far as we can take it!
One year ago this month we tried an innovative modular RFP process which opened up the design of the new eBay Phoenix Data Center to the industry. As I write this blog, we are knee deep in the commissioning of this ground breaking design dubbed Project Mercury. The challenge we put out through Data Center Pulse has yielded one of our most innovative designs to date. The goal was to unleash the creative minds in the design and consulting arena by outlining the business and technical challenges then letting them tackle the "how". I am proud to announce that the process works. It works very well! The collaborative, partnership nature of this project has made it one of the best I have ever worked on. Barriers were shattered, competitors became partners, and the impossible became possible while the project rapidly evolved and our design requirements were exceeded. But I digress! This blog entry is not to announce the details of Project Mercury (more on that in Oct/Nov as we open it up). This blog is a heads up to the industry that eBay will be kicking off round two of the Modular RFP process! But this time, we're taking it to the next level - Salt Lake City, Phase II!
The process will begin August 19, 2011! Let the design competition begin! Stay tuned to the Data Center Pulse YouTube channel and the modular RFP page for more information. For more information, please email firstname.lastname@example.org.
It was a little less than a year and a half ago when I made the announcement that I was moving to ServiceMesh, now I've made another change. What the heck, do I have ADD? Well, I do have ADD, but that's not why I've made a move to a new company.
I joined ServiceMesh because I believed in their vision for cloud management and I still do. During my time there I was able to delve deeply into the world of cloud in general, and specifically as it applied to large enterprises. ServiceMesh was on to something when I joined, and that's not changed. They still have an amazing story in the cloud management space and it's only improving.
So How Did This Change Come About?
As part of my role at ServiceMesh I would occasionally work with partners, helping to develop an adoption strategy. Over the last few months I've been leading a project to do just that at Switch . It was during this project that it became obvious that my data center background, combined with cloud experience was the perfect fit to help Switch achieve its goals. So, while I have moved from ServiceMesh to Switch, it was more like an "employee" transfer than a resignation. At Switch I will have several responsibilities related to Data Center Tech, in combination with ensuring that the solution vision shared by ServiceMesh and Switch is realized. So, in effect, I'm working for both companies now.
Needless to say, I'm extremely excited about my new role. I'm back closer to the data center again, and I still get to play seriously in the cloud. The move to Switch was made easier by the fact that I believe they have the best data center solution on the market. So, as a data center guy, how could I resist.
If you find yourself in Vegas and would like to chat, be sure to look me up.
Ownership has several important connotations and I use it to define my take on personal responsibility for pretty much every aspect of my life. However, it can also mean a "systems" approach to "owning" all aspects of a specific service, solution or function (I.e., I own the Data Center top to bottom). While both of the previous "ownership" definitions are positive, there is a "darker" aspect of owning (hugging) "things" in IT.
The audible symptoms of being Positive for Ownership Disease
I've been in IT for over 20 years now and I've seen all the symptoms as we move from one service strategy to another or make an effort to transform the technical architecture of a component of the infrastructure. Maybe some of you will recognize the following paraphrased quotes that are examples of the symptoms of ownership disease;
"We can't move to client server, there's no way it will ever be as rock solid as big iron"
"No, Mr. CFO, we shouldn't move to VoIP, it's not ready for production. Our current switch has been running without interruption for years, why take the risk?"
"Virtualization isn't production yet, when you really want to ensure performance you have to stay on hardware"
"If you want storage performance and dependability it has to be SAN, and it has to be from (three letter acronym here)"
"We can't use outside air for the data center. It's too dirty and it will increase the failure rate of our hardware"
In it's current form the disease is most often identified through the following phrase:
"We shouldn't adopt cloud, it's not secure"
The reality is that in any one of the above quotes there is just enough truth to scare the unaware executive into staying the course. What is generally not explained is lost opportunity vs. a realistic risk value. I'm sure most of us can quickly look at these examples and see the issues (I.e., how important was 100% uptime on the legacy phone switch vs. saving $2 million a year by going with a VoIP solution or even outsourcing it?). In reality if the CEO/CFO understood the real risk value vs. the savings or business benefit, they would likely have said make the change.
You can call the above quotes myths, FUD, or lies, but in most cases it really boils down to "ownership" and the fear that not being the owner of something (that I know better than anyone else) will put my job at risk. The truly unfortunate aspect of this fear of job loss or at a minimum status reduction, is often generated or perpetuated (mostly inadvertently) by leadership.
The visible effects of being Ownership Disease Positive
There a number of places you can look to find the visible symptoms of the disease, but the most obvious is the delayed adoption of solutions that can bring real change to your business and consequently to your IT team. While it's true that technology should never be adopted for the sake of "technology", it's also true that real opportunity for improvements in cost, management agility, and business execution can be lost when the disease strikes. Let's look at a couple current examples:
Virtualization - As a commonly utilized building block for efficient IT, and improved execution, virtualization is also the most often used tool for on-ramping to the cloud. Companies that failed to make real investment in a strategy for virtualizing their environments now find themselves behind the curve in the adoption of cloud computing.
Data Centers - Being the beating heart of the IT organization, it's strange that the data center is often overlooked as nothing more than an expensive room that we occasionally have to throw lots of money at. Lack of adoption of things like virtualization or outside air will now mean that you're stuck with a beast that is at max capacity with only one quarter of the actual space utilized. You're also stuck with an inflexible design that won't allow you to quickly take advantage of new business opportunity or to just as quickly scale back to avoid wasting cash during a difficult business climate.
Why am I Blaming The Leadership?
Who do you blame when your favorite team doesn't win the championship? You might blame the team owner, or maybe there's one player you really dislike, but in most cases it falls on the coach. So, if your IT team isn't dealing effectively with ownership disease, you have to look to leadership first. What can we do to reduce the risk? This is the tricky part, I know I don't have all the answers, but the following are a few of my recommended strategies to help protect your team from becoming infected.
Recognition and Job Comfort (or freedom from fear)
I'm a huge believer in appropriate recognition:
Example Role: In a high impact support function with lots of customer interaction, getting regular positive feedback (daily) is critical. I'm also a believer that you find the job for the person. If you find yourself perpetually having to explain customer service principles to the same person, you need to find what they are good at. If you spend too much time trying to change negative behavior, instead of reinforcing positive behavior you end up with nothing. Why expend 80% of your energy to get a 20% solution when if you focus on natural abilities the good stuff practically happens by accident and even weirder is the employee likes their job.
When recognizing someone be sure that the message you convey is the right one. If you find yourself saying "Gosh George, without you knowing the phone switch so well, I don't know what we'd do" then you're reinforcing the wrong things.
Recognize your teams for working themselves out of a job. One of my favorite leaders used to tell me all the time "if you work yourself out of a job, I'll find you a better one" and he meant it. This particular form of recognition is the riskiest for the manager, because if you don't plan to follow through (yes that means you have to actually work at future proofing your team) you will immediately lose the trust of your folks.
Be sure to recognize individuals in the manner that they not you are most comfortable with. Yes, I know its novel, but everyone is different.
You must recognize taking "smart" risk as a positive behavior. Talking about it once a quarter isn't enough, it has to be visible and real (promotions, new job options, mentions, etc).
If you do the above and maintain regular communication (not just monthly 1:1s) with your staff, you're likely to build a strong team who is willing to speak out about waste and inefficiency, even if they're talking about their own function. The hard part is that you'll know you're successful when your team tells you when you've screwed up. When your team no longer believes that their future is tied to knowing a specific vendor technology, or architectural strategy then the natural fear we all have of change will be dramatically reduced.
Now go out there and vaccinate your team. Or if you recognize some weakness in your leader(s), then point them to this blog. You'll know your disease free when you're the first one to say "we should consider reviewing alternatives" with full understanding that the alternatives could affect your role.
When I read about a data center like the new Facebook facility in Princeville Oregon being commissioned and learn of all the innovations I'm heartened at the headway our industry is making, but I'm also forced to think of an analogy. The Facebook facility is very much like the NASA space program, there's lots of great tech created, but it takes a while before Tang is in everyone's fridge. Ewe, I can still remember the taste of that orange colored vile brew.
This blog is in no way a negative on what Facebook has done, quite the contrary. This new facility is an excellent example of how real innovation can occur when you break down the assumptions that most of us operate under. Things like high temp or outside air being a problem, assuming what our vendors told us is "all there is" is true, etc., etc..
I led a team that built a very efficient facility for VMware in Wenatchee Washington two years ago. Many of the basic characteristics of the Facebook facility (not their IT equipment) mirror what we did in Wenatchee. We used outside air, we conserved water through a grey water system, we heated the offices with hot air from the servers and didn't use ducting. We also had hot air containment, no raised floor, and a modular design for build out of the larger pods and the smaller containment units. I'm no longer at VMware, so I don't know what the PUE is, but during commissioning and first use, we were seeing 1.25 or less as the expected efficiency. However, the point is that Facebook has taken several known opportunities and improved on them and they've pushed the boundaries on equipment design with their partners and suppliers.
Why won't the Facebook design apply to everyone?
The Facebook design won't apply to everyone, just like it probably doesn't apply for some of Facebook's own IT application environments. The variety of hardware and legacy application and physical architectures in most large IT shops mean that it's a non starter to consider building something that is one size fits all. That being said, it doesn't mean there aren't one size fits all environments, they just aren't designed to the same efficiency ratings being claimed by Facebook. Also, besides the fact that Facebook can buy large numbers (1000s at a time) of servers with every order, they can also buy the same kind of server. The goal of homogeneity is still extremely elusive to enterprise IT environments.
What are the positive learning's to take away from the Facebook solution?
- Higher temp in the data center is OK. If you're still running your facility at the standard 68-72 degrees F, you're wasting a bunch of energy.
- Using outside air is being proved out yet again. After some early adopters began using it as early as 5-6 years ago, we're finally starting to accept it as a fact.
- You can and should push back on your suppliers to give you gear that does the job without being wasteful.
- Reduce packaging
- Eliminate unnecessary additions to servers that don't add to functionality, efficiency or availability
- Demand higher efficiency power supplies
- Look for modularity in virtually everything you implement
- Server design
- Data Center building design
- Power distribution
- UPS capacity
- Network design
When you push your suppliers you'll be surprised what you can get. But remember, you have to know what you need and why you need it or others will define what you need for you.
Most of us know what to do, and we just have to decide to do it. Just remember that even the coolest sounding efficiency benefit can sometimes cost more to implement than you'll get in reduced energy or management costs, so do your homework.
I'd like to close by saying that this Facebook data center generally supports the message in several of my previous blogs (Manufactured Data Center and Cookie Cutter Data Center ). As data center builders, many of us hold on to our creations like they are our personal Frankenstein, it is time to let go. The complexity of building, owning and operating your own facility effectively is just too much risk and overhead for the average IT organization and for the enterprise itself.
Paul Sun Director of Cloud Computing for ITRI Taiwan Joins the Data Center Pulse Board of Directors
Data Center Pulse expands the board of directors to continue the goal of influencing the datacenter industry through their exclusive, global end user community.
UNION CITY, CA, February 8th, 2011 - Today, Data Center Pulse added Paul Sun to the Board of Directors as the Asia Pacific Regional Director
Paul Sun is the Director of the Cloud Computing Center for Mobile Applications at ITRI (Industrial Technology Research Institute), www.itri.org.tw. ITRI is a national research organization that serves to strengthen the technological competitiveness of Taiwan. ITRI's 6,000 employees conduct advanced research and development in Communication and Optoelectronics, Precision Machinery and MEMS, Materials and Chemical Engineering, Biomedical Technology, Sustainable Development, and Nanotechnology.
Mr. Sun is responsible for cloud computing hardware and system infrastructure research and development at ITRI
Paul recently articulated his motivation for accepting the DCP board position. "The Data Center Pulse Board of Directors is made up of an outstanding team of veterans and experts from the data center industry. It is an honor for me to join and help to make DCP into a leading, international organization. Asia is experiencing tremendous growth in the data center industry. My main goal is to focus on the APAC countries, help to foster the exchange of ideas and needs in Asia as the DCP representative."
The Data Center Pulse core membership has reached 1,850 people in 62 countries representing almost 100 different industries. The interest level is increasing and the timing is right to focus more energy on our membership in the Asia Pacific region. With Paul in place, DCP will have a representative in a position to help gather information on local issues, while also representing DCP to new potential members.
Data Center Pulse has the ability to reach a significant population of Data Center customers ranging from a single rack to some of the largest Data Centers in the world. DCP continues to search for candidates to fill the remaining board positions as well as participation in the Industry Alignment Board (IAB) and the Technical Advisory Board (TAB). To become a Data Center Pulse Member, click here. For more information on DCP or local chapter interest, please email email@example.com.
About Data Center Pulse: Data Center Pulse (DCP) is a growing, non-profit, datacenter industry community founded on the principles of sharing best practices amongst its exclusive membership. Founded in late 2008, DCP is quickly becoming an industry nexus for the explosive datacenter industry's operators and influencers. DCP's mission is to align end users to share information thereby influencing the industry by defining, adopting and driving best practices and next generation solutions. The DCP members are the individuals that evaluate, recommend and purchase the products and services for the datacenter. They represent billions of dollars of annual purchases that drive the IT economy. Information is available at datacenterpulse.org.
As part of our partnership with The Green Grid, Data Center Pulse is assisting The Green Grid's Thermal Sub-work group by asking our DCP members to participate in the following survey regarding Economizers:
Please click on the link or paste it into your browser and participate. The survey is open until Feb 5th 2011.
Project Mercury is born...
Today, we are pleased to announce that EDI, along with their partners AHA Consulting Engineers and Winterstreet Architects, have been selected as the winner of the Modular Data Center RFP - now dubbed as Project Mercury.
This has been an extremely interesting process for us with an unexpected result. EDI, a small company that we had never even heard of before, was able to meet all of the challenging requirements we had proposed to the industry through the Modular RFP process in a cost effective, simple design. In addition, a very compelling ultra dense product named "eHive" emerged from Skanska, one of the RFP finalists. It has not been released publicly yet (stay tuned for follow up). While Skanska was not selected for the RFP, their modular product was innovative enough to warrant further consideration in this data center deployment. All in all, the open RFP process did exactly what we had hoped. It enabled design engineers the opportunity to shed the traditional barriers, consider the difficult challenges and start with a clean slate. The outcome was new and compelling solutions as well as new innovative products driven by the free cooling, density and flexibility requirements.
In the video below we go into more details about the process, the team and the challenge that still lies ahead. We made the decision to engage with EDI and explore the Skanska product in mid November. From that point the team has been working non-stop to make up the time we lost during the RFP process. While the selection process took almost two months longer than anticipated, the end date did not change. We are still laser focused on completing the Data Center by summer 2011.
In the design sessions that started in mid-November there has been lively discussions and debate where phrases such as "Hot Water Cooling","Extreme Density", "Rapid Deployment", "Multi-Tier", and "Rack & Roll" were common. Day one, I sat down with the team and tasked them to solve the challenge holistically, not just from a Data Center availability perspective. I detailed the IT equipment that will be going into the Data Cener and how it must be fully integrated with the facility. it is all about density, rapid deployment and sustained efficiency even under varying work load. This balance is where extreme efficiency can emerge. Data Center Facilities and IT equipment are not mutually exclusive, they should work in harmony.