N Plus One is on hiatus

The N Plus One blog is currently on hiatus. We encourage you to visit other Uptime Institute and The 451 Group offerings at the following sites:

Comments Off

Filed under Availability, Green Enterprise IT Awards, Optimization, Scalability, Sustainability, Symposium 2011, Technologies

Green IT Gets Real As The Hype Disperses

When it comes to hyped areas of IT, you’d be hard pressed to come up with two more impressive examples of technology marketing department excess than ‘cloud computing’ and ‘Green IT’. True, some of the momentum behind the whole green phenomenon has dissipated over the last 12 months, but cloud is still going strong and is probably at the “peak of the hype cycle” as some other analyst groups would have it.

While all the flak being thrown up around both subjects might serve the short term interests of certain suppliers, it’s actually rather detrimental to sensible debate and analysis of what are two very fundamental, and disruptive trends in enterprise IT.

While I wouldn’t go as far as Richard Stallman in dismissing everything bearing the cloud moniker as merely marketing, the old adage that less is more certainly holds true as far as cloud is concerned. More judicial use of the term would not only help to clarify its true meaning but also avoid the cold shiver the term now elicits from many in the industry – myself included.

The bubble around cloud computing has to burst, but green IT seems to have seen its fortunes wane of late. The term is no longer sandwiched into every supplier press release or marketing announcement going.

But behond the lack of hype, the concept of sustainable and energy efficient technology is just as solid as it was three years ago, if not more so.

Rising energy prices, concerns about nuclear safety, new carbon legislation such as the UK’s CRC Energy Efficiency scheme, continue to reinforce the idea that more IT performance per kWh is fundamental to the development of enterprise IT. What we have really witnessed is the marrow being sucked out of Green IT as a PR tool, with voracious marketeers casting aside one bone to wave another in the air.

Whether the problem is too much hype or not enough, the reality is that innovative projects around cloud and green IT are happening out there. But many are not getting the attention they deserve either because they are lost amid the flood of information around cloud computing, or suffering from the switch in fashions around green IT.

I am actually involved in a project at the moment which is focused on the intersection point between cloud and Eco-efficiency (The 451 group’s preferred term for energy and economically efficient IT).

Optimis is a EU project to develop an open and independent cloud brokerage platform which customers can use to connect with cloud service providers. One of the key aspects of Optimis is that it will allow customers of cloud services to assess service providers according to a number of criteria. Alongside trust, risk, and cost, customers will also be able to assess the energy efficiency of a particular cloud service.

Developing the technology to define cloud services in these terms is as close to the cutting edge as you can get without falling off. This is especially true in the area of energy efficiency where tasks that may appear straightforward are actually anything but. For example, it’s highly likely that service providers will be running applications or platforms on virtual machines. But working out the energy efficiency of an individual workload on a virtual machine is extremely complex. In fact the technology and models to do it don’t really exist, certainly not in any organized form at the moment.

One of the challenges and opportunities of working on an EU-backed project such as Optimis is being given the latitude to experiment and push boundaries, free from the commercial imperatives you would find inside most high-tech suppliers. True, innovative organisations such as Google have their 70:20:10 approach where the developers are allotted time to work on niche or personal projects, but that isn’t the norm.

The first code release, or code drop, from Optimis is due to be made public in June, so it will be great to be able to share our work so far. The project runs for another two years, and is being funded to the tune of €7m Euros, so there is still plenty of work to do.

However, when complete, Optimis will provide a way to assess the energy efficiency of cloud computing in a very granular way. Some of that will be achieved by developing a methodology to measure the energy efficiency of a virtual or physical workload, but we’ll also use existing metrics and schemes.

Power Usage Effectiveness (PUE) has its flaws and detractors but it’s still a useful measure of data centre energy efficiency. Likewise the EU Code of Conduct for data centres, and US EPA Energy Star rating all provide a shorthand for assessing the energy efficiency of cloud service supplier – or at least the infrastructure they are using.

Projects such as Optimis show that behind the hype, technology trends such as green IT and cloud computing are having a very real impact on the IT is developed, deployed, and paid-for. Marketing plays an important role in promoting genuine innovations but it’s important that it doesn’t end up trampling over the very things it’s trying to promote. It’s fine to screen out the hype as long as you don’t ignore the real achievements in the process.

This article was first published on eWEEK Europe UK.

By Andrew Donoghue. 

1 Comment

Filed under Green Enterprise IT Awards, Sustainability

GEIT 2011 winners and finalists announced

Uptime Institute has just announced the winners of this year’s Green Enterprise IT (GEIT) Awards, which recognize outstanding projects for improved energy and resource productivity in IT and data center operations. There are awards in several categories, to recognize different kinds of innovation.

This year’s winners are:

Category Winner Project title
Audacious Idea Harris Corporation and Lee Technologies, a Schneider Electric Company Multi-Zone Water Containment: A Breakthrough in Site Sustainability
Beyond the Data Center Tieto Distributed Heating Using Data Center Waste Heat
Data Center Design Capgemini UK and Red Engineering Design Merlin, the World’s Most Sustainable Data Center
Facilities Innovation Kaiser Permanente Computer Room Functional Efficiency, a Study in Raised Floor Optimization
Outstanding Facilities Product in a User Deployment Opengate Data Systems and Agriculture and Agri-Food Canada Intelligent Containment – Beyond Hot and Cold
Innovation in a Smaller Data Center (> 1,000 sq ft) CEMEX Global Data Center – Smart Energy Management
Innovation in a Smaller Data Center (< 1,000 sq ft) University of Hertfordshire Reduction and Reuse of Energy in Institutional Data Centers (RARE-IDC)
IT Innovation AOL You’ve Got……AOL Cloud Computing
Outstanding IT Product in a User Deployment Viridity Software Multi-Billion Dollar Business Deploys Viridity DCIM Software – Saves Over $1M

.
.
The following companies are also honored as 2011 GEIT finalists:

  • Audacious Idea – Akamai Technologies – “Creating financial incentives for off-peak energy efficiency through metered-power data center contracts”
  • Beyond the Data Center – KPMG – “Sustainable IT next practice: Influencing employee travel patterns”
  • Data Center Design – Brocade Communications Systems – “Brocade Data Center Consolidation and Energy Efficiency”
  • Facilities Innocation – Yahoo! – “Yahoo! HQ Air-Side Economizer Retro-Fit Project”
  • Outstanding Facilities Product in a User Deployment – SynapSense and AT&T – “Wireless Automation Optimizes Data Center Cooling”
  • Outstanding IT Product in a User Deployment – HP Brasil and Bradesco – “HP Brasil and Bradesco moving together towards the green data center”

Case studies for these projects will be presented at Uptime Institute’s Symposium this May 9-12, 2011, in Santa Clara, CA.

Comments Off

Filed under Green Enterprise IT Awards, Symposium 2011

Compelling line up for Uptime Symposium

By Andy Lawrence. After nearly six months working on our speaker line up for the Uptime Symposium, we made our formal announcement of our main program last week. Our goal was to make this a seminal, milestone event in datacenter thinking, with speakers who are really in the forefront of the disruption of that is sweeping through our industry. Modestly, I think we’ve achieved this. In fact, my colleague John Stanley and I have discussed and in some cases previewed presentations with many of the speakers below and I can pretty well guarantee that if you are interested in anything to do with datacenters, you will find some of these presentations seriously challenging.

We have a second wave of speakers to announce, and this list here is not complete..there’s more on the Symposium site.  But hopefully you’ll agree its pretty compelling. Here’s just some…not in any particular order:

• Mark Ascolese, Chairman and CEO, Power Analytics (formerly EDSA Power Analytics)

• Christian Belady, GM, Data Center Research, Microsoft Corporation

• Rami Branitzky, Managing Director, SAP Labs North America, SAP

• Kenneth Brill, Founder, Uptime Institute

• Peter Panfil, VP and GM, Liebert AC Power

• Steve Carter, VP, Digital Infrastructure Services, Uptime Institute

• Neil Rasmussen, Chief Innovation Officer, Schneider Electric

• Rachel Chalmers, Research Director, Infrastructure Management, The 451 Group

• Rob Coupland, COO, TelecityGroup

• Bob Gilligan, VP, GE Digital Energy Services

• Jon Haas, Director, Eco-Technologies Programs Office, Intel Corporation

• Steve Hassell, President of Avocent, a Division of Emerson Network Power

• Chris Malone, Thermal Technologies Architect, Google

• Michael Manos, SVP of Technologies, AOL

• Dean Nelson, Senior Director, Global Datacenter Services, eBay

• Peter Panfil, VP and GM, Liebert AC Power

• Jay Park, Director of Data Center Engineering, Facebook

• Antonio Piraino, VP and Research Director, Tier1 Research

• Vince Renaud, VP and Managing Principal, Uptime Institute Professional Services

• Andrew Stokes, Chief Infrastructure Architect, Deutsche Bank

• David Struhs, VP – Sustainability Strategy, C3

• Jack Tison, Chief Technology Officer, Panduit

• Jim Trout, President and CEO, Vantage Data Centers

• George Slessman, CEO, i/o Data Centers

• Byron Washom, Director of Strategic Energy Initiatives, UC San Dieg

• Andy Lawrence, Research Director, Datacenter Technologies, The 451 Group

The conference, themed The Disrupted Data Center: Cloud, Cost, Capacity and Carbon, will focus on strategies and best practices for anticipating and responding to disruptive changes such as cloud computing, virtualization, volatile demand and rapidly evolving modular architectures. Symposium is scheduled for May 9-12, 2011, at the Santa Clara Convention Center in Santa Clara, California.

You can still register for the event here.

Comments Off

Filed under Uncategorized

Want an EPA ENERGY STAR rating? Better get your meter in place ASAP

In early March at the Green Grid Technical Forum, Mike Zatz of the EPA pointed out that data centers will need IT energy meters in place by June 2011 (about three months from now) in order to receive an ENERGY STAR for data centers rating after June 2012.

The requirement stems from the fact that in June 2012, the ENERGY STAR program will begin requiring 12 consecutive months of IT energy data measured at the UPS output. This means that data centers will need to start measuring by June 2011. Estimates were acceptable as the program was first getting off the ground, but that option expires in June 2012. If I remember correctly, it also means that any facility currently certified will see its certification evaporate in next June if its input data is still based on estimates at that time.

The EPA presentation is publicly available on Green Grid’s website.

The current list of ENERGY STAR data centers is also online.

–by John Stanley

Comments Off

Filed under Legislation, Sustainability

Disconnect your datacenter for Earth Hour

It’s Earth Hour time again. Tomorrow night (Saturday 8.30 to 9.30 pm, March 26th) we are all being encouraged to switch off our lights, and other household appliances, for an hour to help save energy, and the planet to boot.

It’s a nice idea, and has produced some great photos and video of whole city blocks gone dark. It probably won’t change behaviour much though. Pretty much everyone is going to flick those lights and gadgets right back on again afters but at least it gets people thinking about the issue.

One company that has clearly been thinking is Candian datacenter operator Cogeco Data Services which has decided to go one step further with Earth Hour. The company has pledged to disconnect from the electrical grid during Earth Hour and run its DC from a natural gas generator back-up.

Now this might all be a way of turning a bit of resiliency testing into a marketing opportunity but hats off to them for thinking it up. By not using grid energy for an hour, the company claims it will reduce power consumption in Toronto by approximately 750kW, or the equivalent of 12,000 60-watt lightbulbs.

“Our unique ability to disconnect from the electrical grid and power our data centre using clean burning natural gas has a positive impact on the environment and we are looking forward to doing our part for Earth Hour. However, as the demand for cloud computing services grows and the need for colocation facilities increases, the IT industry must go above and beyond Earth Hour to create long term strategies and standards to reduce power consumption in data centres,” said Tony Ciciretto, President, Cogeco Data Services.

Nice sentiment. But the truth is that energy efficiency is still the poor neighbor of uptime and availability in the datacenter industry (I wonder what Cogeco told its customers when it thought up this stunt?). That is beginning to change, thanks more to rising fuel prices than any heart-felt environmental altruism, but it’s progress at least.

Cogeco’s Earth Hour stunt chimes with a comment piece I read on eWEEK Europe UK. Editor Peter Judge questioned why datacenters still mostly depend on diesel generators for back-up. The piece was a reaction to the recent events in Japan where the power grid was severely affected but Internet connectivity survived relatively unscathed.

The article points out one aspect of energy efficiency which often gets overlooked: “That is yet another good reason to make your data centre more energy-efficient. If you are burning less energy, your back-up will last longer.”

Being energy self-sufficient is another obvious answer to improving DC reliance. But we are still a long way from being able to trust renewables to that task. Still, at least efforts such as Earth Hour help to highlight the issues even if they don’t provide many answers.

By Andrew Donoghue

Comments Off

Filed under Availability, Sustainability

Green IT standards: EPEAT and UL SPC

About a month ago, I read a blog post about Lenovo’s ThinkPad T420 earning recognition under a new green IT certification scheme from Underwriters Laboratories, the  UL Environment Sustainable Production Certification (SPC).

“Wait a minute,” I said to myself. “I thought we already had a green IT certification, and I thought it was called EPEAT, from the Green Electronics Council.” Things would definitely get confusing for sustainability managers if there were two conflicting standards out there for green IT products.

Fortunately, upon closer inspection, the EPEAT and UL SPC standards don’t appear to conflict. Both standards are based on IEEE 1680, the “Standard for Environmental Assessment of Electronic Products.” (See the IEEE 1680 page, the IEEE 1680 group page, and the EPEAT notes and UL SPC notes on where their standards come from.)

However, the EPEAT and UL SPC certification processes are entirely separate, and have slightly different requirements, so it’s certainly possible for a product to end up with one stamp and not the other. For example, as of March 21, 2011 when I last checked, the ThinkPad T420 shows up in UL SPC’s registry, but not EPEAT’s. One big difference in the requirements is that UL tests every product before stamping it, while EPEAT allows manufacturers to self-certify products and then conducts random audits. In theory, this could give EPEAT an advantage in greater speed and flexibility (and therefore likely greater adoption), and it could give UL SPC an advantage in reduced risk of unwittingly certifying a product that isn’t up to snuff.

I can’t speculate on how much of a difference either of these theoretical advantages makes, though. I do know that EPEAT is still very credible. It’s good enough for the federal government, too–Federal agencies are required to purchase EPEAT-certified electronics when possible. (See Executive Order 13423 and the Federal Acquisition Regulations.) Similarly, UL has probably made its certification process as streamlined as possible, since it presumably wants high adoption of its SPC rating.

The good news for sustainability managers is that at least the two programs drive toward the same criteria. So, from an environmental perspective, the two stamps may not be that different.

Finally, sustainability managers should note that the IEEE 1680 standards currently focus on personal computer products. This means that IT managers buying office PCs need to pay attention, but data center managers buying servers don’t need to think about this…yet.

–by John Stanley

Comments Off

Filed under Sustainability

Take the Uptime Institute 2011 Data Center Survey

To take the survey now, use the following link – https://www.surveymonkey.com/s/JLBGF9K

Take fifteen minutes to shape the future of the data center industry! The Uptime Institute’s inaugural data center survey will help data center professionals identify key industry challenges and allow them to benchmark their organization against peers. Datacenter and IT managers who complete the survey will receive a free copy of the results and analysis document ($799 value) and will be entered to win an Apple iPad or free passes to Uptime Institute’s Symposium May 9-12, 2011.

Comments Off

Filed under Symposium 2011

Videos on DCIM, Modular

As a lead in to my presentation on DCIM (datacenter infrastructure management) software at the Uptime Symposium in May, Matt Stansberry interviewed me and produced a short video here.

Matt is very well known in the datacenter world as the former editor of SearchDatacenter, and joined Uptime Institute as director of content earlier this year.

Matt also interviewed Uptime’s Vince Renaud on modular and container datacenters  – Vince will be speaking on this at Symposium.

By Andy Lawrence

Comments Off

Filed under Uncategorized

Japan, nuclear power and datacenters

Following the recent tragedy in Japan, a huge amount of nonsense has appeared in the popular and business press about nuclear power.  As I have no great knowledge of nuclear power generation, I am not going to add to it.

But, as I have been asked about this more than once in recent days, it might be worth a setting down a few notes on the implications of the recent disaster for datacenters.  For some more systematic thinking on dealing with risk, I draw you attention to Ken Brill’s forthcoming Uptime Symposium presentation, “The Characteristics of High-Reliability Organizations”. Ken was partly inspired by errors that were made by BP and others that led to the Gulf oil spill, but undoubtedly the Japanese quake adds to his thesis.

I see five consequences for datacenters, the last two of which are trends that are already underway.

1) Governments will become more cautious about nuclear power, and it will more difficult for power companies to get permits, at least in some countries and states. As a result, electricity supply will be more constrained in some geographies, or will be more dependent on oil, pushing up prices. Bottom line: higher energy bills, greater focus on energy efficiency, greater focus on finding secure locations.

2) Governments and utilities will become more cautious about nuclear power, and as a result, coal and other fossil fuels will take a higher share of power generation, at least for a period. This will put more pressure on governments to introduce and properly manage carbon emissions through trading or tax schemes. Bottom line: More carbon taxes and trading schemes will attempt to drive down power consumption, greater focus on energy efficiency.

3) Datacenter managers will review their resilience, focusing on the traditional but apparently increased risk of losing grid suppliers and/generators. This will cover, obviously, risks from natural events, such as floods, as well as sustained availability of generator fuels. Some datacenters in Japan have found it difficult, although not impossible, to secure sufficient fuel as the grid has come under pressure. Similar problems occurred after Hurricance Katrina in the US.

4) Business and datacenters  may be expected to participate in electricity demand-reduction programs initiated by governments and utilities. This means that IT organizations may have to develop strategies, with their datacenter suppliers or managers, to reduce or scale down application availability or services in times of high demand or constrained supply. Sony has, for example, temporarily shut down some of its online games services.

5) Overall availability planning will increasingly entail incorporating the ability to switch work loads from one datacenter to another, preferably in a different region.  Bottom line: new thinking on IT availability required, investments in advanced software necessary.

By Andy Lawrence

1 Comment

Filed under Uncategorized