Sentilla Blog

This week, I'm super excited to give a talk at PICNIC. This is the TED of Europe and everyone from Richard Branson to Genevieve Bell (ala Intel) to Werner Vogels to a million other amazing individuals will be speaking. They all have a huge interest in changing the world, and they are doing it today. They're focusing on energy, environment, IT, media, and really everything else. I'm honored to give a talk in the presence of such luminary individuals.

Data centers and the equipment they house have come a long way since the days of the mainframe. Despite how different the computing systems may look today, they actually share a lot of similarities. Cloud computing and virtualization look a lot like distributed mainframes, and systems are moving back to old benchmarks that incorporate power and workload.

Energy and Green IT have become the hot topic, and a recent article illustrated to me how much confusion is out there in the market. Vendors (like Sentilla) are all competing for the same resources and budget, whereas the media simply doesn't have enough time to investigate the industry as fully as analysts have done (and are still doing). I've been encouraging analysts from Gartner, IDC, Forrester, and others to put out a "data center energy management landscape" document, to set the record straight on how each of the vendors interact and compete. Unfortunately no such document exists yet.

On a recent flight, I had the pleasure of watching a number of episodes of E2, a PBS show narrated by Morgan Freeman.  The shows cover a number of environmental topics and features Stephen Chu, former energy expert at LBL and now Secretary of Energy.  If you haven't seen them, I highly recommend finding a few episodes. Unfortunately, only the "design" episodes are on hulu, the "energy" and "transport" episodes have not been posted online.

A couple of things stood out at me, and I came away with a feeling that we could do better at achieving energy efficiency and energy independence.  Not just because we should as a planet-saving activity, but because we've actually gone backwards in efficiency in the last 100 years.

First, consider this statistic:  The first Ford Model T accepted both Ethanol and Gasoline, and got 25 miles to the gallon.  Today's cars average 21 miles to the gallon, and while it only costs $50 to outfit a gasoline car at production with Ethanol capabilities, few cars are actually "FlexFuel".

Figuring out how to implement the right efficiency plan in the data center is a daunting task.  I often start discussions with customers by asking "How much electricity does your data center use?" and "How much are you paying for electricity?"  While these may seem like such amazingly fundamental questions, you'd be surprised by how many people don't know the answers or respond with "Let me go and look that up."

Further confusing the issue is vendors claims about what they can achieve.  Have you heard the marketing campaign about virtualization reducing your energy bill by 80%?  Sounds compelling right?  The trouble is, you're not going to virtualize every server in your facility and shut off every chiller.  Let's say 20% of your servers can be virtualized (such as staging, dev, and test), and your servers consume 70% of your IT energy consumption, which is 45% of your facility's energy consumption.  So if we add that up, .8*.2*.7*.45 = 5%.  That means that, for your data center, 80% savings just became 5%.

This means that data centers really need to figure out what their strategy is.  And, of course, you want the biggest bang for your buck.  By measuring your baseline, applying the various strategies to it, and calculating ROI, a plan can be formulated that makes sense of the efficiency measures for your specific facility.  I've written an eWeek HowTo titled How To Achieve 40% Energy Savings In Your Data Center.

Sentilla is excited to win the American Business Award for Best Software Product of the Year for Sentilla Energy Manager for Data Centers. There are essentially two ways to solve the global warming problem: find alternative sources of energy or reduce the energy you're currently using. Wind, solar, and tide power are great.

They've come a long way in a short amount of time. But it's going to take a lot more than that to solve this problem because the world's appetite for energy is so immense. In addition to using alternative energy sources, we need to take a closer look at how we use today's energy and find ways to cut back. Improving energy efficiency can be tremendously valuable because the cheapest power plant ever built is the one you don't have to build.

This has never been more profound than in the case of data centers, whose energy consumption, globally, is greater than the country of Israel. This is clearly an area where much can be done and Sentilla Energy Manager for Data Centers is directly addressing this issue.

We appreciate the award and consider it validation that this market segment is not only in need of a solution, but that we are well positioned to provide an answer to this growing problem.

As you develop your IT strategy for the next five to seven years, you, an IT executive, will be asked to support a growing number of new business initiatives within the constraints of a shrinking to flat IT development budget. You are already spending a disproportionate amount of your budget on maintaining the existing data centers. You have insufficient capital expenditure to develop your environment necessary for meeting these new and growing business services. You need to think outside the box. You need to think “energy.”

Here’s one perspective: To efficiently manage your data center, you need to know the actual energy usage of all your assets to prevent over-provisioning while meeting peak demands. By using energy consumption and cost as the base indicator for your data center performance, you can quickly and dynamically balance your compute load with your IT infrastructure -- giving you the elasticity you need to match demand cycles.

This asset-centric approach to energy management simultaneously balances workloads with IT footprint, frees up operational cash flow, and gives you an elastic data center infrastructure. This approach helps you to meet your data center performance objectives, by dynamically balancing your workloads with IT capacity allocation.

In so many ways, the Cisco announcement was a throwback to the last century. Yes, it was BIG. Satellite uplinks, worldwide audience, a parade of Cisco executives and a roomful of attendees in a world-class venue. John Chambers is extremely compelling and I always enjoy his presentations. I have tremendous respect for him, and for what Cisco has done and is doing for the industry.

In the process of releasing application tracking in Sentilla Energy Manager, we had some great discussions with Dave Cappuccio and Rakesh Kumar.  Our conversations turned to "How do you measure efficiency in the data center?"  Of course, we all argue that PUE is not the right metric for determining the efficiency of a data center.  Instead, there needs to be more metrics around the applications, their workload, and the overhead they incur.  These metrics are at the core of the new application tracking release of Sentilla Energy Manager.

"The new efficiency metric for data centers is compute per kilowatt," said David Cappuccio, managing vice president and chief of research for the infrastructure teams, Gartner. "Application energy and cost information is becoming critical for enterprises concerned with optimizing their infrastructures. IT organizations armed with this information will increase their adoption of virtualization and realize energy savings and consolidation benefits more quickly."

The theme at this year's SAP TechEd conference is "Innovation without Disruption".  What a fitting theme for Sentilla as well, as we share SAP's desire to bring innovation to the data center without causing any disruption. The idea is simple: if you build a platform that leverages the existing architecture but bridges it to new technologies, you can migrate a customer from a legacy approach to a modern approach without disrupting their applications. As you can imagine, this is important with big ERP, CRM, PI, etc systems that must be available.

With a lot of the data center energy efficiency focus on facility improvements and virtualization, I've decided we need to take a step back and look instead at the applications.  As Moore's Law has increased the capabilities of servers, and disk density has doubled (roughly) every year, we're no longer as constrained as we used to be by physical resources.

I am, at the heart of it all, a computer scientist.  In grad school, I spent a lot of time working on optimizing software that runs on embedded systems -- little microcontrollers with limited resources.  At the core of these systems was power.  If you wrote your code inefficiently, not only would it fail to fit on the device, it would also burn through batteries.  And as data center operators know, a server without power is a VERY BAD thing.

So why is it then, that we don't look at how efficiently our applications are written?  There's no apples-to-apples energy comparison, and few people take energy into account when buying a software package.  My prediction is that will change.  In a few years, when you evaluate whether to buy (or renew) SAP or Oracle, you'll ask about the energy operating cost of those software packages over the life of the contract.

On the same day, two opposing articles have been published with completely opposite points of view.  On one side, Brian Fry argues that location no longer matters and we should build data centers in the most efficient locations and supply fiber connectivity to them.  On the other side, Paris Burstyn argues that location and latency are business critical for companies resulting in Equinix's acquisition of Switch and Data.

At the center of Brian's argument is that low cost, low carbon power is good for data centers.  If you can live with your data centers a few milliseconds away from everyone else's, then your IT operations can be greener and cheaper, a win-win scenario.  But in the world where every millisecond counts, that's not the right solution for all industries.

Recently, there's been a slew of articles about how IT managers have identified that standardized access to power information over SNMP is one of the top ten problems that they face when managing energy.

We were listening. And today we announced the release of Sentilla Energy Manager for Data Centers version 2.1 which integrates all the great things in the original release like the ability to analyze your energy profile at each piece of equipment and we extended that to include third party equipment. That means that, within an hour of installation, you can get a high level overview of what's going on across the entire data center.

If you are attending the Gartner Symposium/ITXpo next week, you are in for a treat!

Renowned Expert and CTO, Dr. Joe Polastre, will present on the Modern Data Factory.  A winner of Silicon Valley/San Jose Business Journal 40 Under 40 award and named one of BusinessWeek’s Best Young Tech Entrepreneurs, Dr. Polastre will bring to center stage his innovative concept of managing data centers like modern factories: industrializing them to optimal output with minimized capital investment and operating costs. He will share best practices for data center managers faced with immense challenges in aligning activities with business objectives while keeping costs low.

Bring your questions and ideas to Dr. Polastre’s presentation. Also, come by the Sentilla booths to see a demonstration of Sentilla data center analytics!

BNP Media put on the Green Intelligent Buildings conference in Arlington, VA this spring. I was there, and gave a talk on what sustainability means to green buildings. Utlimately the message was this: There are 3 steps to maintaining a sustainable building -- measure, analyze, and act. While most think they can measure once and expect the result months later to be the same, the collective view was that people are the most disruptive force to a green building. Continually measuring, analyzing, and acting is what differentiates an audit from sustainability. There were a number of sophisticated companies there, and it was great to learn the different ways that they keep buildings running efficiently. Tridium was quite impressive -- they are wrestling the large Honeywell, Johnson Controls, and Schneider Electric of the world and attacking the SMB market (and, it appears, winning).

As you and your team are grappling with the ever increasing complexity of delivering more services with less resources, increased governance and shrinking budget’s, you are now also being asked to deliver these services and applications within a sustainable environment and have a Green IT initiative….its enough to make you see Red.

There are a number of socio-political considerations driving sustainability and Green IT for CIO’s and it is shaping their approach to the challenge.

European IT teams have conversations that revolve around the current and pending legislation by country. You are focused on compliance and meeting the requirements of the law – that will usually mean meeting the minimum of the requirements. Your focus will be on cost of compliance, planning for delivering your Green IT response, based on meeting the law of the land, not on what the ROI will be or how you and your team will benefit.

Randy Katz, professor at the University of California, Berkeley, has embarked on a new project called LoCal. Short for any number of things, including "Lo(w) Cal(orie)", "Local Information", and of course "Cal", Katz has thrown Berkeley's hat in the ring to make energy smart in the 21st century. I had the privilege of bouncing ideas off Katz while a graduate student at Cal, and he gave a talk about LoCal at the 2009 EECS BEARS Symposium.

When it comes to "efficiency", there's a lot of different definitions for what makes a data center efficient.  Fundamentally, efficiency is how much work is done per unit of energy.  And when it comes to Cyber Monday, today is the most efficient day for many data centers.

Just think about it -- normally servers are sitting around idle, waiting for a job a to do.  They have low utilization, but are needed for availability to ensure that all customers can access your websites and make purchases without any hassles when the demand picks up.  Most servers in data centers run at a utilization rate of around 8 to 12%, but what's even more frightening is 83% of data center managers don't even know what the utilization of their systems is.  When it comes to ensuring availability of compute resources on heavy demand days like Cyber Monday, it is scary to think that most don't even know how much headroom they have.

I had the pleasure of attending and speaking at this year's Data Centres Europe (DCE) conference, held in June in Sophia Antipolis, France by BroadGroup.  It was delayed due to the ash cloud mess back in April, when I had planned to originally attend but ended up stuck in London instead.  The event was well attended despite being rescheduled, although the weather in Côte d'Azur was a bit warmer than the organizers had anticipated.

At DCE, I was honored in two ways that I want to share with you.  The first is with an award!  The annual Data Centres Europe awards are given out in a number of categories to recognize innovative data centre design.  With a panel of judges consisting of end users and analysts, I was surprised and excited to hear them announce that I am the recipient of the European CTO Award for Innovation in Data Centres!  In the introduction of the award, the judges introduced me and Sentilla with these nice words:

Sentilla is honored to win another product award.  We’re very excited that energy management has gained so much traction this year.  But it’s no surprise.  As the UK embraces the Carbon Reduction Commitment and with the publication of the European Code of Conduct for Data Centres, it has become clear that a solution is needed for managing and tracking data center energy use. The award validates the need for energy management in data centers and Sentilla is humbled to be part of this movement.

Recent studies indicate that, as IT professionals look to trim power expenditures and move toward ever greener IT solutions, energy efficiency metrics will become an ever more critical data point in their decision-making process.  According to a recent Gartner
study, many IT and data center managers consider green IT a top priority, but have yet to embrace measurement and monitoring infrastructure critical to determine energy-saving measures and satisfy government regulations.

Efficiency metrics span the gamut from the GreenGrid's PUE and DCIE to the E.U.'s Code of Conduct for Data Centres Energy Efficiency.  In all cases, energy utilization measurement and analysis is a necessary step to understanding how IT architecture and provisioning can be improved.  Data indicating a server's idle power usage, work done per watt, comparison of energy usage for comparable equipment, and sustained efficiency over time merely scratch the surface.

Data Center Infrastructure Management, or DCIM, has become top-of-mind these days, and for good reason. There are a number of important trends out there fueling DCIM  market growth, including data center consolidation, the move toward virtualization and cloud computing, the need to accelerate service delivery growth, and the desire to reduce power consumption costs.

I had the privilege of being one of the only -- and possibly the only -- startup person attending Google's Green Data Center Summit held at their Mountain View campus. While there were a wide variety of talks (which I'll cover with some analysis in a subsequent post), the gadget hacker in me loved how Google unveiled and described their servers.

Completely custom designed and built in Google-style, here's what they did:

• Instead of having a centralized UPS, Google integrated a battery into every server with a charge controller and test circuit. The battery is sealed lead-acid. Basically, it is a car battery. The goal is to keep the server running for "about a minute" until the generator turns on or the A/C power source is switched.
• The power supply ONLY provides 12VDC (notice the yellow 12V wires coming out of the power supply), and the power only goes to the motherboard. The motherboard then directly supplies power to the disks (at 5VDC) and no other voltage conversions occur (with the exception of 1.2-1.8V for the processor). As such, the AC feed of either 208VAC or 230VAC flow directly to the server with no UPS in the middle. The battery backup directly supplies 12VDC during power outages, so no more inverters.
• The power supply is about 92% efficient according to Google.
• The "distributed UPS" solution is estimated by Google to be "99.9%" efficient since there's no power distribution losses by operating directly at the DC voltage of the server.

Sentilla was at the Uptime Institute Symposium a few weeks back in Santa Clara, and we sponsored the event with our partner SAP.  We had a lot of great discussions, from government policy makers to data center managers of varying size enterprises to cloud providers and infrastructure vendors.

But the most impactful takeaway was from Christian Belady and Mike Manos.  They put it different ways, but ultimately there message was this: Instead of coming to the conference and saying "there's nothing new", come back next year and actually implement the improvements we've been talking about for the last 10 years!

You can see a snippet in Uptime's video from Mike Manos' presentation.

ECN asked technology leaders to brainstorm solutions that would make the most impact, specifically which green technology has the most commercial promise. I responded, along with 7 others, and our thoughts were published in the March issue. I've included my response below, and you can view the ECN March issue for the full set of responses.

New energy is really expensive. The most commercially viable option is to reduce your own energy use, which, as stated by the Obama Administration, is the cleanest fastest easiest, cheapest way to generate renewable energy. In order to reduce your energy consumption, you need to know precisely where and when you're using it.

SAP and Sentilla have written a white paper that ties corporate sustainability to the data center and IT operations.   A fully integrated energy management approach in your data center can deliver the energy visibility you need. this approach can help you achieve improvements in system utilization, capacity planning, workload management, and smart capital investment.

As we previously mentioned, Sentilla Energy Manager 3.1 has built-in integration to SAP Carbon Impact and Business Objects Explorer.  Sentilla Energy Manager collects information about everything in the data center, from storage to servers to CRAC units to application utilization, analyzes and summarizes this massive amount of data, and then provides carbon and energy consumption snapshots to SAP's toolset.  This integration brings the data center into enterprise-wide carbon, energy, and resource tracking.  Since data centers are the third or fourth worst carbon and energy offender in enterprises, tools like Sentilla's are needed to bring visibility into IT operations and align IT with corporate goals.

This white paper was written as part of SAP's Green IT community, where Sentilla chairs one of the working groups.  SAP's Green IT initiatives can be found on their website.