The first in June was an overview of ways in which governments can use IT to save money while reducing environmental impact.  Regardless of where policy makers are on the political spectrum, reducing energy use through IT seems like a no-brainer in this financial and environmental climate.  You can read the first post here.

The second article, which gets me the title of this post, was on about Google's "How Green is the Internet" event that took place in June.  As I wrote in the second article, I think the more pertinent question is not so much "How Green is the Internet?", but "How can the Internet make us Greener?".  As it happens, I wrote about some of the ways in my first post. :-) 

http://corporateecoforum.com/ecoinnovator/?p=7035

As incredible as it sounds, and as I wrote in The IT Energy Efficiency Imperative, a typical data center uses less than 5% of the energy it consumes for actual computing –the rest is lost due to various overheads and inefficiencies.  If that wasn’t bad enough, data centers also generate huge piles of e-waste each year as computer servers are replaced by more powerful and, ironically, more energy efficient models.

While this may sound terrible, first keep in mind that it usually takes less energy and natural resources to communicate and consume goods digitally rather than physically, so more investment in IT is often net positive for the environment.

However, underlying all those boasts about new, “energy efficient” data centers, there is usually a dirty little secret: the servers are often woefully under-utilized.  Most work at less than 10 percent of their capacity, but suck up about half of the energy that they would when running at full capacity. So all over the world, millions of servers operate like mostly empty delivery trucks, consuming lots of resources, but delivering very little.

True, most IT departments have tried to tackle poor server utilization by consolidating applications onto fewer servers through the use of virtualization technology.  Unfortunately, many consolidation projects have stalled due to financial constraints, organizational politics and staffing shortages.   As such, server utilization across the enterprise remains very low – often less than 10 percent.  Even worse, as each generation of server hardware has become more powerful, utilization has tended to decrease even further as applications and IT operational practices have been unable to take advantage of it.

Continued...

In my white paper on IT energy efficiency, I wrote about the need for software developers to design their applications to dynamically scale, so we aren’t simply trading idle servers for idle VMs when applications move to public or private clouds.   Unlike physical server infrastructure which is often statically assigned to specific applications for years, private and public clouds typically provide incentives for developers to scale up and down dynamically because the billing increment can be quite small – often by the hour.  So you can control costs and reduce the environmental impact of your apps at the same time – if you design them appropriately!

Up until now, scaling instances on Windows Azure was a manual process or required you to develop your own functionality to do this automatically.

In response to this opportunity for improved resource efficiency and utilization, the Microsoft Patterns & Practices team has worked with Microsoft Consulting Services and an advisory board of Windows Azure customers to develop the Windows Azure AutoScaling Block, which RTW’ed last week as part of the new Enterprise Library Integration Pack for Windows Azure.

WASABi, as it’s affectionately known, helps developers to automatically scale both web and worker roles in Windows Azure by dynamically provisioning/decommissioning roles or throttling. These scaling actions are based on timetables or on metrics collected from the application and/or Windows Azure Diagnostics.

WASABi addresses the following scenarios:

1. Autoscaling both web and worker roles in Windows Azure by dynamically changing instance counts or performing application throttling.

2. Autoscaling Windows Azure roles based on timetables.

3. Autoscaling Windows Azure roles based on metrics collected from the application and/or Windows Azure but constrained by upper and lower bounds on the instance count per role.

4. Preventing fast oscillations in the number of role instances through the use of a “stabilizer”. The stabilizer can also help to optimize costs by limiting scaling up operations to the beginning of the hour and scaling down operations to the end of the hour.

5. Monitoring and logging autoscaling activity.

6. Sending notifications to preview any scaling operations before they take place.

7. Encrypting the rules and other configuration in Windows Azure blob storage or in local file storage.

8. Managing the autoscaler configuration by using Windows PowerShell.

WASABi scales apps through rules that a developer and/or an IT Pro creates and updates over time.  The excellent blog post by Julian Dominguez notes that there are 2 types of rules in WASABi:

• Constraint Rules- Defines lower and upper boundaries for the number of instances that should be present at all times. You can define boundaries that apply at all times, and you can also define boundaries based on recurrence patterns, to proactively scale your application based on expected load.
• Reactive Rules- Allows you to scale your application based on conditions, such as when a certain performance counter goes beyond a certain level (e.g. average CPU usage for all instances of a role is greater than 80%), when a Windows Azure Queue has too many messages, or if you have your own metric you are interested in, you can extend and support it.

The recommended way to obtain the Enterprise Library Integration Pack for Windows Azure is as NuGet packages. Alternatively, you can download self-extracting zip files with binaries, sources (including tests) and the reference implementation from MSDN.  The reference implementation also includes an Autoscaling management site, which lets you do authoring of the rules, and monitoring of the autoscaled application, so you can see how Tailspin (the reference implementation application) responds to load.  The configuration tool is available as a Visual Studio extension package (VSIX) from the Visual Studio Gallery.

The buzz and uptake on WASABi has been very positive thus far and which leads me to believe that WASABi will be a key ingredient for any Windows Azure app that expects (or hopes) to scale up significantly.  Next up – Pickled Ginger?

P.S.   If you were wondering, the root pictured above is the real Wasabi root, which you can read more about here.

Once again, I've had a long hiatus from blogging.  But I've got an excuse (this time!).  I've been heads down on finalizing a new white paper on IT energy efficiency.   I also just returned from Corporate EcoForum's annual meeting where a number of the attendees confirmed the key assertion in the paper - that in most organizations, server utilization is embarassingly low and falling, and (as the chart above shows) improving it provides the biggest opportunity for increasing IT energy efficiency.  Move over PUE! (we still love you though).

As I posted on Microsoft's sustainability blog, the paper explores the critical importance and substantial benefits of embracing IT energy efficiency in a world where almost every facet of business and society is dependent on IT-based services. The exponential demand for these IT services is beginning to strain the capacity and cost effectiveness of many organizations’ IT infrastructures. It shows why adopting strong IT energy efficiency principles and practices – from the top of the organization down to the IT pro - can significantly mitigate these challenges and improve the cost effectiveness, capacity and flexibility of tomorrow’s IT infrastructures.

In this paper, we:

• Highlight the financial, productivity and environmental benefits of embracing IT energy efficiency.
• Illustrate why increasing IT resource utilization from today’s low levels offers the most significant energy efficiency gains and why current efforts to improve utilization are failing.
• Provide motivation and actionable guidance for IT and business decision makers to improve IT energy efficiency and resource utilization through technologies such as cloud computing.

If you are looking to reduce costs and increase the ROI from IT (particularly within data centers), or simply contribute to reducing the environmental impact of IT, please download this onscreen-readable whitepaper and let me know what you think.

The IE team recently published the results from a set of tests comparing the energy efficiency of the latest competing browsers.  Spoiler-alert - IE9 was the most efficient. :-)

http://blogs.msdn.com/b/ie/archive/2011/03/28/browser-power-consumption-leading-the-industry-with-internet-explorer-9.aspx

The System Center  team blog featured a guest post by one of their MVPs yesterday on his experiences with the centralized power management features in SCCM 07 R3.

While not the only feature in the release, the centralized power management functionality certainly seems to be getting the most amount of airtime.  Which is not really surprising given that the savings gained from implementing power management would seem to easily pay for the product many times over.  As Panu Saukko reports, his customer was spending about $2500/weekday for electricity without power management.  In early September they enforced power management on 5000 computers using SCCM 2007 R3, and decreased the power bill by to about $300 per day!  And this was with fairly lenient settings - 3 hours of idle before automatic sleep during the day, which means that most PCs are on all day during the work day.   The opportunity for more power savings is still significant as they crank down the idle-to-sleep timeouts.

Read the post on the System Center blog here.

I analyzed some of the raw data from the tests and discovered that in addition to saving over 1/2 the power on idle (54%) as mentioned in the study, the newer server  provides almost 300% more transactions at any given CPU load than the 2006 model year server, resulting in about a 300-400% improvement across the board on performance-to-power ratios.

This is really great news.  Of course, it also implies that in order to really take advantage of these new servers, you must make sure you are really driving up utilization or that extra performance is going to be wasted.  Simply consolidating even more idle VMs to these new servers really isn’t going to cut it.  Your ops management infrastructure needs to be aware of what is going on inside those VMs so it can shut some of them down if appropriate.  This is going to require applications to take a more active role in resource utilization in the future. 

Download the study here  (2.67MB)

As I laid out in my presentation, IT efficiency has (or should) become the next critical focus for IT.  Efficiency shares characteristics similar to security, in that everyone involved, including Business Decision Makers, developers, IT Pros, and even users, together with hardware and software vendors, need to do their part to improve IT efficiency. 

Readers of my blog know that applications have been the missing link that prevents PCs and servers from being highly efficient.  Now is the time to fill that void.  With Windows 7 and Windows Server 2008 R2, IT Pros, developers and enthusiasts are able to use in-box diagnostic tools to determine why power management is not working as expected.  Microsoft uses these tools every day to identify problems with its own and even third party applications, and we are asking the rest of the Windows ecosystem to do the same to ensure that their applications are “energy smart”. 

So where should a developer start?  During my talk, I unveiled the “Energy Smart Software Top 10”, which provides recommendations for how developers should prioritize their energy smart software efforts:

For developers of client software, steps 1 through 8 are relevant.  For server developers 3-10 are relevant.  Additionally, I outlined some of the key investments that will help keep systems idle.

We have published a collection of guidance on www.microsoft.com/energysmart that covers the majority of these recommendations to help developers build energy smart software.  More guidance will be added over the coming weeks and months.

There are also links on the site to tools, including an update of the Joulemeter research project by Microsoft Research that we also announced at IDF.  The new version of Joulemeter helps democratize computer power measurement by allowing laptop users to use their laptop battery to calibrate the power model to their specific machine.

Developers who have questions on developing energy smart applications can post questions in our MSDN forum.

While there are many factors that are forcing the need to focus on IT efficiency (e.g. increased demand for IT, power constraints, regulations and stakeholder pressure), there are a lot of benefits as well.  Users can be more productive due to better power management experiences and TCO can be reduced through smaller energy bills and less idle equipment.  And it won’t hurt that our planet’s life support systems will be just a little less stressed as we slow down the rate of extraction and waste disposal (including carbon into the atmosphere) as a result of more efficient IT utilization.

As I closed my presentation, I noted that Intel calculated that the world’s 1 billion PCs collectively consumed more than 320 terawatt hours of electricity in 2007.  They estimate that by 2014, advances in computing will allow 2 billion PCs to consume less than half of that amount (151 twH), but provide more than 17 times the computing power.

This is certainly would be an impressive feat for energy-efficient performance, and one that I hope the industry can meet.  But, as I pointed out, hardware and operating system advances alone will not get us there.  Developers will need to take an active role and ensure their software is energy smart if we are to get close to achieving this goal.

Again, if you write software for a living, you owe it to yourself, your customers and the environment to become educated on energy smart software development.  Please visit http://microsoft.com/energysmart today!

For those of you who couldn’t attend my talk, a recorded interview of me talking about the topic is available on MSDN’s channel 9.   If and when we get the video from IDF, I’ll post a link to that too.

Possibly one of the most interesting tests they did was the effect of processor chipset drivers on the power consumption of Windows XP and Windows 7. (table here is copy and pasted from the appendix of the doc, and painfully reformatted)

What is notable is that while W7 beats XP on every test, the “out of box” differences are really significant. If you look closely, Windows 7 out of the box numbers are the same as with the updated drivers – this means that Windows 7 is taking care of the chipset drivers, even on older hardware. The same cannot be said for XP. And even with updated drivers (obtained manually), it’s still worse than Windows 7. How many folks have reinstalled Windows XP and not installed the chipset drivers. Quite a few I would guess.

The Mindteck green computing white paper can be obtained at http://www.mindteck.com/whitepapers/Enabling-Green-Computing.pdf

If you’re just getting started on making your software “energy smart”, I strongly encourage you to take a look. Enjoy!