On Monday this week, I had the pleasure of presenting at the 2009 Technology Conference sponsored by the American Society of Association Executives (ASAE). I was asked to present at a pre-conference session: “Green Computing – IT’s Social Responsibility.
First a bit on them:
- ASAE (American Society of Association Executives) is the membership organization and voice of the association profession. Founded in 1920, ASAE now has more than 22,000 association CEOs, staff professionals, industry partners, and consultant members.
- The Center for Association Leadership is the premier provider of learning and knowledge for the association community. The Center was founded in 2001.
- ASAE & The Center serve approximately 10,000 associations that represent more than 287 million people and organizations worldwide.
I discovered that their member associations range from a few dozen employees to a few thousand so their IT environments vary. A lot.
Most of the time my interactions are with customers who run large to very large IT operations and so I take my examples from those environments. The efficiency principles are the same for smaller environments but the scale is different enough that the savings may seem hardly worth it. Although, I’d say that every little bit helps.
Why waste any asset?
A few in the audience asked for tips on what smaller shops or even individuals can do to help. Specifically they wanted checklists. So I looked at the fundamentals, and then did some digging for a few good lists and came up with the following to pick and choose.
Three Biggies:
At it’s most fundamental, EMC’s advice is: consolidate, optimize, automate.
These aren’t new principles. We just have more compelling reasons to adhere to them – where it makes sense – and we have the added benefit that new technologies – some in just the last year or so – that generate much higher payback from taking these steps.
Consolidate, means consolidate everything. Data Centers. Storage. Servers. Networks. Applications. For example, we estimate a 2:1 payback just from the consolidation of storage. Consolidation of servers through virtualization is much bigger 10:1, 15:1 or even higher.
Optimization brings much bigger payback once all of that consolidation has taken place. Tiering data can and applications can decrease total cost of ownership by 25%. Robust disaster recovery and business continuity is better enabled – optimized - by virtualization.
Then, software and business process automation can decrease operating costs by 30% or more. Improvements from data deduplication, for example, are off the charts.
And a big benefit in all of this is that you needn’t do it all at once. Probably couldn’t even if you wanted to. It’s more a matter of step through the three - consolidate, optimize and automate – rinse and repeat.
A Call-to-Action Checklist
I remind myself that some in the audience are new to this and that revisiting the basics now and then is worthwhile. So, here’s a checklist that EMC developed as a starter set almost two years ago. Still makes sense though if you haven’t done hit them already.
- Get an executive sponsor
- Determine your data center power capacity
Are you near or at limits?
- Assess your utilization rates and power efficiency
Are you managing your current assets wisely?
- Identify and address low hanging fruit first
Do you have stranded servers, over cooling or inefficient data
center cooling, under-utilized storage?
- Use TCO when planning new projects
- Some technology refresh projects can fund themselves through operational cost savings.
- Create an energy strategy
Employ a cross-functional team that treats the data center as a
core business operation of your organization.
Since this list was developed, I’ve learned that item one is most important. Try convincing the Oracle application team to tier their application. Or convince server and storage “owners” to share through virtualization. Nice to have a bit of executive arm-twisting on call.
Looking back a bit more, have a look at my post from October 2007 that included some other checklists and links to more tips: Branch Circuits
Clean Up
A January 29th article on internetnews.com features an interview with Ken Brill from the Uptime Institute (another source of energy advice) extolling the virtues of data center clean up as a “free” savings on IT energy and space.
Ken’s short list is:
- Turn it off. Make sure that every device in the data center is doing useful work – he estimates that about 10-15% are not.
- Measure it. Calculate your power usage efficiency (PUE). This is a measure of how much datacenter power goes to IT infrastructure and how much to cooling. The first digit is the operating power, and everything above 1.00 is power for cooling. Generally anything 1.5 or lower is considered very good.
- Plug up. That means plug any blank panels and holes where cables penetrate the floor, as that wastes air conditioning. (This one isn’t in the article but I have heard Ken say that just doing this has allowed some data centers to turn off power sucking air conditioning units altogether.)
You can also have a look at the links below that offer some low-cost or no-cost tips.
GreenerComputing.com: “Fifteen Simple Ways to Green Your Data Center on the Cheap”
These are mainly about cooling options – although their idea of “cheap” is strained.
And one more that is a reasonable simple primer:
http://en.wikipedia.org/wiki/Green_computing
Comments