Dynamic Data Center, Dynamic Metrics

I'm happy to see that the Green Data Center Blog points out that data center efficiency metrics like PUE (Power Usage Effectiveness) need to be necessarily dynamic if/when the data center is dynamic:

"I've been meaning to write about PUE, and have been stumped in that It is defined as a metric, and in the Green Grid document referenced it makes no reference that is dynamic. In reality PUE will be a dynamic # that changes as the load changes in a room. How ironic would it be that your best PUE # is when all the servers are running at near capacity, and shutting down servers to save power will [actually] increase your PUE?

A broader explanation of Assessing the New Data Center Metrics is on an earlier entry.

If you look at Gartner's description of "Real-Time Infrastructure" (RTI), IDC's "Dynamic IT" or Forrester's "Organic IT" concept, they all point to evolutionary steps where data centers will adapt/respond to demand by re-purposing equipment in real-time. As this happens, the metrics that will define efficiency will need to track this as well.

EDS Outlines Key Data Center Energy Saving Measures for 2008

I Just came across this very validating article in the datacentres.com newsletter, Data Centre News from the BroadGroup in the U.K.

The EDS Fellows – a team of recognized IT leaders in EDS, the US outsourcing and IT company - have predicted 8 measures that need to be taken to reduce the impact of the Data Center on the environment and also prolong the use of IT components. They include:

1. Users should move away from deploying one application per server. By using server virtualization server utilization can be boosted from 15 to 20 per cent to up to 90 per cent.
2. Users should turn off their IT hardware – servers - when they are not being used. Automated operating processes should be deployed to bring on servers when they are required. [Personal note: I'm happy to see that Server Power Management made it to #2 on this list!]
   
3. Users should deploy power saving techniques which reduce server operating speeds and increase power supply efficiencies.
4. Users should reduce the number of software applications that are deployed within the Data Center and standardize as much as possible, and cut out the under-used applications.
5. Users should deploy higher density multi-core CPU’s which will as a result lead to fewer blades required per rack.
6. Users should implement a rigorous maintenance and replacement policy to improve the Data Center layout and IT configuration within the Data Center
7. Users should focus on the entire IT infrastructure operation rather than reducing the cost of computing, and should seek to measure the overall power impact of their facility. [Personal note: this is consistent with the metrics & recommendations I've found from the Uptime Institute]
   
8. Finally users should consider the implications of networked and integrated IT services, in particular to control heat, cooling and lighting in line with the use of the facility.

Data Center Energy Efficiency Survey: Redux

I received many cards and letters regarding my previous blog about the survey Cassatt ran last month - and yet I forgot to reference the raw results, which are here.

Meanwhile, I've also come across a few more sources of data and industry surveys on the state of Data Centers and IT energy use and efficiency:

• SearchDataCenters' study on Data Center construction: In 2007, 22% of survey respondents said they were involved in a data center construction project. More than 60% of respondents will be involved with construction in 2008. And only 18% of respondents will not undertake a data center construction project in 2007 or 2008...
• This very interesting study by AMD on power consumption and cooling in the data center; what makes it so intriguing are the trends it began to forecast back in 2005. Plus a more recent survey they conducted on Power and Heat Management.
• A more current study sponsored by AMD (but performed by a guy I hold in high esteem, Jonathan Koomey of Lawrence Berkeley National Laboratory) tracks shifting patterns in data center energy use across five world regions. It's a tour-de-force of estimating total power consumption around the world based on quantities of servers in use.

Survey Results in on Data Center Energy Efficiency

Here at Cassatt, we recently completed a survey of IT and Facilities professionals regarding data center energy efficiency; we were trying to get a handle on how many are pursuing what types of efficiency initiatives, why they did so, and what data informed them in making decisions.

Some results were surprising, some not. But one thing is for sure: This is a rapidly-developing area with few hard facts available as enterprises develop their own strategies. Maybe this explains why there were so many related surveys in 2007: These included surveys from Emerson Network Power and the Data Center User's Group in Sept '07, Digital Realty Trust in Sept '07, and Symantec in Nov '07. I'm sure there are many others.

Follows are some interesting factoids from our own study:

• Massive waste in Dev/Test: Looking at Development & Test labs, 30% of companies had 25% or more of their servers in Dev/Test. But consider this: Nearly 30% of those companies had over 60% of their Dev/Test servers idle during "off-peak" hours. And, fully 50% of companies had 20% or more of their serves idle in Dev/Test during off-peak. There's a lot of room for savings through Active Power Management here :)
• Hitting the ceiling: Nearly 43% of companies have a data center within 25% of their maximum power capacity, and 25% of companies are within 25% of their maximum cooling capacity.
• Motivation is still, ultimately, $: Even though we're talking about the benevolence of being "green," the strongest motivators for initiatives were still "Clear economic benefit" (68%) and "Constrained by power or cooling" (57%). But it's consoling to know that "Environmental responsibility" ranked reasonably well (39%)
• Self-serving advice?: We also asked where users look for guidance. Who came out on top? Systems vendors (49%) and Power & cooling vendors (39%). Now maybe that makes sense, but maybe it indicates the wolf guarding the sheep. Why was it that "neutral" sources like EPA and EnergyStar (29% and 33% respectively) narrowly scored below Peers/colleagues (34%)? (where do those peers get their advice, anyhow?)
  
• A little confused?: Also, an observation made by our friends at the Green Data Center Blog was that 61% of respondents were aware that the EPA recommends turning off idle servers to save energy -- but over 30% of respondents also said they didn't know how many servers are idle at any given time!
• Poor self-image: And finally, only about 19% of respondents felt that their data centers were "better than average" -- either we had a lot of pessimistic respondents, or there's a lot of room in the market for improvement :)

2007 Darwin Awards are final

I've got to say that each year, this stuff is pathetic.... it's all about folks who come from "the shallow end of the gene pool". The tally of the 2007 Darwin Awards is out, and my personal favorites are:

• (28 July 2007, Czech Republic) A pack of thieves attempted to steal scrap metal from an abandoned factory in Kladno. Unfortunately for them, they selected the steel girders that supported the factory roof. When the roof supports were dismantled, the roof fell, fatally crushing two thieves and injuring three others.

My other favorite (a vintage one, from a few years ago):

• (3 February 1990, Washington) The following mind-boggling attempt at a crime spree appeared to be the robber's first, due to his lack of a previous record of violence, and his terminally stupid choices:
  1. His target was H&J Leather & Firearms. A gun shop.
  2. The shop was full of customers - firearms customers.
  3. To enter the shop, the robber had to step around a marked police patrol car parked at the front door.
  4. A uniformed officer was standing at the counter, having coffee before work.
  Upon seeing the officer, the would-be robber announced a holdup, and fired a few wild shots. The officer and a clerk promptly returned fire, covered by several customers who also drew their guns, thereby removing the confused criminal from the gene pool. No one else was hurt. (BTW, this true story fits perfectly with one of the Darwin Awards Mottos: "Never bring a knife to a gunfight")
  

As you would expect, there are formal rules for nominations to the competition, the summary of which are:
- Reproduction - Out of the gene pool: dead or sterile.

- Excellence - Astounding misapplication of judgment.
- Self-Selection - Cause one's own demise.
- Maturity - Capable of sound judgment.
- Veracity - The event must be true.

A day at DOE's Save Energy Now workshop

Earlier this week I had the privilege of attending a workshop sponsored by the Department of Energy's Save Energy Now program , in collaboration with the EPA/EnergyStar group. The purpose was to help DOE define and engineer tools to help assess data center efficiency performance. They're reaching-out to industry to ensure they're using the right metrics, asking the right questions, and producing a product that will be useful in real-life. The tool will (presumably) help firms "score" the efficiency of their data center from 1-100, much the same way there are already tools to score the energy-efficiency of buildings, etc. BTW, see my earlier blog on some proposed metrics from The Green Grid and Uptime Institute.

In the room were mostly vendors such as AMD, Emerson Electric, HP, Intel, NetApp and Sun. Plus, folks from DOE, EPA, Lawrence Berkeley National Labs, the Green Grid, and the Uptime Institute attended as supporting resources to DOE. And first-off, it was great to see the overlap/interaction between these groups -- especially that DOE, Green Grid and Uptime Institute were cooperating.

The tool itself is expected to have 4 components of assessment: IT (servers, storage, networking, softwar), Cooling (chillers, AC units, fans), Power Systems (UPS, distribution) and Energy Sources (generation, etc.).

But the real core of the day's conversation is around what's meant by "Efficiency" - all agreed that at the highest level it was the ratio of "useful computing output" to the "energy supplied to the overall data center". That second number is sort of the easy part: it includes all of the power used for lighting, cooling, power distribution, etc. etc.; sometimes it's hard to measure empirically, but it all comes down to kWhs. The real issue, it turns out, is what's meant by the first number, "useful computing output".

In our afternoon IT breakout group, about 10 of us debated for a good hour or so about just that: how do we define "output"? Is it MIPS? Is it measured in SPEC units? Is it CPU utilization? And what about storage and network "output" as well? In the end, we agreed that we should define it the way any IT professional would: As an application service with an associated SLA. So for example, the useful output would be a service called "Microsoft Exchange" with SLA traits of X users, Y uptime, and Z latency. And most important, this approach makes the output independent of underlying technology and implementation. Thus, two data centers could implement this service/SLA combination in vastly different ways, with vastly different power requirements and therefore energy efficiencies.

When the DOE tool (or set of tools) is complete in mid-2008, it will represent a seminal event: Data Center operators will have a standard benchmark against which to self-rate -- and to (confidentially) compare their efficiencies against their peers. It will also begin to put pressure on IT vendors (!) to focus on the "gestalt" efficiency of their systems, not just of their components. (IMHO, this will make me *very* happy)

And, I hope, this benchmark will begin to accelerate (a) the move toward dynamically-managed shared compute resources, and (b) the technical & organizational bridge between IT system management and facilities management.

End-of-the-year holiday shopping for IT

The end of the fiscal year us upon most of us - and for some that means spending any remaining dollars on some quick-hit (read: fast ROI) initiatives.

So in the spirit of the holidays, here are some interesting links:

1. Bridget Botelho at SearchDataCenter writes "Servers get no rest during the holidays" -- that most enterprises (and especially small/medium businesses) leave all servers on during the holidays, even if the company is closed. It's a waste of power/money, even though many solutions abound. (And did you even think about the security risk of keeping your IT on but essentially unsupervised?)

2. And, for those of you with money left in your budget, Rick Vanover from TechRepublic chimed-in to suggest 10 good ways to use your remaining IT budget before the end of the year - I particularly like the following:

• #3: Purchase power management: Many new power management devices are available now that can be a good replacement for your limited power distribution units (PDUs). These PDUs can add management layers to individual power sockets for power consumption, naming, grouping, and power control. The new devices can also add more ports should you need to power more computer systems in your racks.
  

I would only add that in addition to switched PDUs, you should consider purchasing policy-based software to control them.

3. And, Rick went on to also cite an earlier blog of his, 10 things you should know about advance power management - another topic near-and-dear to my heart, especially:

• #7: Turn off retired or unused devices: This will reduce your power consumption — and possibly accelerate your removal of the device so as not to overprovision power unnecessarily...
  

The net-net is the following: If you need a quick-hit, easy-to-implement (and money-saving) and pragmatic solution, pursue intelligent operation of your IT - no need for a technology refresh, no need to re-architect anything.