January 05, 2009
The Battle for Data Center Efficiency - Who Owns it?
It is clear that most every company now realizes the importance of controlling the efficiency of data center environments, regardless of company size. We are finally starting to see all IT vendors realize that they need to be part of the solution. Server vendors were the first to recognize this, since the increase in chip power consumption and heat dissipation was the first obvious problem. However, all elements in the data center need to contribute to improved power and cooling metrics.
We are finally starting to see network equipment manufacturers own up to this, with Cisco Systems Inc. starting to increase its focus and marketing on energy awareness. I also expect to see a great deal of hype and marketing in the storage community. We started to hear a lot last year about solid state disks (SSDs), which dramatically reduce power while increasing performance of storage arrays. The costs today are prohibitive to consider a complete replacement of fixed disks for most business, but I expect to see them used in niche high-performance environments and also bundled in tiered storage solutions by companies such as EMC Corp. and Hitachi Data Systems (HDS).
There is still some confusion about whether the data center environmentals are the responsibility of data centers or facilities, but there is no disagreement on the fact that both of these groups need to work very closely together to make sure current and future data center requirements are feasible from all aspects of power, cooling, IT systems performance and costs.
I have been working on several committees over the last couple years to work through the trends and best practices on these issues, including the Next Generation Data Center and the Blade Systems Insight. While the focus is shifting from being green for green's sake to being green to save money, the good news is most of my research has shown that increased attention to environmental considerations usually saves companies a tremendous amount of money, so it makes good business sense to be environmentally aware in the data center. Some of the new topics I expect to come up in future conferences are how to analyze the return on investment (ROI) for IT equipment replacement decisions, as they relate to power and cooling, and whether the dynamics for managed services change are a way for companies to look at outsourcing strategies for saving money.
The bottom line, as I see it, is that EVERYONE owns responsibility for data center efficiency, especially the responsibility for coordination across all IT disciplines and facilities management to make sure any initiatives have taken all aspects of efficiency into consideration for future data center growth.
|by Jerald Murphy||January 5, 2009 in Energy Efficiency |
Permalink | Comments (16) | TrackBack (0)
December 20, 2008
Facebook: 600K new users a day - $1 million monthly electricity bill
datacenterknowledge.com reports that Facebook is adding 600,000 new users a day, with more than 140 million users worldwide, and monthly electrity costs of more than $1 million. The need for more data center infrastructure led Facebook to supplement its Santa Clara data center with an additional facility in Ashburn, Virginia to address latency issues for European users. More than 70 percent of Facebook users are now outside the United States.
With more than 10 billion user photos now hosted (see overall Facebook statistics) and 200 million users expected by early next year, the company's data center capacity requirements will only continue to grow.
|by Will Runyon||December 20, 2008 |
Permalink | Comments (19) | TrackBack (0)
Internet Downtime - 2008's Major Incidents
pingdom.com has a interesting review of the "top 10" Internet downtime incidents in 2008. Everything from severed underwater cables, fires, denial of service attacks, to data center power outages, etc. Good lessons all. Wishing you only "uptime" in 2009.
|by Will Runyon||December 20, 2008 in SysAdmins |
Permalink | Comments (0) | TrackBack (0)
October 21, 2008
The mainframe for mid-sized businesses
An IBM mainframe priced at $100,000? Yes, you read that right. The newest System z mainframe, the z10 Business Class, is a follow up to last February's launch of the z10 Enterprise Class, and is priced and featured for mid-sized companies.
According to InformationWeek, "The system is priced at less than $100,000, making it affordable for companies in developing nations. IBM is offering zero-interest, zero-payment financing on the system for the first 90 days. The z10 offers users big opportunities for server consolidation. It holds the capacity of up to 232 x86 servers within a footprint that's 83% smaller. One company that plans to use the system is Transzap, a provider of electronic payment services for the oil industry. 'We're a small company but our transaction data volumes are growing upwards of 100 percent, annually,' said Transzap CEO Peter Flanagan."
Caption: Created for mid-sized businesses, the IBM z10 BC simplifies commercial computer operations with "specialty engines" to run popular business and consumer applications (email, website hosting, transaction processing, etc) on one of the world's most trusted and secure computer platforms. IBM co-op student Sean Goldsmith surveys the new z10 BC mainframe in IBM's Poughkeepsie, NY, plant to add an extra 1,000 email users with the energy of a 100 watt light bulb. Goldsmith, a senior at Marist College, anticipates a bright future with the mainframe.
CRN also reported that "IBM is working with more than 130 solution providers and systems integrators worldwide who are certified to sell IBM System z mainframes. The certification of IBM System z sales and technician skills has increased 300 percent during the first half of 2008 compared with the same time period in 2007. IBM expects about 70 percent of z10 BC sales to go through IBM's solution providers."
|by Will Runyon||October 21, 2008 in Design, Energy Efficiency, Power & Cooling, SysAdmins, Virtualization |
Permalink | Comments (49) | TrackBack (0)
September 15, 2008
Data center barges on the sea
Last February, we posted about Bill Weihl, Google's green energy czar, and Google's RE<C energy strategy. RE<C stands for Renewable Energy Cheaper than Coal, and is Google's strategy to produce 1 gigawatt of energy capacity through renewable sources like solar, wind and geothermal, to power its data centers.
Two weeks ago, Google filed a patent to house data centers on barges at sea. The idea is to use waves to power the data centers, ocean water to cool them, and a moored distance or seven miles or more to avoid paying taxes. Google concedes the idea is one of many and may not be pursued but they get a lot of credit for thinking differently about the challenges facing data centers.
Picture credit to TimesOnline.
|by Will Runyon||September 15, 2008 |
Permalink | Comments (10) | TrackBack (0)
September 09, 2008
WSJ: Cutting Tech's Energy Bill
The Wall Street Journal's Bill Bulkeley reports in today's edition that "Power use in data centers -- the large, climate-controlled rooms that house a company's computer servers, storage devices and communications switches -- doubled from 2000 to 2006 and now accounts for about 1.5% of U.S. electricity consumption, according to the Environmental Protection Agency."
Bulkeley also reports that IBM, HP and others are finding opportunities and profits in helping companies improve their data centers' efficiency and energy use through better data center design, more energy efficient servers, virtualization and improved cooling. "IBM surprised Wall Street this year when it said its new "Green Data-Center Services" business -- which redesigns customers' data centers and sells energy-efficient products -- signed $300 million in orders in the 2007 fourth quarter. IBM Chief Executive Samuel Palmisano recently told analysts that IBM expects more than 70% of the world's biggest companies 'will modify their data centers significantly in the next five years' to deal with energy shortfalls and rising costs."
Bulkeley also cites that "Hartford Financial Services Group, a Connecticut-based insurer that hired IBM to help it redesign its computer rooms, decided to shut down six of its seven data centers and put its back-up site inside a "green" IBM data center in Boulder, Colo. The new configuration is saving electricity costs and was also "appealing as a good corporate citizen caring about the environment," says Elaine Martinelli, the insurer's senior vice president for technology shared services."
|by Will Runyon||September 9, 2008 |
Permalink | Comments (3) | TrackBack (0)
September 02, 2008
Batman, Iron Man . . . Meet Green Data Center Man
|by Will Runyon||September 2, 2008 in Design, Energy Efficiency, Power & Cooling, Services, SysAdmins |
Permalink | Comments (0) | TrackBack (0)
July 27, 2008
What is Cloud Computing, Anyway?
Irving Wladasky-Berger, chairman emeritus of IBM's Academy of Technology, recently posted a comprehensive overview on Cloud Computing titled "What is Cloud Computing, Anyway?" on his blog. Irving also wrote about this subject in an earlier post, "The Promise and Reality of Cloud Computing." Excerpts below:
"Cloud computing is the kind of wide-ranging initiative that different people can look at from their own point of view and come up with their own, somewhat different definitions. This is not surprising in the early stages of such a comprehensive initiative. When the Internet first broke into the wider world in the mid 1990s, you similarly heard lots of different opinions on what it was and what it would be good for. In reading through the assorted cloud definitions, five key themes keep coming up.
- Well designed consumer and business services
- The evolution of the Internet
- Massive Scalability
- Well engineered cloud delivery centers
- How relevant is cloud computing to most companies?"
Very worthwhile reading.
|by Will Runyon||July 27, 2008 in Design, Services, Who's Who |
Permalink | Comments (2) | TrackBack (0)
July 22, 2008
Greenmonk: Data centers as energy exporters, not energy sinks!
Tom Raftery at Greenmonk recently published a thoughtful post titled Data Centers as energy exporters, not energy sinks! Tom's post includes quotes from Intel's Nick Knupffer and Steve Sams at IBM on progress being made to reduce heat at the chip level.
Tom reports . . . "However, according to the video below, which I found on YouTube, IBM are going way further than I had thought about. They announced their Hydro-Cluster Power 575 series super computers in April. They plan to allow data centers to capture the heat from the servers and export it as hot water for swimming pools, cooking, hot showers, etc. This is how all servers should be plumbed."
|by Will Runyon||July 22, 2008 in Design, Energy Efficiency, Power & Cooling, SysAdmins, Who's Who |
Permalink | Comments (0) | TrackBack (0)
July 02, 2008
Lessons Learned from IBM's Big Green Initiative
GreenerComputing's David Metcalfe has just published a "lessons learned" story on what IBM has learned since announcing its Project Big Green initiative in May 2007 and its Project Big Green 2.0 announcements last month.
The article reports . . . "After 12 months of customer dialogue IBM's Big Green initiative expanded from a focus on hardware energy efficiency in the data centre to a consulting-lead offering for corporate energy efficiency and carbon management. Through its customer outreach IBM learnt six lessons about the commercialization of Green IT:
1. Exploit IT's information management role.
2. Hitch Green IT to data centre refurbishment projects.
3. Tackle corporate energy efficiency and emissions.
4. Differentiate offerings by industry and country.
5. Plan for slow customer adoption.
6. Prepare for investment barriers to IT energy efficiency."
Read the whole story here.
|by Will Runyon||July 2, 2008 in Assessments, Design, Energy Efficiency, Power & Cooling, Services |
Permalink | Comments (1) | TrackBack (0)
The postings on this site are our own and don't necessarily represent the positions, strategies or opinions of our employers.
© Copyright 2007 the respective authors of The Raised Floor Weblog.