ie8 fix
madison

Denial and the coming "data meltdown"

By | December 10, 2007, 8:53pm PST

Summary: Subodh Bapat, Sun Microsystems eco-computing vice president, believes we’ll soon see the first world-class data center meltdown. According to News.com: “You’ll see a massive failure in a year,” Bapat said at a dinner with reporters on Monday. “We are going to see a data center failure of that scale.” “That scale” referred to the problems [...]

Denial and the coming “data meltdown”

Subodh Bapat, Sun Microsystems eco-computing vice president, believes we’ll soon see the first world-class data center meltdown. According to News.com:

“You’ll see a massive failure in a year,” Bapat said at a dinner with reporters on Monday. “We are going to see a data center failure of that scale.”

“That scale” referred to the problems caused by the worm created by Cornell grad student Robert Morris Jr. in 1988. His worm infected about 5 percent of the Unix boxes on the Internet, freaked people out, and helped jump-start the security industry.

The Register adds:

In addition, Sun VP Subodh Bapat warned that 2008 will bring a data center failure of unprecedented scale, causing not only tremendous pains for users but also possible “national security issues.” So, there’s that to look forward to, as our data centers reach enormous proportions.

My new friend (here and here), Nick Carr, wonders about the cause:

It’s unclear what Bapat believes will cause the meltdown, but it appears to be related to the vast electricity requirements of today’s utility-scale server farms. Bapat pointed to a new data center currently being built for a “national lab” that will suck up 50 megawatts of power, more “than a small city would consume.” Bapat said that “utilities are going to become a real problem” for such megacenters.

Seeing this news, Vinnie Mirchandani, fellow Enterprise Irregular, asks why there has been so little media and blogger reaction:

I waited a few days to see how media and blogworld would react. Not much of a reaction at all. Wow, if salesforce.com goes down a few hours everyone’s on it like white on rice, but no statements from IBM, HP, EDS, Accenture (or even Sun’s outsourcing business)…

This muted reaction can be summed into one word: denial. For reasons arising deep in the psyche, ignoring painful truths seems inherent in the human condition. Unfortunately, denial is frequently a component of failed projects, and is one reason many fail so miserably. If project teams and management had greater ability to recognize warning signs in advance of trouble, more projects would succeed.

If you run an IT organization, your best defense against denial, which is the handmaiden of failure, is creating an environment where people feel it’s safe to speak up and be heard. Allowing people to speak their mind is the first step; listening closely, with respect and interest, is the second. Follow steps one and two, and success will follow naturally as step three.

Kick off your day with ZDNet's daily e-mail newsletter. It's the freshest tech news and opinion, served hot. Get it.

Topics

Michael Krigsman is a recognized authority on the causes and prevention of IT failures.

Disclosure

Michael Krigsman

Michael Krigsman writes and speaks about technology in a manner that most observers consider to be fair and balanced. Michael believes that writing about IT failures, which often have complex causes, creates a unique obligation to be reasonable and accurate in both reporting and analysis.

Michael maintains active personal and professional relationships with enterprise technology buyers, vendors, analyst firms (or individual analysts), consultants, and system integrators. As CEO of Asuret, Michael sells and delivers paid services to members of these same groups.

Vendors regularly reimburse Michael's out-of-pocket travel expenses to attend industry conferences and events. Conference organizers frequently waive entry fees when Michael attends industry events. Michael often speaks at industry conferences and events.

He is a member of the Enterprise Irregulars, a loose association of consultants, investors, industry representatives, analysts, and users of enterprise software.

For daily updates on Michael's activities, follow him on Twitter.

Biography

Michael Krigsman

Michael Krigsman is CEO of Asuret, Inc., a consulting company dedicated to reducing technology implementation failures. Asuret's suite of software tools improve the success rate of enterprise software deployments by quantifying and measuring governance issues that cause most project failures. Michael led the research effort underlying Asuret's model of collective intelligence and its practical application to reducing IT failures in consulting environments. He is a recognized authority on the causes and prevention of IT failures and is frequently quoted in the press on IT project and related CIO issues. He is considered an enterprise software industry "influencer" and provides advice to technology buyers, vendors, and services firms.

Previously, Michael served as CEO of Cambridge Publications, which develops tools and processes for software implementations and related business practice automation projects. Michael has been involved with hundreds of software development projects, for companies ranging from small startups to Fortune 500 organizations. Michael graduated with an M.B.A. from Boston University and a B.A. from Bard College. He is a Board member of the America's Cup Hall of Fame and the Herreshoff Marine Museum in Bristol, RI.

7
Comments

Join the conversation!

Just In

Re: Leaving a Lot to the Imagination
elizab 12th Dec 2007
Subodh Bapat doesn't really say what would cause or trigger a data center meltdown. It is the volume of data being transferred to off-site data centers (Internet, bandwidth issue)? Is it the age and capacity of exsiting data centers or replacement rate of existing data centers? Is the power consmption or grid weakness?

Or is he saying, the probability of the event is high enough just based on duration (20 years --- we're bound to have one soon)!

Tangent: This place claims to be 100% solar powered--- http://www.aiso.net/solar-powered-network.asp --- They're probably no where as big as the places Subodh is talking about, but it's most cool.
0 Votes
+ -
I'm surprised...
Erik Engbrecht 11th Dec 2007
That he didn't bring up project Blackbox. A lot of catastrophic failures are
geographically localized.

http://erikengbrecht.blogspot.com/2007/12/data-center-meltdown.html
0 Votes
+ -
Contributr
It is odd he didn't bring it up
mkrigsman@... 11th Dec 2007
In fact, the whole comment was kind of weird, since there was no context, no background, etc.
0 Votes
+ -
RE: Denial and the coming
robertj@... 11th Dec 2007
I'm sure what we're talking about is putting all our eggs in one basket.

One economical(read cheap, insecure, and ill managed) basket.
0 Votes
+ -
RE: Denial and the coming
xfer_rdy 11th Dec 2007
This interesting comment from Bapat. Sun now has some of the lowest power CPUs in the industry addressing the power consumption issues. In fact, upgrading to these systems can earn you a tax credit is some cases. Hmmm there couldn't be some other motive behind the statement, although we all remember those rolling blackouts.

Now, national security issues ?? As Mr Rogers said: "Can we say "selling Fear, Uncertainty and Doubt"? There is going to be a meltdown, but not in the data center due to electricity or cooling. As companies move to outsourced processing on server farms, the weakest like is not in the electrical power, but in the communication systems connecting your business to the farms. All it takes is one smartly placed backhoe kansas and you could be out of luck for a week.
0 Votes
+ -
RE: Denial and the coming
cigar364@... 11th Dec 2007
As a retired electrician (worked both industrial and power co's) all of us know how bad the grid is already maxed out, so this is no suprise to us. No data center should be built without it's own power supply backup independant of the grid (power co).
0 Votes
+ -
Re: Coming Data Meltdown
L3LLL1931@... 11th Dec 2007
As the retired electrician commented, you need a good power supply. On the technical side, there are solar collectors that drive Sterling engines that drive alternator / generators and can be mounted on the roof of building along with wind turbines. These provide sufficient power to run the whole complex in the daytime. With backup batteries and some large generators, there is a good chance there will be adequate power in almost all instances. Critical systems should have Uninterruptable Power Supplies [UPS???s]. If the data center can be placed in a basement, temperature control will be easier to moderate. Add LED lights to reduce power requirements. These are some possible technical solutions for powering the centers. Company politics and other resistance to doing what must be done are just business as usual.
0 Votes
+ -
Subodh Bapat doesn't really say what would cause or trigger a data center meltdown. It is the volume of data being transferred to off-site data centers (Internet, bandwidth issue)? Is it the age and capacity of exsiting data centers or replacement rate of existing data centers? Is the power consmption or grid weakness?

Or is he saying, the probability of the event is high enough just based on duration (20 years --- we're bound to have one soon)!

Tangent: This place claims to be 100% solar powered--- http://www.aiso.net/solar-powered-network.asp --- They're probably no where as big as the places Subodh is talking about, but it's most cool.

Join the conversation!

Formatting +
BB Codes - Note: HTML is not supported in forums
  • [b] Bold [/b]
  • [i] Italic [/i]
  • [u] Underline [/u]
  • [s] Strikethrough [/s]
  • [q] "Quote" [/q]
  • [ol][*] 1. Ordered List [/ol]
  • [ul][*] · Unordered List [/ul]
  • [pre] Preformat [/pre]
  • [quote] "Blockquote" [/quote]
ie8 fix
Click Here
ie8 fix

The best of ZDNet, delivered

ZDNet Newsletters

Get the best of ZDNet delivered straight to your inbox

Facebook Activity

White Papers, Webcasts, & Resources
ie8 fix
ie8 fix