I’m starting day two of the Data Center Summit at Interop, back-to-back with user group events in New York and Boston this week. The first day was a very interesting set of topics:

• Steve Shah of Risingedge presented a session on the “State of the cage.” This fascinating presentation looked at the evolution from mainframes to clustered computers, and from local procedure calls to intra-data-center delays.
• A panel of folks including Michael Baum (CEO of Splunk), James Sayles (CCO of Ecora) and Michael Weider (CTO of IBM’s recently-acquired Watchfire) discussed issues of compliance and privacy in data centers.
• John Carton, formerly at Accenture and now the Senior Director of Web Services at Nature’s Bounty, presented a model for thinking about disaster recovery in data centers.

There was lots to learn from the sessions. Steve pointed out that with the adoption of service-oriented architectures, back-end procedure calls that used to take microseconds now take milliseconds, and that virtualization will make this worse since developers don’t know whether an often-reached machine is local or remote. Michael observed that compliance, which tells people to keep everything, conflicts with privacy, which says that more data makes a breach more risky. And John suggested that companies need to evaluate not only availability, but also how much data they can afford to lose, when setting recovery policies for data centers.

Today’s tracks look at the range of data center models (from on-demand to full colocation with a content delivery network); the issues of power, cooling, and efficiency in greening modern data centers; and automating changes in within data center environments.

I recently wrote a blog for GigaOm’s Earth2Tech site on “Green Code.” The idea is that the quality of code matters. Two coders, writing code for the same application, can have a tremendous difference in efficiency. And that can translate to big differences in power consumption and resource costs — particularly in a virtualized or on-demand environment.

Over here on the Coradiant blog, I can speculate a bit more specifically about what this means. One of the interesting things you can do with user experience is to measure the total processing involved in a page or a user visit.

Because much of the delay on the Internet comes from network performance, two applications with significantly different host efficiency might seem as fast as one another to an end user, so you can’t really measure this just by trying two sites.

But the precision of Real User Monitoring technologies makes even millisecond differences in host processing time clear. And while web operators usually look at average (or percentile) host time, one of the more unusual ways to measure host time is to sum it. This effectively shows you the “total thinking done” for a user’s session.

This can be the start of some pretty fascinating math. Once you know host time per session, you can see how many host-seconds your infrastructure devotes to a visitor. This can show you things like whether a certain class of users is consuming more than its fair share of “heavy” searches.

(Incidentally, on the Coradiant.com site, this often reveals blog spammers from China posting comments about their various vitamins, and more questionable offerings.)

But you can also tie this host time back to IT costs.

I’m teaching a course on data center growth as part of Interop’s Data Center Summit in New York next week (more on this in a later post.) In preparing for that session, I spent a lot of time looking at the cost models behind on-demand hosting, managed servers, collocation, and global CDNs. And it made me realize there are good ways to model IT costs that vary widely according to each business.

Let’s look at combining these two metrics — host time and IT costs — to better understand the business impact of IT.
If you have a good model for IT costs (such as collocation, power, cooling, and storage) and you divide your monthly IT costs by the sum of host time for the month, you know your IT-cost-per-host-second. You don’t want to include bandwidth costs, which aren’t related to the host time.

If you then multiply host-seconds for each user session by that IT cost, you can calculate how much each user session costs you.

This is an excellent basis for evaluating change across releases. It will reflect increased costs in hosting (such as the introduction an application accelerator,) reductions in delay (such as a drop in host time from the AFE’s application acceleration functions reducing the load on servers,) and even changes in pages per session.

You can actually report average IT cost per user session.

As a result, you’ll now know the actual impact of that deployment: Did the reduction in IT-cost-per-host-second outweigh the investment in the AFE? How many weeks did it take to pay the cost back? Is the additional site navigation costing us more?

Of course, there are many other benefits to reducing host time, from user satisfaction to increased capacity to reduced SLA refunds. But this idea of IT-cost-per-host-second is a nice, concrete way to think about what code changes or other modifications to your operations do to your business.

Now back to the fascinating sessions at Web 2.0.

I’m in San Francisco for three days of Web discussions. The Web2 series is always interesting, and offers a good look at what might happen in the future. I just attended a presentation on eBay and open services. The presenter compared eBay’s decision to open its APIs to developers to that of AT&T’s decision to allow third-party devices to connect to the telephone network (the presentation will be available at http://innovation.ebay.com/)
In both cases, the openness led to tremendous advantage.

• For US phone users, billions of dollars in revenue — from answering machines to faxes to modems — were added to the economy. One could even argue that the addition of all these components paved the way for today’s Internet. Imagine if we had to get the telco-approved home router and what that would do to stifle innovation.
• For eBay the market was already building a number of tools for modifying and optimizing both the buying and selling process. At the time, this was achieved through screen-scraping: Pulling down pages and extracting the HTML from them. Not only was this inefficient and error-prone, but every time eBay changed its site, this broke the applications.

Some eye-opening mash-ups — including the combination of Craig’s List housing properties and Google Maps — prompted the folks at eBay to open and document the interfaces to their user base. The results were impressive: Over 55 percent of listings on eBay are submitted by their APIs rather than the traditional eBay web application. That’s billions of dollars in transactions over non-human-web interactions.

The idea of openness is one we spend a lot of time working on at Coradiant. We have a wide range of APIs, from legacy protocols such as SNMP (used by practically every Enterprise Management Software package) to more cutting-edge interfaces like real-time streams of user traffic that can be visualized in interesting ways through browsers or desktop applications.

Our openness has been a deciding factor in many of our customers’ decision to buy TrueSight. Of course, the main focus of our Real User Monitoring appliances is their own interfaces, which operators use to troubleshoot and optimize web apps. But a secondary use is delivering real user data to other destinations. Our ability to get to the individual user sessions and objects, and then to step back and aggregate huge amounts of traffic in ways that make them interesting to the business, is a cornerstone of what we do.

We’re big believers that if we’re open, our customers will surprise us with new things. So far, they haven’t disappointed us.

Now for Jonathan Zittrain, an Oxford Law professor, with the provocatively named Web 2.NO, which I’m choosing over the alternate “Print 2.0″ (or, as Jonathan’s labelling it, “how do I fix my printer driver?”)

Shots from the show at http://www.flickr.com/photos/tags/web20summit/