Anyone who sets up a measure of performance should be very careful, because people will adapt their behaviour to the metric – instead of to the real world problem.
So for instance, when the government says it will cut waiting lists for medical treatment, it means it will reduce the number of people who have been diagnosed but are waiting for operations.
Hospitals and doctors buckle down and try to reduce the waiting lists by doing the operations more quickly, but when they have done that, they might also do other things. They could take a long while diagnosing people, for instance.
That way, it lakes longer for people to get onto the waiting list. The list of diagnosed people will be shorter, but there will be an invisible waiting list of people who are not yet diagnosed.
It is for this reason that I always take new measures of data centre performance with a pinch of salt. Will they help people improve their performance, or just reward particular behaviour that improves the metric?
Since last week, anyone anxious to assess their data centre for performance has a whole new way to do so – in fact several new ways. The Green Grid has issued new specifications at its recent Members Meeting in Santa Clara.
The vendor/user group’s PUE metric has proved incredibly popular among data centre owners, giving a very simple – arguably too simple – figure to measure the energy efficiency of a data centre. Now the group’s Data Center Maturity Model gives operators a way to plot their progress on a journey from inefficient operation to a super-slick site with no energy waste.
The output of the model will, eventually, be a set of graphic bars, which show progress towards goals in the data centre’s power and cooling, and also in the efficiency of its computing and storage sections.
My first reaction to this is to snort: “that Maturity Model doesn’t look very mature!” And I think even the Green Grid would admit that, in some ways, it doesn’t. But all data centre metrics have to go through a teething process, putting something out for people to use, and then refining it.
Even something as simple as the PUE specification has had to go through refinement, and new metrics such as ERE (which includes reuse of energy) and CUE (which measures the site’s carbon footprint) are evidence that this is happening.
I was told that this Model is really a way to encapsulate best practice, and provide users with a roadmap – a set of guided steps to take.
In that way, it could seem to overlap a lot with EU Code of Conduct for data centres. But even that is not a big issue, according to people involved in the process… because when it comes down to it, there is actually quite a small group of experts on this subject, and they all know each other.
“The Maturity Model is an assessement tool,” said Mark Acton, director of the Uptime Institute’s EMEA network. Uptime focuses on business continuity, but it is also involved in the Green Grid and the EU Code of Conduct.
“You work through the spreadsheet, and check the boxes to find out where you are,” said Acton. With any metric like this, he says you have to decide: “Do you release it, or wait for the perfect metric.”
In the end, you have to applaud people who get something like that out – but we’ll have to watch out for imperfections.
US widening lead over China on AI development, as UK places third in Stanford index…
Amazon to invest a further $4bn into AI start-up Anthropic, doubling its investment as it…
The demand for tech skills is surging, driving economic growth but revealing challenges. Financial costs,…
US Supreme Court tosses Meta's appeal over Cambridge Analytica-linked investor lawsuit, meaning case must proceed
Uber reportedly seeks $10m stake in Chinese autonomous driving firm Pony AI via US IPO,…
iPhone maker reportedly developing next-generation AI large language model for Siri for spring 2026 as…