Data Center Operators Double As Energy Brokers
mattOzan writes "When data centers first opened in the 1990s, the tenants paid for space to plug in their servers with a proviso that electricity would be available. As computing power has soared, so has the need for electricity, turning that relationship on its head: electrical capacity is often the central element of lease agreements, and space is secondary. While lease arrangements are often written in the language of real estate, they are essentially power deals. 'Since tenants on average tend to contract for around twice the power they need, Mr. Tazbaz said, those data centers can effectively charge double what they are paying for that power. Generally, the sale or resale of power is subject to a welter of regulations and price controls. For regulated utilities, the average "return on equity" — a rough parallel to profit margins — was 9.25 percent to 9.7 percent for 2010 through 2012.'"
You mean, exclusive contracts can screw you over? WHO KNEW? See also: Every antitrust investigation. Ever.
#fuckbeta #iamslashdot #dicemustdie
sysadmins and parents of newborns get the same amount of sleep.
Wouldn't this be two different things?
At least if the lease agreement is written to the _potential_ power consumption, rather than the actual power used, that is a fundamentally different proposition than taking power and marking it up, or otherwise being registered as a utility.
e.g. you could potentially increase your power draw to the maximum provisioned, and the data center would be contractually bound to be able to provide that level of power, certainly on a particular circuit if there is a redundant power supply failure or other sudden spike in load.
I know in my space (telecom), the logistics of ensuring the correct provisioning of power (access to 48V power, cooling, not overdrawing power slack batteries) can be as complicated as the provisioning of other, more typical server services.
AC power is a scam. First the electricity runs one way, then it stops and it runs the other way. They're selling you the same electrons over and over and over again.
When our name is on the back of your car, we're behind you all the way!
Sure, if you have a half-rack you are going to get screwed in terms of power provisions; you aren't really paying for power. Likewise, if you go with a site that is designed for high density and you are a low density tenant (or the reverse), you are going to get screwed.
For a bigger tenant (every lease I have ever worked on at least), you generally negotiate a fixed PUE* of 1.4-1.6 to cover cooling and system losses. The annualized PUE is usually closer to 1.3-1.4 for most general-purpose sites at reasonable load factors, so they effectively have 14% markup. If you have a really good lease, that markup gets factored out based on actual operating costs.
*PUE(tm) = Power Utilization Effectiveness(tm) = Peak Critical Demand (kW) / Peak Total Demand (kW)
its often not so easy - they need the power because of modern inefficient programs running on supercomputers - your web app goes slow, just add a dozen more cores and a hundred gig of RAM, and then wonder why the cost of hosting is so high.
As a result, you can take your business elsewhere, but they'll be charging as much. That power is not cheap in the first place.
It is time to reduce the requirements of our programs, I understand that datacentres are 2nd only to the airline industry for co2 emissions.