PUE

The electricity bill of a datacenter can be very high. Apart from the IT equipment the datacenter itself uses power as well. This power is needed mainly for cooling, but also for lighting, heating of the operator rooms, etc.   

To measure the power used by a datacenter the Power Usage Effectiveness (PUE) is the most used metric. In a white paper published by the Green Grid in February 2007 called "Green Grid Metrics: Describing Data Center Power Efficiency" the use of the PUE metric was introduced.

The Green grid is planning to also publish metrics for carbon usage effectiveness (CUE) and water usage effectiveness (WUE); specially focused on the usage of cooling water.

PUE is determined by dividing the amount of power entering a datacenter by the power used to run the computer equipment in it. PUE is therefore expressed as a ratio, with efficiency improving as it decreases toward 1.

Running a datacenter with a PUE of 2 means that for each watt of power used by the IT equipment an extra watt is used by the datacenter (mainly for cooling). For a datacenter with 1 MW of equipment, another MW is wasted by the datacenter (not leading to better service to customers). In this example, with an average electricity cost of $0.10 per kWh, every year 1000 kW * 24 hours * 365 days * $0.10 = $876,000 is spent on running the datacenter alone (not including the actual IT equipment)! In this example by optimizing the datacenter to a PUE of 1.5 almost half a million dollar per year can be saved.

Over the years the trend has been to decrease the PUE from more than 2 a few years ago to a typical value of 1.8 to 1.5 today. Google claims its datacenters reach a PUE of 1.22, and a Facebook datacenter built in 2011 even claims to reach a PUE of only 1.07, due to excessive optimizations and large scale operations.

It is important to understand that PUE only measures datacenter power efficiency, and not for instance server efficiency, the efficiency of the power supplies used, let alone the amount of useful work that is done by the IT equipment! To optimize the efficiency of the total chain, the useful work done in the datacenter should be divided to the amount of energy entering the datacenter – a metric that is not used in practice today.

Another thing to remember is that a high PUE is not always bad. If a datacenter uses its IT systems very efficiently, for instance by virtualizing all servers to a few large physical systems, much energy is saved. The PUE however will be relatively high as much cooling is needed for the fully loaded physical machines.

Where to build your datacenter

A datacenter can occupy one room of a building, one or more floors, or an entire building. Finding a good location to build a datacenter can be a complex task. There are many variables that determine where a datacenter should be installed. If the datacenter is used for one organization only, it makes sense to place the datacenter inside one of the office buildings. When the datacenter is used by multiple organizations, like in case of an Internet service provider, choosing the location of the datacenter is more delicate.   

Below is a checklist on choosing the location for a datacenter:  

Environment

• Is enough space available to expand the datacenter in the future? The initial datacenter should be designed with as much free space and overcapacity in utilities as possible to allow for growth.
• Is the location vulnerable to flooding (some countries like most parts of The Netherlands are below sea level, are in a vulnerable delta, or are close to a river)? In that case make sure the datacenter is not located at the ground floor or (worse) the basement, but for instance on the third floor.
• Is the datacenter located in a hurricane area?
• What is the chance of an earthquake?
• What is the climate? Datacenter cooling can be easier accomplished and much cheaper in places with a low ambient temperature.
• Is the datacenter close to possible external hazards like a fireworks storage, a waste dump or a chemical plant?
• What is the crime rate? Are there many burglaries in the neighborhood? What about vandalism?
• Is the datacenter near an airport (chance of crashing airplanes)?
• Is the datacenter near an area that is likely to be closed due to unforeseen circumstances (like a car crash on a nearby highway, a forest fire, a military location or a nuclear plant)?
• Is the location close to the home or office of maintenance staff, system managers and external expertise?
• Can the datacenter be reached easily in case of emergencies, maintenance or expansion?
• Are hospitals, police and fire fighters close by?

Visibility

• Is it preferable to have markings on the building showing that this building contains a datacenter?
• Which neighbors does the datacenter have?
• Is the location of the datacenter included in public maps (like http://www.datacentermap.com/)?
• Does the building have windows? Windows are not preferred as they are easy to break into the building and they show to outsiders what is inside the building.

Utilities

• Is it possible to have two independent power providers and internet providers?
• Can cabling routes to the building be determined? Is it possible to have double power and data connections leave the building from two different places?
• Can cabling routes inside the building be determined in a flexible way? Are there multiple paths available to the patch panels, floors and end users?
• Is the datacenter located in a shared building? What if the building must be evacuated? What if the power must be shut down due to maintenance activities of another user of the building?
• Is enough power available to feed the datacenter? How reliable is the power supply?
• Is cheap power available? Can the datacenter use renewable energy like wind power or water power?
• What is the available bandwidth of the external data connections? Is the datacenter close to an Internet exchange point? Are dark fiber connections possible? How reliable are the data connections?

Foreign countries

• Can the country be reached at all times?
• What is the political stability of that country?
• Are there specific laws and regulations you need to adhere to?
• Does the country have a high level of corruption? How reliable is the staff?
• What is the legal status of the data and the datacenter itself?