The CERN Data Centre is the heart of CERN’s entire scientific, administrative, and computing infrastructure. All services, including email, scientific data management and videoconferencing use equipment based in the data centre. A remote extension of the CERN Data Centre is hosted at the Wigner Research Centre for Physics in Hungary, and it provides the extra computing power required to cover CERN’s increasing needs.
See http://cern.ch/go/datacentrebynumbers for the latest numbers on our computing, storage and networking infrastructure.
State-of-the-art network equipment and over 35 000 km of optical fibre provide the infrastructure that is essential for running the computing facilities and services of the laboratory as well as for transferring the colossal amount of LHC data to and from the data centre. The Wigner Data Centre in Hungary, 1200km away, increases the overall capacity, with three 100 Gb/s (gigabit per second) fibre optic lines linking the two sites.
The LHC experiments produce over 50 petabytes of data per year, plus there is around 25 petabytes produced per year for data from other (non-LHC) experiments at CERN. Archiving the vast quantities of data is an essential function at CERN.
Magnetic tapes are used as the main long-term storage medium. Tape is actually a very sophisticated storage material and it can store huge volumes of data. Tape is inexpensive, compact, doesn't consume much electricity, and is durable for long-term storage. With the data tsunami from the LHC, being able to quickly retrieve petabytes of stored data is essential for physicists to make ground-breaking discoveries.
We have many tape robots to ensure efficient storage and retrieval of data. We also have kept all the data from other and previous experiments, although this only represents a small fraction of the total (eg. the LEP experiments is about 400 terabytes. Peanuts compared to today's hundreds of petabytes!).
Accessing tape data is relatively slow, about 1-3 minutes from tape being located, mounted, read and data sent. Often physicists need to access the latest data immediately, so it is also made available on disk servers, where access time is significantly faster.
Air conditioning is a major problem for data centres everywhere in the world at the moment. As processors get faster they also get hotter and at the moment we are getting a greater increase in heat than in performance. Rack machines are even worse as they are densely packed with processors.
Some of the racks at the computing center contain only a few machines in them since there's not enough cooling capacity now to fill them with more machines. The room was designed with one supercomputer in a corner in mind, not several thousand processors!
To maximise the cooling efficiency, we use a Hot/Cold aisle configuration: the front of the racks are facing each other on the 'cold' aisle and expel heat out in their backs to the 'hot' aisle. The doors and roofs placed in the cold aisles increase efficiency by only allowing cold air into the machines. The cold air comes out from the grills in the floor inside the 'cold' aisles.
The cold air is introduced in the building through the big blue pipes coming from the roof and going down to the floor. 3 chillers on the building roof are responsible for cooling down the air. This process consumes no energy during the winter months where cold air is directly taken from outside.
Wigner Data Centre
The Wigner Data Centre in Budapest is managed and operated from CERN, in the same way as the equipment in the CERN Data Centre. Only activities requiring physical access to the equipment are being performed by the Wigner Data Centre staff, e.g. installation of equipment into racks, repairs to the servers, etc.
A capacity of around 30% of the capacity of the CERN Data Centre has been installed at the Wigner Data Centre. We are increasing resources available to us at Wigner all the time, with the aim by 2020 to have a similar level of resources available there as at the CERN Data Centre in Geneva.
The CERN and Wigner Data Centres are connected via three independent and dedicated 100Gb/s (gigabit per second) fibre optic lines. Network latency (the time taken between sending data and receiving on the other end) between the two sites, which are 1800km apart, is about 25 milliseconds.