Feasibility of Raspberry Pi 2 based Micro Data Centers in Big Data Applications

Many new data centers have been built in recent years in order to keep up with the rising demand for server capacity. These data centers consume a lot of energy, need a lot of cooling equipment and occupy big stretches of land. Energy efficiency of data centers is becoming an increasingly hot topic. Researchers and companies continuously look for ways to bring down energy consumption. This paper takes a look at the possibilities and benefits of using cheap, low-power and widely supported hardware in a micro data center with big data as its main focus. For this purpose, the Raspberry Pi 2 Model B was used as a basis. It was analyzed for performance, scalability, energy consumption and manageability in a data center environment. The result was a fully functional distributed Hadoop setup with a very low power consumption and moderate performance. A highly concurrent, low power setup in a small 1U form factor was proposed as an alternative to traditional rack servers.