Data Center Design
Data center design typically follows a key principle called ‘economics of scale’ that fits very well with the ever increasing amount of big data that needs to be stored. ‘Economics of scale’ means the larger the data center, the lower the operational cost. Interesting is the fact that many data Centers are built with commercially available commodity components. One of the largest data centers in world is about eleven times the size of a football field. The massive large-scale data centers are often built with a large number of servers connected through a big interconnection network. In fact many data centers are built with commodity components available on the market. But it is important to consider that the components used in data centers are often very different from those in building supercomputer centers.
It is important that the design of data centres assumes failures as default. All computing Services and user data should never be lost in failure situations and the user data needs to be accessible even during errors and failures in data centers. Errors and failures in data Centers occur on a daily basis. Those are CPU failures, disk I/O problems, network failures, or scheduling software Errors. Often this high reliability criteria is achieved by redundant hardware and servers in data centers that need to be considered in the design. When operating one data center the RAID principle and replication and backups of user data is important. When operating multiple data centers it is further important to keep multiple copies of the data in different geographical locations.
More about Data Center Design
We recommend the following video about this subject: