chapter  9
22 Pages

Thermal and Power-Aware Task Scheduling and Data Placement for Storage Centric Datacenters

WithBing Shi, Ankur Srivastava

The primary objective of large-scale storage centric datacenters is to provide data. The advent of applications such as YouTube, etc., has inspired the development of such datacenters. Storage centric datacenters are comprised of large-scale storage devices and CPUs that support data access. People produce and access a tremendous amount of data everyday or even every minute. For example, YouTube serves up to 100 million videos a day [1]; Facebook has 400 million active users and 3 billion photos uploaded each month [2]. These videos and images are stored in datacenters. Therefore, large storage centric datacenters have become an essential component of modern information and communication technology.