Conserve the EnvironmentProf. Anand Sivasubramaniam, Head, TCS Innovation Labs - Infrastructure ServicesGreen IT is the term commonly used across the industry that refers to the application of the principles of sustainability to IT services that run within an organization. This domain spans an extremely broad spectrum of business opportunities that promise benefits far beyond improving the energy and carbon footprint of an organization's IT infrastructure. Under our “Improve Environment” theme, we have continued to focus on building sustainable IT services that can broadly be classified into two categories: Greening the IT infrastructure which looks at optimizing Data Center Facilities and Distributed IT infrastructure within an organization, and Using IT as Enabler for Greening which aims at reducing the carbon footprint of a company's operations and the environment in which it operates at large. As part of our endeavour to build practical expertise in this area we focused on carrying out several in-house studies. In this, we conducted two independent Proof of Concept (PoC) studies at TCS' office in Velachery. The first involved studying the Production Data centre in order to correlate power consumption of production servers with their activity to identify any easily implementable improvements. Our study revealed that production servers are underutilized from an activity perspective; are overrated from a power perspective; execute temporally similar workloads over a granularity of weeks; do not idle efficiently; and have power consumptions that are well tracked by their CPU utilizations. Staggering periodic activities on servers to enable deeper sleep states and provisioning based on measurement are effective steps recommended by our study for saving energy. As part of a second PoC study carried out at our IT facility at TCS Velachery we studied the patterns of energy consumption at the activity and business process levels in order to identify focus points where investments on reducing energy consumption should be made. Our study showed that air conditioning was the largest energy consumer with computing being the second largest. Our findings have translated into recommendations for abatement and transformations in terms of policy measures, capex investments and alternative energy which could be applicable to a cross-section of customers. In the last year, we have also focused on theoretical research. Specifically, we studied the problem of consolidating servers to minimize power consumption through frequency control of the servers while complying with performance guarantees. Our research shows that this problem is computationally complex (NP-complete), and have therefore developed an approximation algorithm with a solution cost within a provable ratio of the optimal value. Our experimental simulations show that the algorithm works well in practice even while relaxing some of our analytical assumptions. Tool development is another of our key focus areas and last year we prototyped two new tools:
|