Overview
Reducing the energy consumption of IT systems is a major challenge. To date, power management for storage systems has been limited. This aim of this project is to study the factors that affect power usage in storage systems and develop models that will enable the calculation and management of energy resources.
- Power modeling
- Embed power calculations with capacity planning techniques
- Identify the power consumption components for storage controllers
- Provide trade-offs between performance and power
- Develop fine-grained and accurate power metrics
- Design and build a benchmark that measures the power consumed by a single disk drive power under various disk utilizations
- Understand the trade-offs of I/O rate and response time
We develop capacity planning tools that incorporate the power modeling algorithms. These tools predict the expected performance level given an expected host workload and storage configuration. They can also predict the utilization of the various storage components and compute the workload response time. Our technologies are targeted at helping SMBs, archives, and enterprises better manage their storage power resources.
We are able to estimate power usage of running storage systems. The Storage Energy Estimation Agent displays energy consumption estimates within the Tivoli Enterprise Portal. Estimates are given for storage subsystems, storage pools, arrays and logical volumes. The energy information is displayed in both tabular and graphic forms in the ITM Tivoli Enterprise Portal. The graphs allow the user to monitor the energy consumption as it changes over time.
