Characterizing locality, evolution, and life span of accesses in enterprise media server workloads
NOSSDAV '02 Proceedings of the 12th international workshop on Network and operating systems support for digital audio and video
Machine Learning
Statistical service assurances for applications in utility grid environments
Performance Evaluation - Special issue: Distributed systems performance
iBOM: A Platform for Intelligent Business Operation Management
ICDE '05 Proceedings of the 21st International Conference on Data Engineering
Power provisioning for a warehouse-sized computer
Proceedings of the 34th annual international symposium on Computer architecture
PowerNap: eliminating server idle power
Proceedings of the 14th international conference on Architectural support for programming languages and operating systems
Predictive algorithms in the management of computer systems
IBM Systems Journal
Clustering of time series data-a survey
Pattern Recognition
Hi-index | 0.00 |
Research data centers are often composed of thousands of diverse computer systems used for ongoing research, development, software regression and hardware compatibility testing. The usage patterns of many of these systems result in periodic non-use and extended periods of idleness. Users routinely fail to ensure that idle machines are powered down prior to overnight or extended absence periods. The annual amount of wasted energy in the HP Bangalore development data center is estimated at 14400 MWh resulting in over 8600 tons of CO2 emissions per year. In this paper, we propose Idle Machine Power Savings (IMPS), which seeks to address potential power cost savings and minimize environmental impact. IMPS consists of a low overhead, highly scalable data acquisition framework enabling the development of algorithms (an artificial neural network is used in the initial prototype) for automatic "extended idle" notifications and optional automatic shutdown of unused computers in data centers. This paper describes our approach, the framework, a prototype implementation and provides preliminary results. The results show an enormous potential for energy savings that translate directly into financial savings and lowered greenhouse gas emissions.