PCAP: performance-aware power capping for the disk drive in the cloud

作者: Mohammed G. Khatib , Zvonimir Bandic

DOI:

关键词:

摘要: Power efficiency is pressing in today's cloud systems. Datacenter architects are responding with various strategies, including capping the power available to computing Throttling bandwidth has been proposed cap usage of disk drive. This work revisits throttling and addresses its shortcomings. We show that, contrary common belief, disk's does not always increase as throughput increases. Furthermore, unnecessarily sacrifices I/O response times by idling disk. propose a technique that resizes queues power. Resizing only imposes no delays on servicing requests, but also enables performance differentiation. We present design implementation PCAP, an agile performance-aware system for PCAP dynamically It operates two modes, tail-latency, making it viable systems service-level differentiation. evaluate different workloads drives. Our experiments reduces up 22%. Further, under 60% requests observe service below 100 ms compared just 10% throttling. worst-case latency 50% increases 32% relative

参考文章(41)
Karthick Rajamani, Charles Lefurgy, Soraya Ghiasi, Juan Rubio, Heather Hanson, Tom Keller, Power management solutions for computer systems and datacenters international symposium on low power electronics and design. pp. 135- 136 ,(2008) , 10.1145/1393921.1393956
Xiaobo Fan, Wolf-Dietrich Weber, Luiz Andre Barroso, Power provisioning for a warehouse-sized computer Proceedings of the 34th annual international symposium on Computer architecture - ISCA '07. ,vol. 35, pp. 13- 23 ,(2007) , 10.1145/1250662.1250665
Simona Boboila, Peter J. Desnoyers, Xiaosong Ma, Devesh Tiwari, Sudharshan S. Vazhkudai, Youngjae Kim, Reducing data movement costs using energy efficient, active computation on SSD international conference on power aware computing and systems. pp. 4- 4 ,(2012)
Yanpei Chen, Sara Alspaugh, Dhruba Borthakur, Randy Katz, Energy efficiency for large-scale MapReduce workloads with significant interactive analysis european conference on computer systems. pp. 43- 56 ,(2012) , 10.1145/2168836.2168842
Enrique V. Carrera, Eduardo Pinheiro, Ricardo Bianchini, Conserving disk energy in network servers international conference on supercomputing. pp. 86- 97 ,(2003) , 10.1145/782814.782829
David Lo, Liqun Cheng, Rama Govindaraju, Luiz André Barroso, Christos Kozyrakis, Towards energy proportionality for large-scale latency-critical workloads international symposium on computer architecture. ,vol. 42, pp. 301- 312 ,(2014) , 10.1145/2678373.2665718
Qingyuan Deng, David Meisner, Luiz Ramos, Thomas F. Wenisch, Ricardo Bianchini, MemScale Proceedings of the sixteenth international conference on Architectural support for programming languages and operating systems - ASPLOS '11. ,vol. 46, pp. 225- 238 ,(2011) , 10.1145/1950365.1950392
Yan Zhang, Sudhanva Gurumurthi, Mircea R Stan, SODA Proceedings of the 44th annual conference on Design automation - DAC '07. pp. 865- 870 ,(2007) , 10.1145/1278480.1278694
Charles Weddle, Mathew Oldham, Jin Qian, An-I Andy Wang, Peter Reiher, Geoff Kuenning, PARAID ACM Transactions on Storage. ,vol. 3, pp. 13- ,(2007) , 10.1145/1288783.1289721
Qingbo Zhu, Zhifeng Chen, Lin Tan, Yuanyuan Zhou, Kimberly Keeton, John Wilkes, Hibernator: helping disk arrays sleep through the winter symposium on operating systems principles. ,vol. 39, pp. 177- 190 ,(2005) , 10.1145/1095809.1095828