Abstract
We demonstrate that predictive grouping is an effective mechanism for reducing disk arm movement, thereby simultaneously reducing energy consumption and data access latency. We further demonstrate that predictive grouping has untapped dramatic potential to further improve access performance and limit energy consumption. Data retrieval latencies are considered a major bottleneck, and with growing volumes of data and increased storage needs it is only growing in significance. Data storage infrastructure is therefore a growing consumer of energy at data-center scales, while the individual disk is already a significant concern for mobile computing (accounting for almost a third of a mobile system's energy demands). While improving responsiveness of storage subsystems and hence reducing latencies in data retrieval is often considered contradictory with efforts to reduce disk energy consumption, we demonstrate that predictive data grouping has the potential to simultaneously work towards both these goals. Predictive data grouping has advantages in its applicability compared to both prior approaches to reducing latencies and to reducing energy usage. For latencies, grouping can be performed opportunistically, thereby avoiding the serious performance penalties that can be incurred with prior applications of access prediction (such as predictive prefetching of data). For energy, we show how predictive grouping can even save energy use for an individual disk that is never idle.
Predictive data grouping with effective replication results in a reduction of the overall mechanical movement required to retrieve data. We have built upon our detailed measurements of disk power consumption, and have estimated both the energy expended by a hard disk for its mechanical components, and that needed to move the disk arm. We have further compared, via simulation, three models of predictive grouping of on-disk data, including an optimal arrangement of data that is guaranteed to minimize disk arm movement. These experiments have allowed us to measure the limits of performance improvement achievable with optimal data grouping and replication strategies on a single device, and have further allowed us to demonstrate the potential of such schemes to reduce energy consumption of mechanical components by up to 70%.
- Akyürek, S. and Salem, K. 1995. Adaptive block rearrangement. ACM Trans. Comput. Syst. 13, 2, 89--121. Google Scholar
Digital Library
- Amer, A., Long, D. D. E., and Burns, R. C. 2002. Group-Based management of distributed file caches. In Proceedings of the 22nd International Conference on Distributed Computing Systems (ICDCS), Vienna, Austria. IEEE Computer Society Press. Google Scholar
Digital Library
- Amer, A. M. 2002. Predictive data grouping using successor prediction. Ph.D. thesis, University of California at Santa Cruz. Google Scholar
Digital Library
- Craven, M. and Amer, A. 2005. Predictive reduction of power and latency (purple). In Proceedings of the 22nd IEEE/13th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST). Google Scholar
Digital Library
- Curewitz, K. M., Krishnan, P., and Vitter, J. S. 1993. Practical prefetching via data compression. In Proceedings of the ACM SIGMOD International Conference on Management of Data (SIGMOD), Washington, DC, 257--266. Google Scholar
Digital Library
- Douglis, F., Krishnan, P., and Bershad, B. 1995. Adaptive disk spin-down policies for mobile computers. Comput. Syst. 8, 4 (Fall), 381--413.Google Scholar
- Douglis, F., Krishnan, P., and Marsh, B. 1994. Thwarting the power-hungry disk. In Proceedings of the Winter USENIX Conference, Boston, MA, 293--306. Google Scholar
Digital Library
- Flinn, J. and Satyanarayanan, M. 1999. Energy-Aware adaptation for mobile applications. In Proceedings of the Symposium on Operating Systems Principles (SOSP), 48--63. Google Scholar
Digital Library
- Ganger, G. R. and Kaashoek, M. F. 1997. Embedded inodes and explicit grouping: Exploiting disk bandwidth for small files. In Proceedings of the USENIX Annual Technical Conference, Anaheim, CA, 1--17. Google Scholar
Digital Library
- Golding, R., Bosch, P., Staelin, C., Sullivan, T., and Wilkes, J. 1995. Idleness is not sloth. In Proceedings of the USENIX Technical Conference, New Orleans, LA, 201--212. Google Scholar
Digital Library
- Golding, R., Bosch, P., and Wilkes, J. 1996. Idleness is not sloth. Tech. Rep. HPL-96-140, Hewlett-Packard Laboratories, Palo Alto, California.Google Scholar
- Greenawalt, P. M. 1994. Modeling power management for hard disks. In Proceedings of the 2nd International Workshop on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS). IEEE Computer Society Press, 62--66. Google Scholar
Digital Library
- Griffioen, J. and Appleton, R. 1994. Reducing file system latency using a predictive approach. In Proceedings of the USENIX Summer Technical Conference, 197--207. Google Scholar
Digital Library
- Helmbold, D. P., Long, D. D., and Sherrod, B. 1996. A dynamic disk spin-down technique for mobile computing. In Proceedings of the 2nd Annual ACM International Conference on Mobile Computing and Networking. Google Scholar
Digital Library
- Helmbold, D. P., Long, D. D. E., Sconyers, T. L., and Sherrod, B. 2000. Adaptive disk spin-down for mobile computers. ACM/Baltzer Mobile Netw. Appl. 5, 4. Google Scholar
Digital Library
- Herbster, M. and Warmuth, M. K. 1995. Tracking the best expert. In Proceedings of the 12th International Conference on Machine Learning, Tahoe City, CA. Morgan Kaufmann, 286--294.Google Scholar
- Kistler, J. J. and Satyanarayanan, M. 1991. Disconnected operation in the Coda file system. In Proceedings of the 13th ACM Symposium on Operating Systems Principles (SOSP), Pacific Grove, CA, 213--225. Google Scholar
Digital Library
- Krishnan, P., Long, P., and Vitter, J. S. 1995. Adaptive disk spin-down via optimal rent-to-buy in probabilistic environments. In Proceedings of the 12th International Conference on Machine Learning (ML), Tahoe City, CA. Morgan Kaufman, 322--330.Google Scholar
- Kroeger, T. M. and Long, D. D. E. 1999. The case for efficient file access pattern modeling. In Proceedings of the 7th Workshop on Hot Topics in Operating Systems (HotOS-VII), Rio Rico, AZ. IEEE Computer Society Press, 14--29. Google Scholar
Digital Library
- Kroeger, T. M. and Long, D. D. E. 2001. Design and implementation of a predictive file prefetching algorithm. In Proceedings of the USENIX Annual Technical Conference, Boston, MA. Google Scholar
Digital Library
- Larkby-Lahet, J. A., Santhanakrishnan, G., Amer, A., and Chrysanthis, P. K. 2005. Step: Self-Tuning energy-safe predictors. In Proceedings of the 5th International Conference in Mobile Data Management (MDM). Google Scholar
Digital Library
- Lei, H. and Duchamp, D. 1997. An analytical approach to file prefetching. In Proceedings of the USENIX Annual Technical Conference, Anaheim, CA, 275--288. Google Scholar
Digital Library
- Lorch, J. R. and Smith, A. J. 1998. Software strategies for portable computer energy management. IEEE Personal Commun. 5, 3.Google Scholar
Cross Ref
- McKusick, M. K., Joy, W. N., Leffler, S. J., and Fabry, R. S. 1984. A fast file system for UNIX. ACM Trans. Comput. Syst. 2, 3 (Aug.), 181--197. Google Scholar
Digital Library
- Mummert, L. and Satyanarayanan, M. 1996. Long term distributed file reference tracing: Implementation and experience. Softw. Practice Exper. 26, 6, 705--736. Google Scholar
Digital Library
- Papathanasiou, A. E. and Scott, M. L. 2002. Increasing file system burstiness for energy efficiency. Tech. Rep. TR792, University of Rochester. Google Scholar
Digital Library
- Ruemmler, C. and Wilkes, J. 1993. Unix disk access patterns. In Proceedings of the USENIX Winter Technical Conference, 405--420.Google Scholar
- Russell, S. and Norvig, P. 2003. Artificial Intelligence: A Modern Approach, 2nd ed. Prentice-Hall, Englewood Cliffs, NJ. Google Scholar
Digital Library
- Sedgewick, R. 1992. Algorithms in C++. Addison-Wesley Longman. Google Scholar
Digital Library
- Shriver, E., Gabber, E., Huang, L., and Stein, C. 2001. Storage management for Web proxies. In Proceedings of the USENIX Annual Technical Conference, Boston, MA, 203--216. Google Scholar
Digital Library
- Shriver, E., Small, C., and Smith, K. 1999. Why does file system prefetching work? In Proceedings of the USENIX Annual Technical Conference, Monterey, CA, 71--83. Google Scholar
Digital Library
- Smith, K. A. and Seltzer, M. 1996. A comparison of FFS disk allocation policies. In Proceedings of the USENIX Technical Conference, San Diego, CA, 15--25. Google Scholar
Digital Library
- Staelin, C. and Garcia-Molina, H. 1990a. Clustering active disk data to improve disk performance. Tech. Rep. CS-TR-283-90, Department of Computer Science, Princeton University. February. revised June 1990.Google Scholar
- Staelin, C. and Garcia-Molina, H. 1990b. File system design using large memories. In Proceedings of the 5th Jerusalem Conference on Information Technology (JCIT). IEEE Computer Society Press, 11--21. Google Scholar
Digital Library
- Staelin, C. and Garcia-Molina, H. 1991. Smart filesystems. In Proceedings of the Winter USENIX Conference, 45--51.Google Scholar
- Steere, D. C. 1997. Using dynamic sets to reduce the aggregate latency of data access. Ph.D. thesis, School of Computer Science, Carnegie Mellon University, Pittsburgh, Pennsylvania. Google Scholar
Digital Library
- Vitter, J. S. and Krishnan, P. 1996. Optimal prefetching via data compression. J. ACM 43, 5 (Sept.), 771--793. Google Scholar
Digital Library
- Warmuth, M. and Littlestone, N. 1994. The weighted majority algorithm. Inf. Comput. 108, 2, 212--261. An extended abstract appeared in COLT 89. Google Scholar
Digital Library
- Weissel, A., Beutel, B., and Bellosa, F. 2002. Cooperative I/O: A novel I/O semantics for energy-aware applications. In Proceedings of the Symposium on Operating Systems Design and Implementation (OSDI), Boston, MA. Google Scholar
Digital Library
- Wilkes, J. 1992. Predictive power conservation. Tech. Rep. HPL-CSP-92-5, Concurrent Systems Project, Hewlett-Packard Laboratories, Palo Alto, California. February.Google Scholar
- Zedlewski, J., Sobti, S., Garg, N., Zheng, F., Krishnamurthy, A., and Wang, R. 2003. Modeling hard-disk power consumption. In Proceedings of the USENIX Conference on File and Storage Technologies (FAST), San Francisco, CA. Google Scholar
Digital Library
Index Terms
Predictive data grouping: Defining the bounds of energy and latency reduction through predictive data grouping and replication
Recommendations
Sustainable predictive storage management: on-line grouping for energy and latency reduction
SYSTOR '11: Proceedings of the 4th Annual International Conference on Systems and StorageThe divergence of processor and storage system speeds is one of the most intensely investigated problems in computing. Yet the performance disparity remains, and further, storage energy consumption is rapidly becoming a new critical problem. While ...
Data prefetching to reduce energy use by heterogeneous disk arrays in video servers
NOSSDAV '13: Proceeding of the 23rd ACM Workshop on Network and Operating Systems Support for Digital Audio and VideoAs the storage requirements of a video server increase, disks of different types are often added. This makes it essential to manage the energy requirements of heterogeneous disk arrays. We examine how seek operations affect disk energy consumption, and ...
Radio-Triggered Wake-Up for Wireless Sensor Networks
Power management is an important technique to prolong the lifespan of sensor networks. Many power management protocols employ wake-up/sleep schedules, which are often complicated and inefficient. We present power management schemes that eliminate such ...






Comments