Energy efficient and reliable storage disks

ABSTRACT

Computer hard disks have different failure probability rates that are a function of age and utilization. In this paper the relationship between the disk age, utilization, and failure probabilities are determined. This relationship is used to estimate a safe utilization zone for a disk dependent on its age. Results prove that when disks are operated in safe utilization zones the probability of failure is minimised drastically. Energy consumption is reduced by operating the disks at three power modes: Active, Idle and Sleep. We considered a storage system organized using RAID 1 for the experiments. In RAID 1 data is mirrored to the backup disks from the primary disks. Traditional methods wake up the backup disks when the utilization of the primary disks exceeds 100 percent or both the disks are always active to balance the load between the disks. These methods consume a massive amount of energy. Hence, we designed a policy where we determine the safe utilization levels for the disks to operate with minimum probability failure rates while also conserving energy. This approach is not only reliable but also saves a significant amount of energy.

REFERENCES

• 1: Bianca Schroeder , Garth A. Gibson, Disk failures in the real world: what does an MTTF of 1,000,000 hours mean to you?, Proceedings of the 5th USENIX conference on File and Storage Technologies, p.1-es, February 13-16, 2007, San Jose, CA
• 2: D. Li, and J. Wang, "Conserving Energy in RAID Systems with Conventional Disks," Proc. 3rd Int'l Workshop on Storage Network Architecture and Parallel I/Os, 2005.
• 3: Dakai Zhu , R. Melhem , D. Mosse, The effects of energy management on reliability in real-time embedded systems, Proceedings of the 2004 IEEE/ACM International conference on Computer-aided design, p.35-40, November 07-11, 2004 [doi>10.1109/ICCAD.2004.1382539]
• 4: David P. Helmbold , Darrell D. E. Long , Bruce Sherrod, A dynamic disk spin-down technique for mobile computing, Proceedings of the 2nd annual international conference on Mobile computing and networking, p.130-142, November 1996, Rye, New York, United States [doi>10.1145/236387.236423]
• 5: Eduardo Pinheiro , Wolf-Dietrich Weber , Luiz André Barroso, Failure trends in a large disk drive population, Proceedings of the 5th USENIX conference on File and Storage Technologies, p.2-2, February 13-16, 2007, San Jose, CA
• 6: Emilia Rosti , Giuseppe Serazzi , Evgenia Smirni , Mark S. Squillante, Models of Parallel Applications with Large Computation and I/O Requirements, IEEE Transactions on Software Engineering, v.28 n.3, p.286-307, March 2002 [doi>10.1109/32.991321]
• 7: Enrique V. Carrera , Eduardo Pinheiro , Ricardo Bianchini, Conserving disk energy in network servers, Proceedings of the 17th annual international conference on Supercomputing, June 23-26, 2003, San Francisco, CA, USA [doi>10.1145/782814.782829]
• 8: J. G. Elerath and S. Shah, "Server class disk drives: how reliable are they," IEEE Reliability and Maintainability Symposium, pp. 151--156, Jan 2004.
• 9: Jianfeng Mao , Christos G. Cassandras , Qianchuan Zhao, Optimal Dynamic Voltage Scaling in Energy-Limited Nonpreemptive Systems with Real-Time Constraints, IEEE Transactions on Mobile Computing, v.6 n.6, p.678-688, June 2007 [doi>10.1109/TMC.2007.1024]
• 10: Jianhui Yue , Yifeng Zhu , Zhao Cai, Evaluating memory energy efficiency in parallel I/O workloads, Proceedings of the 2007 IEEE International Conference on Cluster Computing, p.21-30, September 17-20, 2007 [doi>10.1109/CLUSTR.2007.4629213]
• 11: John Zedlewski , Sumeet Sobti , Nitin Garg , Fengzhou Zheng , Arvind Krishnamurthy , Randolph Wang, Modeling Hard-Disk Power Consumption, Proceedings of the 2nd USENIX Conference on File and Storage Technologies, March 31-31, 2003, San Francisco, CA
• 12: L. Lu, P. Varman, and J. Wang, "DiskGroup: Energy Efficient Disk Layout for RAID 1 Systems," Int'l Conf. on Networking, Architecture, and Storage, IEEE Press, pp. 233--242.
• 13: L. Yuan, and G. Qu, "Analysis of energy reduction on dynamic voltage scaling-enabled systems," IEEE Tran. Computer-Aided Design of Integrated Circuits and Systems, vol. 24, no. 12, pp. 1827--1837.
• 14: Sudhanva Gurumurthi , Anand Sivasubramaniam , Mahmut Kandemir , Hubertus Franke, DRPM: dynamic speed control for power management in server class disks, Proceedings of the 30th annual international symposium on Computer architecture, June 09-11, 2003, San Diego, California
• 15: S. W. Son , G. Chen , M. Kandemir , A. Choudhary, Exposing disk layout to compiler for reducing energy consumption of parallel disk based systems, Proceedings of the tenth ACM SIGPLAN symposium on Principles and practice of parallel programming, June 15-17, 2005, Chicago, IL, USA [doi>10.1145/1065944.1065968]
• 16: Seung Woo Son , Guangyu Chen , Ozcan Ozturk , Mahmut Kandemir , Alok Choudhary, Compiler-Directed Energy Optimization for Parallel Disk Based Systems, IEEE Transactions on Parallel and Distributed Systems, v.18 n.9, p.1241-1257, September 2007 [doi>10.1109/TPDS.2007.1056]
• 17: Tajana Šimunić , Luca Benini , Peter Glynn , Giovanni De Micheli, Dynamic power management of laptop hard disk (poster paper), Proceedings of the conference on Design, automation and test in Europe, p.736, March 27-30, 2000, Paris, France [doi>10.1145/343647.344103]
• 18: Yehua Du , Jinxiang Dong , Ming Cai, Dynamic Voltage Scaling of Flash Memory Storage Systems for Low-Power Real-Time Embedded Systems, Proceedings of the Second International Conference on Embedded Software and Systems, p.152-157, December 16-18, 2005 [doi>10.1109/ICESS.2005.49]

INDEX TERMS

Primary Classification:
  B. Hardware
  B.3 MEMORY STRUCTURES
      B.3.2 Design Styles
          Subjects: Mass storage (e.g., magnetic, optical, RAID)

Additional Classification:
  B. Hardware
  B.8 Performance and Reliability
      B.8.1 Reliability, Testing, and Fault-Tolerance

General Terms:
Design, Reliability