How migrating 0.0001% of address space saves 12% of energy in hybrid storage

We present a simple, operating- system independent method to reduce the num- ber of seek operations and consequently reduce the energy consumption of a hybrid storage device consisting of a hard disk and a flash memory. Trace-driven simulations show that migrating a tiny amount of the address space (0.0001%) from disk to flash already results in a significant storage energy reduction (12%) at virtually no extra cost. We show that the amount of energy saving depends on which part of the address space is migrated, and we present two indicators for this, namely sequentiality and request frequency. Our simulations show that both are suitable as criterion for energy-saving file placement methods in hybrid storage. We address potential wear problems in the flash subsystem by presenting a simple way to pro- long its expected lifetime.

[1]  Chi Zhang,et al.  Considering the energy consumption of mobile storage alternatives , 2003, 11th IEEE/ACM International Symposium on Modeling, Analysis and Simulation of Computer Telecommunications Systems, 2003. MASCOTS 2003..

[2]  Alan Jay Smith,et al.  Software strategies for portable computer energy management , 1998, IEEE Wirel. Commun..

[3]  Eduardo Pinheiro,et al.  Failure Trends in a Large Disk Drive Population , 2007, FAST.

[4]  Andrew S. Tanenbaum,et al.  Immediate files , 1984, Softw. Pract. Exp..

[5]  Dirk Grunwald,et al.  Massive Arrays of Idle Disks For Storage Archives , 2002, ACM/IEEE SC 2002 Conference (SC'02).

[6]  C.T. Johnson.  The IBM 3850: A mass storage system with disk characteristics , 1975, Proceedings of the IEEE.

[7]  Tei-Wei Kuo,et al.  Efficient management for large-scale flash-memory storage systems with resource conservation , 2005, TOS.

[8]  P. Krishnan,et al.  Flash memory file caching for mobile computers , 1994, 1994 Proceedings of the Twenty-Seventh Hawaii International Conference on System Sciences.

[9]  John Wilkes,et al.  An introduction to disk drive modeling , 1994, Computer.

[10]  Young-Jin Kim,et al.  Energy-efficient file placement techniques for heterogeneous mobile storage systems , 2006, EMSOFT '06.

[11]  Ethan L. Miller,et al.  Measuring the Compressibility of Metadata and Small Files for Disk/NVRAM Hybrid Storage Systems , 2004 .

[12]  D. A. Thompson,et al.  The Future of Magnetic Data Storage Technology , 2000 .

[13]  M. Frans Kaashoek,et al.  Embedded Inodes and Explicit Grouping: Exploiting Disk Bandwidth for Small Files , 1997, USENIX Annual Technical Conference.

[14]  Arvind Krishnamurthy,et al.  Modeling Hard-Disk Power Consumption , 2003, FAST.

[15]  Naehyuck Chang,et al.  Low-Energy Heterogeneous Non-Volatile Memory Systems for Mobile Systems , 2005, J. Low Power Electron..

[16]  Berend Jan van der Zwaag,et al.  Interposing Flash between Disk and DRAM to Save Energy for Streaming Workloads , 2007, 2007 IEEE/ACM/IFIP Workshop on Embedded Systems for Real-Time Multimedia.