We provide three mechanisms to improve performance of file system on NVRAM system. First, the TempFFS can avoid some unnecessary IO and reduce file fragmentations when short-lived files are deleted in TempFFS. Moreover, TempFFS is put between VFS and file system and it does not modify any codes of file system.
Second, we provide an intelligent write-back policy (relo) that considers both locations of dirty blocks in the disk and recency. It not only can reduce the seek time and rotation delay, but also avoids writing back recently-updated dirty pages. Last, we provide a simple transaction support in NVRAM system. It can make sure the file operation is consistent and does not produce any extra disk IO overhead.
In performance results, we not only show the performance improvement per mechanism, but also show the performance improvement of all combinations among three mechanisms.
References
[1] M. Baker, S. Asami, E. Deprit, J. Ousterhout, and M. Seltzer, “Non-Volatile Memory for Fast, Reliable File Systems”, Proceedings of the 5th International Conference on Architectural Support for Programming Languages and Operating Systems, pp.10-22, October 1992.
[2] M. Baker and M. Sullivan, “The Recovery Box: Using Fast Recovery to Provide High Availability in the UNIX environment”, Proceedings of the USENIX Summer Conference, June 1992.
[3] A. Batsakis and R. Burns, “AWOL: An Adaptive Write Optimizations Layer”, Proceedings of the USENIX Conference on File and Storage Technologies, pp.
67-80, February 2008.
[4] A. Ames, N. Bobb, S. Brandt , A. Hiatt, C. Maltzahn, E. Miller, A. Neeman, and D. Tuteja, “Richer File system Metadata Using Links and Attributes”.
Proceedings of the 13th NASA Goddard Conference on Mass Storage and Technologies (TSST’ 05), pp. 49-60, April 2005.
[5] T. Bray, “Bonnie++ benchmark”, http://www.coker.com.au/bonnie++/
[6] K. Chen, R. B. Bunt and D. L. Eager, “Write Caching in Distributed File Systems”, Proceedings of the 15th International Conference on Distributed Computing Systems, pp.457-466, June 1995.
[7] P. M. Chen., “Optimizing Delay in Delayed-Write File Systems”, Technical Report CSE–TR–293–96, University Michigan, May 1996.
[8] P. Chen, W. NG, G. Rajamani, C. Aycock and D. Lowell “The Rio File Cache:
Surviving Operating System Crashes”, Proceedings of the International
Conference on Architectural Support for Programming Languages and Operating Systems, pp. 74-83, October 1996.
[9] R. Desikan, S. W. Keckler, D. Burger and R. Austin. “Assessment of MRAM Technology Characteristics and Architecture”, Technical Report CS-TR-01-36, University of Texas at Austin, Department of Computer Sciences, April 2001.
[10] B. Dipert. “Exotic Memories, Diverse Approaches”, EDN Magazine, pp.56-70, April 2001.
[11] I. H. Doh, J. Choi, D. Lee and S. H. Noh, “Exploiting Non-Volatile RAM to Enhance Flash File System Performance”, Proceedings of the International Conference on Embedded Software, October 2007.
[12] N. K. Edel, D. Tuteja, E. L. Miller and S. A. Brandt, “MRAMFS: A Compressing File System for Non-volatile RAM”, Proceeding of the IEEE Society’s 12th Annual International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunications Systems, pp. 596-603, Oct. 2004.
[13] N. K. Edel, E. L. Miller, K. S. Brandt and S. A. Brandt, “Measuring the Compressibility of Metadata and Small Files for Disk/Nvram Hybrid Storage Systems”, Proceedings of the 2004 International Symposium on Performance Evaluation of Computer and Telecommunication Systems, July 2004.
[14] B. C. Forney, A. C. Arpaci-Dusseau and R. H. Arpaci-Dusseau, “Storage-aware Caching: Revisiting Caching for Heterogeneous Storage Systems”, Proceedings of the First Usenix Conference on File and Storage Technologies, pp.61-74, January 2002.
[15] G. Ganger and M. F. Kaashoek, “Embedded Inodes and Explicit Grouping:
Exploiting Disk Bandwidth for Small Files”, Proceedings of the 1997 USENIX Technical Conference, pp. 1-18, January 1997.
[16] B. S. Gill and D. S. Modha, “WOW: Wise ordering for writes – combining spatial and temporal locality in Non-Volatile Caches”, Proceedings of the 4th Conference on File and Storage Systems, pp. 129–142, Dec. 2005.
[17] K. M. Greenan and E. L. Miller, “Reliability Mechanisms for File Systems using Non-Volatile Memory as a Metadata Store”, Proceedings of the International Conference on Embedded Software, October 2006.
[18] K. M. Greenan and E. L. Miller, “PRIMS: Making NVRAM Suitable for Extremely Reliable Storage”, Proceedings of the 3rd workshop on Hot Topics in System Dependability, April 2007.
[19] T. Haining and D. Long., “Management Policies for Non-volatile Write Caches”, Proceedings of the 18th IEEE International Performance, Computing and Communications Conference, pp. 321–328, February 1999.
[20] R. Y. Hou and Y. N. Patt., “Using Non-volatile Storage to Improve the Reliability of RAID5 Disk Arrays”, Proceedings of the 27th International Symposium on Fault- Tolerant Computing, pp. 206–215, June1997.
[21] S. Jiang, X. Ding, F. Chen,E. Tan, and X. Zhang, “DULO: An Effective Buffer Cache Management Scheme to Exploit Both Temporal and Spatial Locality”, Proceedings of the 4th USENIX Conference on File and Storage Technologies, pp.101-114, December.2005.
[22] J. Katcher, “Postmark: A New File System Benchamrk”. Technical Report TR3022 Network Appliance Inc, October 1997.
[23] A. Kawaguchi, S. Nishioka, and H. Motoda., “A Flash Memory Based File System”, Proceedings of the 1995 USENIX Technical Conference, pp. 155–164, January 1995.
[24] M. Levy, “Memory Products, Chapter Interfacing Microsoft’s Flash File System”, Intel Corporation, pp. 4-318-4-325, 1993.
[25] B. Liskov, S. Ghemawat, R. Gruber, and P. Johnson, “Replication in the Harp File System”, Proceedings of the 1991 Symposium on Operating System Priciples, October 1991.
[26] S. Lim, H. J. Choi, and K. H. Park, “Journal Remap-Based FTL for Journaling File System with Flash Memory”, Proceedings of the High Performance Computation Conference , pp. 192-203, September 2007.
[27] D. E. Lowell, P. M. Chen, “Free Transactions with Rio Vista”. ACM Symposium on Operating Systems Principles, October 1997.
[28] C. Lumb ,J. Schindler ,G. Ganger ,D. Nagle and E. Riedel, “Towards Higher Disk Head Utilization: Extracting Free Bandwidth from Busy Disk Drives”, Proceedings of the Symposium on Operating SystemsDesign and Implementation, pp. 87-102, October 2000.
[29] M. McKusick, W. Joy, S. Leffler, and R. Fabry, “A Fast File System for UNIX”, ACM Transactions on Computer Systems, pp. 181-197, August 1984.
[30] M. McKusick, G. Ganger, “Soft Updates: A Technique for Eliminating Most Synchronous Writes in the Fast File System”, USENIX Annual Technical Conference, pp. 24-24, 1999.
[31] E. Miller, S. Brandt, and D. Long, “HeRMES: High-Performance Reliable
MRAM-Enabled Storage”, Proceedings of the Eighth Workshop on Hot Topics in Operating Systems, pp.95-99, May 2001.
[32] J. C. Mogul, “A Better Update Policy”. Proceedings of the USENIX 1994 Technical Conference, pp.99-111, June 1994.
[33] R. Ohmura, N. Yamasaki, and Y. Anzai, “Device State Recovery in Non-Volatile Main Memory Systems”, Proceedings of the 27th Annual International Computer Software and Applications Conference, pp.16-21, November 2003.
[34] “Ramfs”, http://lwn.net/Articles/156098/
[35] “Reiserfs”, http://www.namesys.com
[36] D. Roselli, “Characteristics of File System Workloads”, Technical Report CSD-98-1029, University of California at Berkeley, December 1998.
[37] D. Roselli, J. Lorch, and T. Anderson, “A comparison of file system workloads”, Proceedings of the USENIX Annual Technical Conference, pp. 41–54, Jun. 2000.
[38] C. Ruemmler and J. Wilkes, “UNIX disk access patterns”, Proceedings of the Winter 1993 USENIX Conference, pp. 405–20, 25–29, January 1993.
[39] M. Rosenblum, and J. Ousterhot, “The Design and Implementation of a Log-Structured File System”, ACM Transaction on Computer Systems, February 1992
[40] M. Satyanarayanan, H. H. Mashburn, P. Kunar, D. C. Steere, and J. J. Kistler,
“Lightweight Recoverable Virtual memory“, ACM Special Interest Group on Operating Systems, pp. 33-57, February 1993.
[41] W. Vogels, “File System Usage in Windows NT 4.0”, Proceedings of the 17th
Symposium on Operating Systems Principles, pp.93-109, December 1999.
[42] A. Wang, G. H. Kuenning, P. Reiher, and G. J. Popek, “Conquest: Better Performance Through a Disk/Persistent-RAM Hybrid File System”. Proceedings of the General Track: 2002 USENIX Annual Technical Conference, pp.15-28, Jane 2002.
[43] J. Wang and Y. Hu., “PROFS – Performance-Oriented Data Reorganization for Log-structured File System on Multi-Zone Disks”, Proceedings of the 9th International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems, pp. 285–293, August 2001.
[44] D. Woodhouse, “The Journaling Flash File System”, Ottawa Linux Symposium, July 2001.
[45] M. Wu, and W. Zwaenepoel, “eNVy: A Non-Volatile, Main Memory Storage System”, Proceedings of the 6th Conference on Architectural Support for Programming Languages and Operating Systems, pp. 86-97, October 1994.
[46] W. Vogels, “File System Usage in Windows NT 4.0”, Proceedings of the 17th Symposium on Operating Systems Principles, pp.93-109, December 1999.
[47] “XFS: A High Performance Journaling File System”, http://oss.sgi.com/projects/xfs