Abstract
We develop a holistic framework for adaptively scheduling asynchronous requests in distributed file systems. The system is holistic in that it manages all resources, including network bandwidth, server I/O, server CPU, and client and server memory utilization. It accelerates, defers, or cancels asynchronous requests in order to improve application-perceived performance directly. We employ congestion pricing via online auctions to coordinate the use of system resources by the file system clients so that they can detect shortages and adapt their resource usage. We implement our modifications in the Congestion-Aware Network File System (CA-NFS), an extension to the ubiquitous network file system (NFS). Our experimental result shows that CA-NFS results in a 20% improvement in execution times when compared with NFS for a variety of workloads.
- Alonso, M. and Santonja, V. 1999. A new destage algorithm for disk cache: DOME. In Proceedings of the EUROMICRO Conference.Google Scholar
- Amir, Y., Awerbuch, B., Danilov, C., and Stanton, J. 2005. A cost-benefit flow control for reliable multicast and unicast in overlay networks. Trans. Netw. Google Scholar
Digital Library
- Anderson, E., Spence, S., Swaminathan, R., Kallahalla, M., and Wang, Q. 2005. Quickly finding near-optimal storage designs. Trans. Comput. Syst. Google Scholar
Digital Library
- Awerbuch, B., Azar, Y., Plotkin, S. A., and Waarts, O. 1994. Competitive routing of virtual circuits with unknown duration. In Proceedings of the Symposium on Discrete Algorithms. ACM, New York, NY. Google Scholar
Digital Library
- Baker, M. G., Hartman, J. H., Kupfer, M. D., Shirriff, K. W., and Ousterhout, J. K. 1991. Measurements of a distributed file system. In Proceedings of the Symposium on Operating Systems Principles. ACM, New York, NY. Google Scholar
Digital Library
- Batsakis, A., Burns, R., Kanevsky, A., Lentini, J., and Talpey, T. 2008. AWOL: An adaptive write optimizations in layer. In Proceedings of the Conference on File and Storage Technologies. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Capps, D. 2009. The IOzone Benchmark. http://www.iozone.com.Google Scholar
- Carson, S. D. and Setia, S. 1992. Analysis of the periodic update write policy for disk cache. Trans. Softw. Engin. 18. Google Scholar
Digital Library
- Chambliss, D. D., Alvarez, G. A., Pandey, P., Jadav, D., Xu, J., Menon, R., and Lee, T. P. 2003. Performance virtualization for large-scale storage systems. In Proceedings of the Symposium on Reliable Distributed Systems. IEEE.Google Scholar
- Chun, B. N. and Culler, D. E. 2002. User-centric performance analysis of market-based cluster batch schedulers. In Proceedings of the IEEE/ACM International Symposium on Cluster Computing and the Grid (CCGRID'02). ACM, New York, NY. Google Scholar
Digital Library
- Clearwater, S. 1996. Market-Based Control: A Paradigm for Distributed Resource Allocation World Scientific, Kirkkonummi, Finland. Google Scholar
Digital Library
- Druschel, P. and Banga, G. 1996. Lazy receiver processing (LRP): A network subsystem architecture for server systems. In Proceedings of the Symposium on Operating Systems Design and Implementation. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Fengguang, W. 2007. Adaptive read-ahead in the Linux kernel. http://lwn.net/Articles/155097/.Google Scholar
- Floyd, S. 1994. TCP and explicit congestion notification. Comput. Comm. Rev. 24, 5, 10--23. Google Scholar
Digital Library
- Floyd, S. and Jacobson, V. 1993. Random early detection gateways for congestion avoidance. Trans. Netw. 1, 4, 397--413. 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 Annual Technical Conference. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Goyal, P., Jadav, D., Modha, D. S., and Tewari, R. 2003. CacheCOW: QoS for storage system caches. In Proceedings of the International Workshop on Quality of Service (IWQoS'03). ACM, New York, NY. Google Scholar
Digital Library
- Jacobson, V. 1988. Congestion avoidance and control. In Proceedings of the ACM SIGCOMM Communications Festival. ACM, New York, NY. Google Scholar
Digital Library
- Jones, M. B., Rou, D., and Rou, M. 1997. Cpu reservations and time constraints: Efficient, predictable scheduling of independent activities. In Proceedings of the Symposium of Operating Systems and Principles. ACM, New York, NY. Google Scholar
Digital Library
- Kaldewey, T., Wong, T., Golding, R., Povzner, A., Brandt, S. A., and Maltzahn, C. 2008. Virtualizing disk performance. In Proceedings of the Real-Time and Embedded Technology and Applications Symposium. IEEE, Los Alamitos, CA. Google Scholar
Digital Library
- Karlsson, M., Karamanolis, C., and Zhu, X. 2004. Triage: Performance isolation and differentiation for storage systems. In Proceedings of the International Workshop on Quality of Service.Google Scholar
- Kelly, F., Maulloo, A., and Tan, D. 1998. Rate control in communication networks: Shadow prices, proportional fairness and stability. J. Oper. Resear. Soc. 49.Google Scholar
- Ki, A. and Knowles, A. E. 1997. Adaptive data prefetching using cache information. In Proceedings of the International Conference on Supercomputing. IEEE, New York, NY. Google Scholar
Digital Library
- Kustarz, E., Shepler, S., and A.Wilson. 2008. The new and improved filebench file system benchmarking framework. In Proceedings of the Conference on File and Storage Technologies. USENIX, Berkeley, CA.Google Scholar
- Lai, K., Rasmusson, L., Adar, E., Zhang, L., and Huberman, B. A. 2005. Tycoon: An implementation of a distributed, market-based resource allocation system. Multiagent Grid Syst. 1, 3, 169--182. Google Scholar
Digital Library
- Levy, L., Blumrosen, L., and Nisan, N. 2001. On line markets for distributed object services: The majic system. In Proceedings of the USENIX Symposium on Internet Technologies and Systems (USITS'01). USENIX, Berkeley, CA. Google Scholar
Digital Library
- Li, X., Aboulnaga, A., Salem, K., Sachendina, A., and Gao, S. 2005. Second-tier cache management using write hints. In Proceedings of the Conference on File and Storage Technologies. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Lu, Y., Abdelzaher, T., Lu, C., and Tao, G. 2002. An adaptive control framework for qos guarantees and its application to differentiated caching services. In Proceedings of the International Workshop on Quality of Service. IEEE, Los Alamitos, CA.Google Scholar
- Lumb, C., Merchant, A., and Alvarez, G. 2003. Facade: Virtual storage devices with performance guarantees. In Proceedings of the Conference on File and Storage Technologies. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Massoulie, L. and Roberts, J. 1999. Bandwidth sharing: Objectives and algorithms. In Proceedings of the Annual Joint Conference of the IEEE Computer and Communications Societies (INFOCOM). IEEE, Los Alamitos, CA.Google Scholar
- Mogul, J. 1994. A better update policy. In USENIX Summer Technical Conference. USENIX, Berkeley, CA, 1--10. Google Scholar
Digital Library
- Nam, Y. J. and Park, C. 2002. An adaptive high-low water mark destage algorithm for cached RAID5. In Proceedings of the Pacific Rim Internataional Symposium on Dependable Computing. IEEE, Los Alamitos, CA. Google Scholar
Digital Library
- Nelson, M. N., Welch, B. B., and Ousterhout, J. K. 1988. Caching in the sprite network file system. Trans. Comput. Syst. 6, 1, 1--20. Google Scholar
Digital Library
- Patterson, R. H., Gibson, G. A., Ginting, E., Stodolsky, D., and Zelenka, J. 1995. Informed prefetching and caching. In Proceedings of the ACM Symposium on Operating Systems Principles. ACM, New York, NY. Google Scholar
Digital Library
- Povzner, A., Kaldewey, T., Brandt, S., Golding, R., Wong, T. M., and Maltzahn, C. 2008. Efficient guaranteed disk request scheduling with fahrrad. SIGOPS Oper. Syst. Rev. 42, 4, 13--25. Google Scholar
Digital Library
- Revel, D., McNamee, D., Steere, D., and Walpole, J. 1997. Adaptive prefetching for device independent file I/O. Tech. rep. CSE-97-005, School of Science and Engineering, Oregon Graduate Institute. Google Scholar
Digital Library
- Riska, A., Riedel, E., and Iren, S. 2004. Adaptive disk scheduling for overload management. In Proceedings of the International Conference on the Quantitative Evaluation of Systems. IEEE, Los Alamitos, CA. Google Scholar
Digital Library
- Stoica, I., Abdel-Wahab, H., and Pothen, A. 1995. A Microeconomic Scheduler for Parallel Computers. In Proceedings of the Workshop on Job Scheduling Strategies for Parallel Processing. ACM, New York, NY. Google Scholar
Digital Library
- Strunk, J. D., Thereseka, E., Faloutsos, C., and Ganger, G. R. 2008. Using utility to provision storage systems. In Proceedings of the Conference on File and Storage Technologies. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Uttamchandani, S., Yin, L., Alvarez, G. A., Palmer, J., and Agha, G. 2005. Chameleon: A self-evolving, fully-adaptive resource arbitrator for storage systems. In Proceedings of the USENIX Annual Technical Conference. USENIX, Berkeley, CA. Google Scholar
Digital Library
- Varma, A. and Jacobsen, Q. 1998. Destage algorithms for disk arrays with nonvolatile caches. Trans. Comput. 47, 2, 228--235. Google Scholar
Digital Library
- Welsh, M., Culler, D., and Brewer, E. 2001. Seda: An architecture for well-conditioned scalable internet services. In Proceedings of the Symposium of Operating Systems Principles. ACM, New York, NY. Google Scholar
Digital Library
- Zimmermann, Z. and Killat, U. 2002. Resource marking and fair rate allocation. In Proceedings of the International Conference on Communications. IEEE, Los Alamitos, CA, 1310--1314.Google Scholar
Index Terms
CA-NFS: A congestion-aware network file system
Recommendations
Implementation of a stackable file system for real-time network backup
We propose a backup system based on a stackable mirroring file system, general-purpose mirroring file system (GMFS). This file system mirrors data in real-time on the file system layer. It uses the typical network file system (NFS) and backs up data to ...
Using Eager Strategies to Improve NFS I/O Performance
NAS '11: Proceedings of the 2011 IEEE Sixth International Conference on Networking, Architecture, and StorageTypical NFS clients write in a lazy fashion: they leave dirty pages in the page cache and defer writing to the server until later. This reduces network traffic when applications repeatedly modify the same set of pages. However, this approach can lead to ...






Comments