Peter F. Corbett
Dror G. Feitelson
Abstract
The Vesta parallel file system provides parallel access from compute nodes to files distributed across I/O nodes in a massively parallel computer. Vesta is intended to solve the I/O problems of massively parallel computers executing numerically intensive scientific applications. Vesta has three interesting characteristics: First, it provides a user defined parallel view of file data, and allows user defined partitioning and repartitioning of files without moving data among I/O nodes. The parallel file access semantics of Vesta directly support the operations required by parallel language I/O libraries. Second, Vesta is scalable to a very large number (many hundreds) of I/O and compute nodes and does not contain any sequential bottlenecks in the data-access path. Third, it provides user-directed checkpointing of files during continuing program execution with very little processing overhead.
- [1] P.M. Chen, G.A. Gibson, R.H. Katz, and D.A. Patterson, "An evaluation of redundant arrays of disks using an Amdahl 5890". In SIGMETRICS Conf. Measurement & Modeling of Comput. Syst., pp. 74-85, May 1990. Google Scholar
Digital Library
- [2] P. M. Chen and D. A. Patterson, "Maximizing performance in a striped disk array". In 17th Ann. Intl. Symp. Computer Architecture Conf. Proc., pp. 322-331, May 1990. Google ScholarDigital Library
- [3] A. L. Chervenak and R. H. Katz, "Performance of a disk array prototype". In SIGMETRICS Conf. Measurement & Modeling of Comput. Syst., pp. 188-197, May 1991. Google ScholarDigital Library
- [4] E. DeBenedictis and J. M. del Rosario, "nCUBE parallel I/O software". In 11th Intl. Phoenix Conf. Computers & Communications , pp. 117-124, Apr 1992.Google Scholar
- [5] P. C. Dibble, M. L. Scott, and C. S. Ellis, "Bridge: a high-performance file system [or parallel processors". In 8th Intl. Conf. Distributed Comput. Syst., pp. 154-161, 1988.Google ScholarCross Ref
- [6] J. Edler, J. Lipkis, and E. Schonberg, "Memory management in Symunix II: a design for large-scale shared memory multiprocessors ". In Proc. Workshop on UNIX and Supercomputers, pp. 151-168, USENIX, Sep 1988.Google Scholar
- [7] M. P. Herlihy and J. M. Wing, "Linearizabiliy: a correctness condition for concurrent objects". ACM Trans. Prog. Lang. & Syst. 12(3), pp. 463-492, Jul 1990. Google ScholarDigital Library
- [8] R. H. Katz, G. A. Gibson, and D. A. Patterson, "Disk system architectures for high performance computing". Proc. IEEE 77(12), pp. 1842-1858, Dec 1989.Google ScholarCross Ref
- [9] M. Y. Kim, "Synchronized disk interleaving". IEEE Trans. Comput. C-35(11), pp. 978-988, Nov 1986. Google ScholarDigital Library
- [10] J. J. Kistler and M. Satyanarayanan, "Disconnected operation in the Coda file system". ACM Trans. Comput. Syst. 10(1), pp. 3-25, Feb 1992. Google ScholarDigital Library
- [11] C. Koelbel and P. Mehrotra, "Compiling global name-space parallel loops for distributed execution". IEEE Trans. Parallel & Distributed Syst. 2(4), pp. 440-451, Oct 1991. Google ScholarDigital Library
- [12] D. F. Kotz and C. S. Ellis, "Prefetching in file systems for MIMD multiprocessors". IEEE Trans. Parallel & Distributed Syst. 1(2), pp. 218-230, Apr 1990. Google ScholarDigital Library
- [13] L. Lamport, "How to make a multiprocessor computer that correctly executes multiprocess programs". IEEE Trans. Comput. C-28(9), pp. 690-691, Sep 1979.Google ScholarDigital Library
- [14] E.K. Lee and R.H. Katz, "Performance consequences of parity placement in disk arrays". In 4th Intl. Conf. Architect. Support for Prog. Lang. & Operating Syst., pp. 190-199, Apr 1991. Google ScholarDigital Library
- [15] M.N. Nelson, B.B. Welch, and J.K. Ousterhout, "Caching in the Sprite network file system". ACM Trans. Comput. Syst. 6(1), pp. 134-154, Feb 1988. Google ScholarDigital Library
- [16] S. Ng, D. Lang, and R. Selinger, "Trade-offs between devices and paths in achieving disk interleaving". In 15th Ann. Intl. Symp. Computer Architecture Conf. Proc., pp. 196-201, 1988. Google ScholarDigital Library
- [17] D.A. Patterson, G. Gibson, and R.H. Katz, "A case for redundant arrays of inexpensive disks (RAID)". In SIGMOD Intl. Conf. Management of Data, pp. 109-116, Jun 1988. Google ScholarDigital Library
- [18] P. Pierce, "A concurrent file system for a highly parallel mass storage subsystem". In 4th Conf. Hypercubes, Concurrent Comput., & Appl., vol. I, pp. 155-160, Mar 1989.Google Scholar
- [19] R. Rashid, A. Tevanian, Jr., M. Young, D. Golub, R. Baron, D. Black, W. J. Bolosky, and J. Chew, "Machine-independent virtual memory management for paged uniprocessor and multiprocessor architectures". IEEE Trans. Comput. 37(8), pp. 896- 908, Aug 1988. Google ScholarDigital Library
- [20] A.L.N. Reddy and P. Banerjee, "An evaluation of multiple-disk I/O systems". IEEE Trans. Comput. 38(12), pp. 1680-1690, Dec 1989. Google ScholarDigital Library
- [21] K. Salem and H. Garcia-Molina, "Disk striping". In Proc. Intl. Conf. Data Engineering, pp. 336-342, 1986. Google ScholarDigital Library
- [22] M. Satyanarayanan, "Scalable, secure, and highly available distributed file access". Computer 23(5), pp. 9-21, May 1990. Google ScholarDigital Library
- [23] M. Satyanarayanan, J. J. Kistler, P. Kumar, M. E. Okasaki, E.H. Siegel, and D.C. Steere, "Coda: a highly available file system for a distributed workstation environment". IEEE Trans. Comput. 39(4), pp. 447-459, Apr 1990. Google ScholarDigital Library
- [24] D.C. Stokes, "DCE DIS vs. AFS: a disrtributed file system comparison". In UNITE! 92, pp. 1-12, 1992.Google Scholar
Index Terms
- Overview of the Vesta parallel file system
Recommendations
The Vesta parallel file system
The Vesta parallel file system is designed to provide parallel file access to application programs running on multicomputers with parallel I/O subsystems. Vesta uses a new abstraction of files: a file is not a sequence of bytes, but rather it can be ...
Performance of the Vesta parallel file system
IPPS '95: Proceedings of the 9th International Symposium on Parallel ProcessingVesta is an experimental parallel file system implemented on the IBM SPI. Its main features are support for parallel access from multiple application processes to file, and the ability to partition and re-partition the file data among these processes. ...
Comments