Abstract
Remote Direct Memory Access (RDMA) is a technology to update a remote machine’s memory without intervention at the receiver side. We evaluate where RDMA can be usefully applied and where it is a loss in Object-Oriented DSM systems. RDMA is difficult to use in modern OO-DSMs due to their support for large address spaces, advanced protocols, and heterogeneity. First, a communication pattern that is based on objects reduces the applicability of bulk RDMA. Second, large address spaces (meaning far larger than that of a single machine) and large numbers of machines require an address space translation scheme to map an object at different addresses on different machines. Finally, RDMA usage is hard since without polling (which would require source code modifications), incoming RDMA messages are hard to notice on time.
Our results show that even with RDMA, update protocols are slower than invalidation protocols. But RDMA can be successfully applied to fetching of objects in an invalidation protocol and improves performance by 20.6%.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Carter, J.B., Bennett, J.K., Zwaenepoel, W.: Techniques for reducing consistency-related communication in distributed shared-memory systems. ACM Trans. Comput. Syst. 13(3), 205–243 (1995)
Eichner, H., Trinitis, C., Klug, T.: Implementation of a DSM-System on Top of InfiniBand. In: Proc. 14th Euromicro Intl. Conf. on Parallel, Distributed, and Network-Based Processing (PDP 2006), Washington, DC, pp. 178–183 (February 2006)
Falsafi, B., Lebeck, A.R., Reinhardt, S.K., Schoinas, I., Hill, M.D., Larus, J.R., Rogers, A., Wood, D.A.: Application-specific protocols for user-level shared memory. In: Supercomputing, Washington, DC, pp. 380–389 (November 1994)
Fang, W., Wang, C.L., Zhu, W., Lau, F.C.M.: A novel adaptive home migration protocol in home-based DSM. In: Proc. of the 2004 IEEE Intl. Conf. on Cluster Computing, San Diego, CA, pp. 215–224 (September 2004)
Iosevich, V., Schuster, A.: Multithreaded Home-Based Lazy Release Consistency over VIA. In: Proc. 19th IEEE Intl. Parallel and Distributed Processing Symp (IPDPS 2004), Santa Fe, New Mexico, pp. 59–70 (April 2004)
Liu, J., Jiang, W., Wyckoff, P., Panda, D., Ashton, D., Buntinas, D., Gropp, W., Toonen, B.: Design and Implementation of MPICH2 over InfiniBand with RDMA Support. In: Int’l Parallel and Distributed Processing Symposium (IPDPS 2004), Santa Fe, NM (April 2004)
Noronha, R., Panda, D.K.: Reducing Diff Overhead in Software DSM Systems using RDMA Operations in InfiniBand. In: Workshop on Remote Direct Memory Access (RDMA): RAIT 2004 (Cluster 2004), San Diego, CA (September 2004)
Protic, J., Tomasevic, M., Milutinovic, V.: A survey of distributed shared memory systems. In: Proc. 28th Hawaii Intl. Conf. on System Sciences (HICSS 1995), pp. 74–84 (January 1995)
Veldema, R., Hofman, R.F.H., Bhoedjang, R.A.F., Bal, H.E.: Runtime-Optimizations for a Java DSM. In: Proc. ACM 2001 Java Grande Conf., San Francisco, CA, pp. 89–98 (June 2001)
Veldema, R., Hofman, R.F.H., Bhoedjang, R.A.F., Jacobs, C.J.H., Bal, H.E.: Source-level global optimizations for fine-grain distributed shared memory systems. In: 8th Symp. on Principles and Practices of Parallel Programming (PPoPP), Snowbird, Utah, pp. 83–92 (June 2001)
Woo, S.C., Ohara, M., Torrie, E., Singh, J.P., Gupta, A.: The SPLASH-2 Programs: Characterization and Methodological Considerations. In: Proc. 22nd Intl. Symp. on Computer Architecture, Santa Margherita Ligure, Italy, pp. 24–36 (June 1995)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Veldema, R., Philippsen, M. (2008). Evaluation of RDMA Opportunities in an Object-Oriented DSM. In: Adve, V., Garzarán, M.J., Petersen, P. (eds) Languages and Compilers for Parallel Computing. LCPC 2007. Lecture Notes in Computer Science, vol 5234. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-85261-2_15
Download citation
DOI: https://doi.org/10.1007/978-3-540-85261-2_15
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-85260-5
Online ISBN: 978-3-540-85261-2
eBook Packages: Computer ScienceComputer Science (R0)