Abstract
This paper describes a methodology for efficiently implementing the barrier operation, on clusters with the emerging InfiniBand Architecture (IBA). IBA provides hardware level support for the Remote Direct Memory Access (RDMA) message passing model as well as the multicast operation. This paper describes the design, implementation and evaluation of three barrier algorithms that leverage these mechanisms. Performance evaluation studies indicate that considerable benefits can be achieved using these mechanisms compared to the traditional implementation based on the point-to-point message passing model. Our experimental results show a performance benefit of up to 1.29 times for a 16-node barrier and up to 1.71 times for non-powers-of-2 group size barriers. Each proposed algorithm performs the best for certain ranges of group sizes and the optimal algorithm can be chosen based on this range. To the best of our knowledge, this is the first attempt to characterize the multicast performance in IBA and to demonstrate the benefits achieved by combining it with RDMA operations for efficient implementations of barrier. This framework has significant potential for developing scalable collective communication libraries for IBA-based clusters.
This research is supported in part by Sandia National Laboratory’s contract #30505, Department of Energy’s Grant #DE-FC02-01ER25506, and National Science Foundation’s grants #EIA-9986052 and #CCR-0204429.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Banikazemi, M., Govindaraju, R.K., Blackmore, R., Panda, D.K.: MPI-LAPI: An Efficeint Implementation of MPI for IBM RS/6000 SP Systems. IEEE TPDS, 1081–1093 (October 2001)
Huse, L.P.: Collective communication on dedicated clusters of workstations. In: Margalef, T., Dongarra, J., Luque, E. (eds.) PVM/MPI 1999. LNCS, vol. 1697, pp. 469–476. Springer, Heidelberg (1999)
InfiniBand Trade Association. InfiniBand Architecture Specification, Release 1.0, October 24 (2000)
Kini, S.P.: Efficient Collective Communication using RDMA and Multicast Operations for InfiniBand-Based Clusters. Master Thesis, The Ohio State University (June 2003)
Kini, S.P., Liu, J., Wu, J., Wyckoff, P., Panda, D.K.: Fast and Scalable Barrier using RDMA and Multicast Mechanisms for InfiniBand-Based Clusters. Technical Report, OSU-CISRC-05/03-TR24 (May 2003)
Lawrence Livermore National Laboratory. MVICH: MPI for Virtual Interface Architecture (August 2001)
Liu, J., Wu, J., Kini, S.P., Wyckoff, P., Panda, D.K.: High Performance RDMA-Based MPI Implementation over InfiniBand. In: ICS 2003 (June 2003)
Mellanox Technologies. Mellanox InfiniBand InfiniHost Adapters (July 2002)
Mellanox Technologies. Mellanox IB-Verbs API (VAPI), Rev. 0.97 (May 2003)
Mellor-Crummey, J.M., Scott, M.L.: Algorithms for scalable synchronization on shared-memory multiprocessors. ACM ToCS 9(1), 21–65 (1991)
Message Passing Interface Forum. MPI: A Message Passing Interface. In: Supercomputing 1993, pp. 878–883. IEEE Computer Society Press, Los Alamitos (1993)
Network-Based Computing Laboratory. MVAPICH: MPI for InfiniBand on VAPI Layer (January 2003), http://nowlab.cis.ohio-state.edu/projects/mpi-iba/index.html
Gupta, R., Balaji, P., Panda, D.K., Nieplocha, J.: Efficient Collective Operations using Remote Memory Operations on VIA-Based Clusters. In: IPDPS 2003 (April 2003)
Gupta, R., Tipparaju, V., Nieplocha, J., Panda, D.K.: Efficient Barrier using Remote Memory Operations on VIA-Based Clusters. In: Cluster 2002 (September 2002)
Thakur, R., Gropp, W., Lusk, E.: An Abstract-Device Interface for Implementing Portable Parallel-I/O Interfaces. In: Frontiers 1996, IEEE Computer Society, Los Alamitos (1996)
Topspin Communications, Inc. Topspin InfiniBand Host Channel Adapter, http://www.topspin.com/solutions/hca.html
Tipparaju, V., Nieplocha, J., Panda, D.K.: Fast Collective Operations Using Shared and Remote Memory Access Protocols on Clusters. In: IPDPS 2003 (April 2003)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Kini, S.P., Liu, J., Wu, J., Wyckoff, P., Panda, D.K. (2003). Fast and Scalable Barrier Using RDMA and Multicast Mechanisms for InfiniBand-Based Clusters. In: Dongarra, J., Laforenza, D., Orlando, S. (eds) Recent Advances in Parallel Virtual Machine and Message Passing Interface. EuroPVM/MPI 2003. Lecture Notes in Computer Science, vol 2840. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-39924-7_51
Download citation
DOI: https://doi.org/10.1007/978-3-540-39924-7_51
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-20149-6
Online ISBN: 978-3-540-39924-7
eBook Packages: Springer Book Archive