Abstract
Message passing paradigms provide for many to one messaging patterns that result in receive side resource exhaustion. Traditionally, MPI implementations layered over the Portals network programming interface provided a large default unexpected receive buffer space, the user was expected to configure the buffer size to the application demand, and the application was aborted when the buffer space was overrun. The Portals 4 design provides a set of primitives for implementing scalable resource exhaustion recovery without negatively impacting normal operation. A resource exhaustion recovery protocol for MPI implementations is presented, as well as performance results for an Open MPI implementation of the protocol.
This is a preview of subscription content, log in via an institution to check access.
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
Bonachea, D.: Gasnet specification, v1.1. Tech. Rep. UCB/CSD-02-1207 (October 2002)
Brightwell, R., Lawry, W., Maccabe, A.B., Riesen, R.: Portals 3.0: Protocol building blocks for low overhead communication. In: Proceedings of the 2002 Workshop on Communication Architecture for Clusters (April 2002)
Brightwell, R., Hudson, T., Pedretti, K., Riesen, R., Underwood, K.: Implementation and performance of Portals 3.3 on the Cray XT3. In: Proceedings of the 2005 IEEE International Conference on Cluster Computing (September 2005)
Brightwell, R., Maccabe, A.B., Riesen, R.: Design, implementation, and performance of MPI on Portals 3.0. International Journal of High Performance Computing Applications 17(1), 7–20 (2003)
Canonico, R., Cristaldi, R., Iannello, G.: A scalable flow control algorithm for the Fast Messages communication library. In: Proceedings of the Workshop on Communication, Architecture and Applications for Network-based Parallel Computing (CANPC 1999), pp. 77–90 (1999)
Farreras, M., Cortes, T., Labarta, J., Almasi, G.: Scaling MPI to short-memory MPPs such as BG/L. In: Proceeding of the International Conference on Supercomputing, pp. 209–218 (June 2006)
Liu, J., Panda, D.K.: Implementing efficient and scalable flow control schemes in MPI over InfiniBand. In: Proceedings of the 2004 Workshop on Communication Architecture for Clusters (April 2004)
Petrini, F., Chun Feng, W., Hoisie, A., Coll, S., Frachtenberg, E.: The Quadrics network: High-performance clustering technology. IEEE Micro 22(1), 46–57 (2002)
Underwood, K.D., Coffman, J., Larsen, R., Hemmert, K.S., Barrett, B.W., Brightwell, R., Levenhagen, M.: Enabling flexible collective communication offload with triggered operations. In: IEEE Symposium on High-Performance Interconnects (HotI 2011) (August 2011)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Barrett, B.W., Brightwell, R., Underwood, K.D. (2012). A Low Impact Flow Control Implementation for Offload Communication Interfaces. In: Träff, J.L., Benkner, S., Dongarra, J.J. (eds) Recent Advances in the Message Passing Interface. EuroMPI 2012. Lecture Notes in Computer Science, vol 7490. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-33518-1_8
Download citation
DOI: https://doi.org/10.1007/978-3-642-33518-1_8
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-33517-4
Online ISBN: 978-3-642-33518-1
eBook Packages: Computer ScienceComputer Science (R0)