Architectural Support for Managing Communication in Point-to-Point Distributed Systems

Point-to-point networks provide a natural platform for distributed systems due to their high bandwidth, scalability, and the potential for exploiting communication locality. Managing these networks, however, is complicated by the disparate quality-of-service requirements of distributed applications. Since the various routing and switching schemes are best-suited for diierent classes of traac, the network adapter should be able to dynamically tailor them according to the current need. In this paper, we present and evaluate SPIDER (Scalable Point-to-point Interface DrivER), a network adapter that supports a variety of routing and switching schemes, and includes means for dynamically selecting between them. By exercising low-level control over each network link, SPIDER exploits concurrency amongst the links to reduce both packet latency and intrusion on the protocol software. We evaluate a mesh of SPIDERs through cycle-level simulations over a range of routing and switching schemes. We also demonstrate how this architecture can dynamically partition network bandwidth to meet the performance requirements of disparate traac classes.

[1]  Leonard Kleinrock,et al.  Virtual Cut-Through: A New Computer Communication Switching Technique , 1979, Comput. Networks.

[2]  Paul Hudak,et al.  Memory coherence in shared virtual memory systems , 1986, PODC '86.

[3]  William J. Dally,et al.  Deadlock-Free Message Routing in Multiprocessor Interconnection Networks , 1987, IEEE Transactions on Computers.

[4]  David R. Cheriton,et al.  The VMP network adapter board (NAB): high-performance network communication for multiprocessors , 1988, SIGCOMM '88.

[5]  David R. Cheriton,et al.  The VMP network adapter board (NAB): high-performance network communication for multiprocessors , 1988, SIGCOMM 1988.

[6]  Charles L. Seitz,et al.  Multicomputers: message-passing concurrent computers , 1988, Computer.

[7]  H. T. Kung,et al.  The design of nectar: a network backplane for heterogeneous multicomputers , 1989, ASPLOS III.

[8]  Krishan K. Sabnani,et al.  VLSI implementations of communication protocols-a survey , 1989, IEEE J. Sel. Areas Commun..

[9]  H. T. Kung,et al.  The design of nectar: a network backplane for heterogeneous multicomputers , 1989, ASPLOS 1989.

[10]  Constraint Based Evaluation of Multicomputer Networks , 1990, ICPP.

[11]  Domenico Ferrari,et al.  Client requirements for real-time communication services , 1990, IEEE Communications Magazine.

[12]  Masayuki Murata,et al.  Survey of switching techniques in high-speed networks and their performance , 1990, Proceedings. IEEE INFOCOM '90: Ninth Annual Joint Conference of the IEEE Computer and Communications Societies@m_The Multiple Facets of Integration.

[13]  H. T. Kung,et al.  Network-based multicomputers: an emerging parallel architecture , 1991, Proceedings of the 1991 ACM/IEEE Conference on Supercomputing (Supercomputing '91).

[14]  Kang G. Shin,et al.  Real-time communication in multi-hop networks , 1991, [1991] Proceedings. 11th International Conference on Distributed Computing Systems.

[15]  Onat Menzilcioglu,et al.  Nectar CAB: a high-speed network processor , 1991, [1991] Proceedings. 11th International Conference on Distributed Computing Systems.

[16]  Charles L. Seitz,et al.  A framework for adaptive routing in multicomputer networks , 1989, CARN.

[17]  Xiaodong Zhang,et al.  System effects of interprocessor communication latency in multicomputers , 1991, IEEE Micro.

[18]  Kang G. Shin,et al.  Reliable broadcast algorithms for HARTS , 1991, TOCS.

[19]  S. Konstantinidou,et al.  Chaos router: architecture and performance , 1991, [1991] Proceedings. The 18th Annual International Symposium on Computer Architecture.

[20]  Al Davis,et al.  Mayfly: A general-purpose, scalable, parallel processing architecture , 1992, LISP Symb. Comput..

[21]  William J. Dally,et al.  The message-driven processor: a multicomputer processing node with efficient mechanisms , 1992, IEEE Micro.

[22]  Domenico Talia,et al.  Message-routing systems for transputer-based multicomputers , 1993, IEEE Micro.

[23]  Charles L. Seitz,et al.  A family of routing and communication chips based on the Mosaic , 1993 .

[24]  Lionel M. Ni,et al.  A survey of wormhole routing techniques in direct networks , 1993, Computer.

[25]  Gregory G. Finn,et al.  The use of message-based multicomputer components to construct gigabit networks , 1993, CCRV.

[26]  James William Dolter A programmable routing controller supporting multi-mode routing and switching in distributed real-time systems , 1993 .

[27]  K. K. Ramakrishnan,et al.  Performance Considerations in Designing Network Interfaces , 1993, IEEE J. Sel. Areas Commun..

[28]  Kang G. Shin,et al.  SPIDER: flexible and efficient communication support for point-to-point distributed systems , 1994, 14th International Conference on Distributed Computing Systems.

[29]  IEEE Network Magazine , 2004, IEEE Network.

[30]  William J. Dally,et al.  The torus routing chip , 2005, Distributed Computing.