Sockets Direct Protocol over InfiniBand in clusters: is it beneficial?

The Sockets Direct Protocol (SDP) had been proposed recently in order to enable sockets based applications to take advantage of the enhanced features provided by InfiniBand architecture. In this paper, we study the benefits and limitations of an implementation of SDP. We first analyze the performance of SDP based on a detailed suite of micro-benchmarks. Next, we evaluate it on two different real application domains: (1) A multitier data-center environment and (2) A Parallel Virtual File System (PVFS). Our micro-benchmark results show that SDP is able to provide up to 2.7 times better bandwidth as compared to the native sockets implementation over InfiniBand (IPoIB) and significantly better latency for large message sizes. Our experimental results also show that SDP is able to achieve a considerably higher performance (improvement of up to 2.4 times) as compared to IPoIB in the PVFS environment. In the data-center environment, SDP outperforms IPoIB for large file transfers inspite of currently being limited by a high connection setup time. However, this limitation is entirely implementation specific and as the InfiniBand software and hardware products are rapidly maturing, we expect this limitation to be overcome soon. Based on this, we have shown that the projected performance for SDP, without the connection setup time, can outperform IPoIB for small message transfers as well.

[1]  David E. Culler,et al.  An Implementation and Analysis of the Virtual Interface Architecture , 1998, Proceedings of the IEEE/ACM SC98 Conference.

[2]  A. Chien,et al.  High Performance Messaging on Workstations: Illinois Fast Messages (FM) for Myrinet , 1995, Proceedings of the IEEE/ACM SC95 Conference.

[3]  Dhabaleswar K. Panda,et al.  Can user-level protocols take advantage of multi-CPU NICs? , 2002, Proceedings 16th International Parallel and Distributed Processing Symposium.

[4]  Rajeev Thakur,et al.  On implementing MPI-IO portably and with high performance , 1999, IOPADS '99.

[5]  Calton Pu,et al.  High Performance Sockets and RPC over Virtual Interface (VI) Architecture , 1999, CANPC.

[6]  W. Richard Stevens,et al.  TCP/IP Illustrated, Volume 2: The Implementation , 1995 .

[7]  Rich Seifert Gigabit Ethernet , 2001, LCN.

[8]  P. Wyckoff,et al.  EMP: Zero-Copy OS-Bypass NIC-Driven Gigabit Ethernet Message Passing , 2001, ACM/IEEE SC 2001 Conference (SC'01).

[9]  Robert B. Ross,et al.  Noncontiguous I/O through PVFS , 2002, Proceedings. IEEE International Conference on Cluster Computing.

[10]  Jin-Soo Kim,et al.  SOVIA: a user-level sockets layer over virtual interface architecture , 2001, Proceedings 42nd IEEE Symposium on Foundations of Computer Science.

[11]  Dhabaleswar K. Panda,et al.  Impact of high performance sockets on data intensive applications , 2003, High Performance Distributed Computing, 2003. Proceedings. 12th IEEE International Symposium on.

[12]  Howard Frazier,et al.  Gigabit Ethernet: From 100 to 1000 Mbps , 1999, IEEE Internet Comput..

[13]  Charles L. Seitz,et al.  Myrinet: A Gigabit-per-Second Local Area Network , 1995, IEEE Micro.

[14]  Dhabaleswar K. Panda,et al.  High performance user level sockets over Gigabit Ethernet , 2002, Proceedings. IEEE International Conference on Cluster Computing.

[15]  Robert B. Ross,et al.  PVFS: A Parallel File System for Linux Clusters , 2000, Annual Linux Showcase & Conference.

[16]  Dhabaleswar K. Panda,et al.  PVFS over InfiniBand: design and performance evaluation , 2003, 2003 International Conference on Parallel Processing, 2003. Proceedings..

[17]  W. Richard Stevens,et al.  TCP/IP Illustrated, Volume 1: The Protocols , 1994 .