USING MULTITAIL NETWORKS IN HIGH PERFORMANCE CLUSTERS
Using multiple independent networks (also known as rails) is an emerging technique to overcome bandwidth limitations and enhance fault-tolerance of current high-performance clusters. We present and analyze various venues for exploiting multiple rails. Different rail access policies are presented and compared, including static and dynamic allocation schemes. An analytical lower bound on the number of networks required for static rail allocation is shown. We also present an extensive experimental comparison of the behavior of various allocation schemes in terms of bandwidth and latency. Striping messages over multiple rails can substantially reduce network latency, depending on average message size, network load and allocation scheme. The methods compared include a static rail allocation, a round-robin rail allocation, a dynamic allocation based on local knowledge, and a rail allocation that reserves both end-points of a message before sending it. The latter is shown to perform better than other methods at higher loads: up to 49% better than local-knowledge allocation and 37% better than the round-robin allocation. This allocation scheme also shows lower latency and it saturates on higher loads (for messages large enough). Most importantly, this proposed allocation scheme scales well with the number of rails and message sizes.
- Research Organization:
- Los Alamos National Lab., NM (US)
- Sponsoring Organization:
- US Department of Energy (US)
- DOE Contract Number:
- W-7405-ENG-36
- OSTI ID:
- 776953
- Report Number(s):
- LA-UR-01-1735
- Country of Publication:
- United States
- Language:
- English
Similar Records
Simulation studies of round robin contention in a prioritized CSMA broadcast network. [Dynamic assignment]
Implementation of MP_Lite for the VI Architecture