作者: M. Frans Kaashoek , A. S. Tanenbaum , S. F. Hummel
DOI: 10.1145/70730.70732
关键词:
摘要: Many distributed and parallel applications can make good use of broadcast communication. In this paper we present a (software) protocol that simulates reliable broadcast, even on an unreliable network. Using protocol, application programs need not worry about lost messages. Recovery communication failures is handled automatically transparently by the protocol. normal operation, our more efficient than previously published protocols. An initial implementation 10 MC68020 CPUs connected Mbit/sec Ethernet performs in 1.5 msec.