Message Proxies for Efficient, Protected Communication on SMP Clusters

This research addresses the problem ofproviding eficient, protected communication in an SMPcluster without incurring the overhead of system calls or the cost of custom hardware. It anafyzes an approach that uses an idle SMP processor to run a message proxy, a communication process that provides protected access to the nemork. We implement message proxy based communication between a pair of IBM Model G30 SMPs and analyze the resulting overheads. We derive a pevormance model that shows that cachemiss latency within an SMP injluencesmessage proxy performance significantly. Simulations of a suite of ten parallel applications demonstrate that message proxies match the performance of custom hardwarefor three of the ten applications, and are between IO30% slowerfor the other seven applications. A direct cache-update mechanism to reduce cache misses improves the pelformance of message proxies on communication-intensive programs by 7-25%. We conclude that message proxies provide a viable alternative to custom hardware for protected communication.