Distributed Job Scheduling Using Snapshots

A methodology is developed for the solution of dynamic graph problems through the idea of snapshots. A snapshot of a dynamic graph is a collection of consistent local snapshots, where a local snapshot defines the neighborhood of a process. The methodology is illustrated by solving the distributed job scheduling problem. The solution combines a given snapshot algorithm with existing solutions to the dining philosophers and δ-coloring problems. The time and message complexities of the resulting algorithm depend minimally on global parameters.