o

Title: Should we worry about starving large jobs? Abstract: Scheduling policies that give priority to small jobs at the expense of large jobs have long been known to perform well for a variety of performance metrics. For example, it has been nearly 50 years since Schrage proved that that Shortest Remaining Processing Time minimizes mean queue length and mean response time. However, whenever such policies are suggested for use in practice, worries about the "starvation" of large jobs arise. In this talk, we will present some results from a recent stream of work that strives to understand what truth there is behind these worries.