Datacenters are the main infrastructure on top of which cloud computingservices are offered. Such infrastructure may be shared by a large number oftenants and applications generating a spectrum of datacenter traffic. Delaysensitive applications and applications with specific Service Level Agreements(SLAs), generate deadline constrained flows, while other applications initiateflows that are desired to be delivered as early as possible. As a result,datacenter traffic is a mix of two types of flows: deadline and regular. Thereare several scheduling policies for either traffic type with focus onminimizing completion times or deadline miss rate. In this report, we applyseveral scheduling policies to mix traffic scenario while varying the ratio ofregular to deadline traffic. We consider FCFS (First Come First Serve), SRPT(Shortest Remaining Processing Time) and Fair Sharing as deadline agnosticapproaches and a combination of Earliest Deadline First (EDF) with either FCFSor SRPT as deadline-aware schemes. In addition, for the latter, we considerboth cases of prioritizing deadline traffic (Deadline First) and prioritizingregular traffic (Deadline Last). We study both light-tailed and heavy-tailedflow size distributions and measure mean, median and tail flow completion times(FCT) for regular flows along with Deadline Miss Rate (DMR) and averagelateness for deadline flows. We also consider two operation regimes oflightly-loaded (low utilization) and heavily-loaded (high utilization). We findthat performance of deadline-aware schemes is highly dependent on fraction ofdeadline traffic. With light-tailed flow sizes, we find that FCFS performsbetter in terms of tail times and average lateness while SRPT performs betterin average times and deadline miss rate. For heavy-tailed flow sizes, exceptfor tail times, SRPT performs better in all other metrics.
展开▼
