Open AccessImproving Software Middleboxes and Datacenter Task Schedulers
Author(s) -
Hugo Sadok,
Miguel Elias M. Campista,
Luís Henrique M. K. Costa
Publication year - 2019
Language(s) - English
Resource type - Conference proceedings
DOI - 10.5753/ctd.2019.6334
Subject(s) - computer science , stateful firewall , network packet , popularity , task (project management) , software , scheduling (production processes) , distributed computing , computer network , operating system , engineering , psychology , social psychology , operations management , systems engineering
Shared systems have contributed to the popularity ofmany technolo- gies. However, these systems often confront a common challenge: to ensure that resources are fairly divided without compromising utilization efficiency. In this master’s thesis we look at this problem in two distinct systems—software mid- dleboxes and datacenter task schedulers. We first present Sprayer, a system that uses packet spraying to load balance packets to cores in software middleboxes. Our design eliminates the imbalance problems of per-flow solutions and ad- dresses the new challenges ofhandling shared flow states that come with packet spraying. Then, we present Stateful Dominant Resource Fairness (SDRF), a task scheduling policy for datacenters that looks at past allocations and en- forces fairness in the long run. SDRF reduces users’ waiting time on average and improves fairness by increasing the number of completed tasks for users with lower demands, with small impact on high-demand users