Scheduling Mix-flows in Commodity Datacenters with Karuna
Authors:
- Li ChenLCLi Chen
- Kai ChenKCKai Chen
- Wei BaiWBWei Bai
- Mohammad AlizadehMAMohammad Alizadeh
Open Access
- 22 August 2016
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 174-187
Abstract
Cloud applications generate a mix of flows with and without deadlines. Scheduling such mix-flows is a key challenge; our experiments show that trivially combining existing schemes for deadline/non-deadline flows is problematic. For example, prioritizing deadline flows hurts flow completion time (FCT) for non-deadline flows, with minor improvement for deadline miss rate. We present Karuna, a first systematic solution for scheduling mix-flows. Our key insight is that deadline flows should meet their deadlines while minimally impacting the FCT of non-deadline flows. To achieve this goal, we design a novel Minimal-impact Congestion control Protocol (MCP) that handles deadline flows with as little bandwidth as possible. For non-deadline flows, we extend an existing FCT minimization scheme to schedule flows with known and unknown sizes. Karuna requires no switch modifications and is backward compatible with legacy TCP/IP stacks. Our testbed experiments and simulations show that Karuna effectively schedules mix-flows, for example, reducing the 95th percentile FCT of non-deadline flows by up to 47.78% at high load compared to pFabric, while maintaining low (<5.8%) deadline miss rateKeywords
This publication has 27 references indexed in Scilit:
- Efficient Coflow Scheduling Without Prior Knowledge2015
- Inside the Social Network's (Datacenter) Network2015
- Global Minimization for Generalized Polynomial Fractional ProgramMathematical Problems in Engineering, 2014
- Towards minimal-delay deadline-driven data center TCP2013
- FCP2013
- Global optimization algorithm for sum of generalized polynomial ratios problemApplied Mathematical Modelling, 2013
- Better never than late2011
- The ns-3 Network Simulator2010
- Solving sum of quadratic ratios fractional programs via monotonic functionApplied Mathematics and Computation, 2009
- Scheduling Algorithms for Multiprogramming in a Hard-Real-Time EnvironmentJournal of the ACM, 1973