Hipster: Hybrid Task Manager for Latency-Critical Cloud Workloads
Authors:
- Rajiv NishtalaRNRajiv Nishtala
- Paul CarpenterPCPaul Carpenter
- Vinicius PetrucciVPVinicius Petrucci
- Xavier MartorellXMXavier Martorell
- 1 February 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 409-420
Abstract
In 2013, U.S. data centers accounted for 2.2% of the country's total electricity consumption, a figure that is projected to increase rapidly over the next decade. Many important workloads are interactive, and they demand strict levels of quality-of-service (QoS) to meet user expectations, making it challenging to reduce power consumption due to increasing performance demands. This paper introduces Hipster, a technique that combines heuristics and reinforcement learning to manage latency-critical workloads. Hipster's goal is to improve resource efficiency in data centers while respecting the QoS of the latency-critical workloads. Hipster achieves its goal by exploring heterogeneous multi-cores and dynamic voltage and frequency scaling (DVFS). To improve data center utilization and make best usage of the available resources, Hipster can dynamically assign remaining cores to batch workloads without violating the QoS constraints for the latency-critical workloads. We perform experiments using a 64-bit ARM big.LITTLE platform, and show that, compared to prior work, Hipster improves the QoS guarantee for Web-Search from 80% to 96%, and for Memcached from 92% to 99%, while reducing the energy consumption by up to 18%.Keywords
This publication has 37 references indexed in Scilit:
- Human-level control through deep reinforcement learningNature, 2015
- Containers and Cloud: From LXC to Docker to KubernetesIEEE Cloud Computing, 2014
- Towards energy proportionality for large-scale latency-critical workloadsACM SIGARCH Computer Architecture News, 2014
- The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines, Second editionSynthesis Lectures on Computer Architecture, 2013
- Whare-mapACM SIGARCH Computer Architecture News, 2013
- Bubble-fluxACM SIGARCH Computer Architecture News, 2013
- The tail at scaleCommunications of the ACM, 2013
- Clearing the cloudsACM SIGARCH Computer Architecture News, 2012
- SPEC CPU2006 benchmark descriptionsACM SIGARCH Computer Architecture News, 2006
- Web search for a planet: the google cluster architectureIEEE Micro, 2003