Optimizing latency and CPU load in packet processing systems

Paul Emmerich, Daniel Raumer, Alexander Beifuß, Lukas Erlacher, Florian Wohlfart, Torsten M. Runge, Sebastian Gallenmüller, Georg Carle

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

Abstract

High-speed network cards supporting 10 or 40 GbE (Gigabit Ethernet) are available today. Software frameworks for high-speed packet reception and transmission were created to exhaust the performance of these cards. However, these frameworks are not applicable as general-purpose solution. Thus it is necessary to revisit general purpose network 10 software that was designed more than a decade ago. In standard Linux settings, connectivity between applications and physical networks happens via the New API (NAPI). This motivated us to investigate how underlying NIC drivers can be adapted to improve latency in combination with the Linux NAPI. Based on testbed measurements, we propose an optimized algorithm for the NIC driver to dynamically adapt the Interrupt Throttling Rate (ITR). We implemented the algorithm and evaluated it with latency and throughput measurements based on the Linux module of Open vSwitch that operates on top of the NAPI. Our measurements show that our new ITR algorithm improves the packet latency without affecting the CPU load as much as other solutions.

OriginalspracheEnglisch
TitelSymposium on Theory of Modeling and Simulation - DEVS Integrative M and S Symposium, DEVS 2015 - 2015 Spring Simulation Multi-Conference, SpringSim 2015
Redakteure/-innenFranco Davoli, Joel Rodrigues, Imadeldin Mahgoub, Jose Marzo, Malamata Louta, Jose Saldana
Herausgeber (Verlag)The Society for Modeling and Simulation International
Seiten37-44
Seitenumfang8
Auflage9
ISBN (Print)9781510810600
PublikationsstatusVeröffentlicht - 2015
VeranstaltungInternational Symposium on Performance Evaluation of Computer and Telecommunication Systems, SPECTS 2015, Part of the 2015 Summer Simulation Multi-Conference, SummerSim 2015 - Chicago, USA/Vereinigte Staaten
Dauer: 26 Juli 201529 Juli 2015

Publikationsreihe

NameSimulation Series
Nummer9
Band47
ISSN (Print)0735-9276

Konferenz

KonferenzInternational Symposium on Performance Evaluation of Computer and Telecommunication Systems, SPECTS 2015, Part of the 2015 Summer Simulation Multi-Conference, SummerSim 2015
Land/GebietUSA/Vereinigte Staaten
OrtChicago
Zeitraum26/07/1529/07/15

Fingerprint

Untersuchen Sie die Forschungsthemen von „Optimizing latency and CPU load in packet processing systems“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren