As network speed increases, servers struggle to serve all requests directed at them. This challenge is rooted in a partitioned data path where the split between the kernel space networking stack and user space applications induces overheads. To address this challenge, we propose Santa, a architecture to optimize the data path by enabling server applications to (partially) offload packet processing to a generic rule processor. We exemplify Santa by showing how it can drastically accelerate kernel-based UDP packet processingâ\text\euro''a currently neglected domain. Our evaluation focuses on accelerating DNS server for which we find a performance increase by a factor of 5.5.
%0 Conference Paper
%1 Hohlfeld18ITC30
%A Hohlfeld, Oliver
%A Reelfs, Helge
%A Rüth, Jan
%A Schmidt, Florian
%A Zimmermann, Torsten
%A Hiller, Jens
%A Wehrle, Klaus
%B 30th International Teletraffic Congress (ITC 30)
%C Vienna, Austria
%D 2018
%K Session_2:_Mobile_Edge_Computing itc itc30
%T Application-Agnostic Offloading of Datagram Processing
%U https://gitlab2.informatik.uni-wuerzburg.de/itc-conference/itc-conference-public/-/raw/master/itc30/Hohlfeld18ITC30.pdf?inline=true
%X As network speed increases, servers struggle to serve all requests directed at them. This challenge is rooted in a partitioned data path where the split between the kernel space networking stack and user space applications induces overheads. To address this challenge, we propose Santa, a architecture to optimize the data path by enabling server applications to (partially) offload packet processing to a generic rule processor. We exemplify Santa by showing how it can drastically accelerate kernel-based UDP packet processingâ\text\euro''a currently neglected domain. Our evaluation focuses on accelerating DNS server for which we find a performance increase by a factor of 5.5.
@inproceedings{Hohlfeld18ITC30,
abstract = {As network speed increases, servers struggle to serve all requests directed at them. This challenge is rooted in a partitioned data path where the split between the kernel space networking stack and user space applications induces overheads. To address this challenge, we propose Santa, a architecture to optimize the data path by enabling server applications to (partially) offload packet processing to a generic rule processor. We exemplify Santa by showing how it can drastically accelerate kernel-based UDP packet processingâ\text\euro''a currently neglected domain. Our evaluation focuses on accelerating DNS server for which we find a performance increase by a factor of 5.5.},
added-at = {2020-04-29T16:58:36.000+0200},
address = {Vienna, Austria},
author = {Hohlfeld, Oliver and Reelfs, Helge and Rüth, Jan and Schmidt, Florian and Zimmermann, Torsten and Hiller, Jens and Wehrle, Klaus},
biburl = {https://www.bibsonomy.org/bibtex/220fd49f4dd38b2c3353b06c19e539468/itc},
booktitle = {30th International Teletraffic Congress (ITC 30)},
interhash = {2e8ff2eadf37da105bbf052cab79b1eb},
intrahash = {20fd49f4dd38b2c3353b06c19e539468},
keywords = {Session_2:_Mobile_Edge_Computing itc itc30},
timestamp = {2020-05-24T20:14:34.000+0200},
title = {Application-Agnostic Offloading of Datagram Processing},
url = {https://gitlab2.informatik.uni-wuerzburg.de/itc-conference/itc-conference-public/-/raw/master/itc30/Hohlfeld18ITC30.pdf?inline=true},
year = 2018
}