The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Avec le développement de la technologie des réseaux, les réseaux de nouvelle génération doivent répondre à de nombreuses nouvelles exigences en matière de fonctions et de performances réseau. Le traitement des champs de paquets trop longs est l'une des exigences et constitue également la base du routage et de la recherche de contenu basés sur l'ID, ainsi que des mécanismes d'ajout/suppression de champs de paquets. Les commutateurs SDN actuels n'offrent pas une bonne prise en charge du traitement des champs trop longs. Dans cet article, nous proposons une série de mécanismes d'optimisation pour les instructions inconscientes du protocole, dans lesquels nous abordons le problème de la prise en charge insuffisante des données trop longues dans les commutateurs SDN existants en étendant la largeur de bits des instructions et en les accélérant à l'aide de jeux d'instructions SIMD. Nous fournissons également une représentation intermédiaire du jeu d'instructions inconscient du protocole pour améliorer l'efficacité du stockage et de la lecture des blocs d'instructions, et réduire davantage le temps d'exécution des blocs d'instructions en les prétraitant. Les expériences montrent que notre approche améliore de 56 % les performances du traitement des données trop longues. Pour les instructions impliquant l'ajout et la suppression de champs de paquets, l'amélioration des performances atteint 455 %. Dans des scénarios de transfert normaux, notre solution réduit la latence de transfert des paquets d'environ 30 %.
Zhiyuan LING
Chinese Academy of Sciences
Xiao CHEN
Chinese Academy of Sciences
Lei SONG
Chinese Academy of Sciences
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Zhiyuan LING, Xiao CHEN, Lei SONG, "Flow Processing Optimization with Accelerated Flow Actions on High Speed Programmable Data Plane" in IEICE TRANSACTIONS on Communications,
vol. E106-B, no. 2, pp. 133-144, February 2023, doi: 10.1587/transcom.2022EBP3069.
Abstract: With the development of network technology, next-generation networks must satisfy many new requirements for network functions and performance. The processing of overlong packet fields is one of the requirements and is also the basis for ID-based routing and content lookup, and packet field addition/deletion mechanisms. The current SDN switches do not provide good support for the processing of overlong fields. In this paper, we propose a series of optimization mechanisms for protocol-oblivious instructions, in which we address the problem of insufficient support for overlong data in existing SDN switches by extending the bit width of instructions and accelerating them using SIMD instruction sets. We also provide an intermediate representation of the protocol-oblivious instruction set to improve the efficiency of storing and reading instruction blocks, and further reduce the execution time of instruction blocks by preprocessing them. The experiments show that our approach improves the performance of overlong data processing by 56%. For instructions involving packet field addition and deletion, the improvement in performance reaches 455%. In normal forwarding scenarios, our solution reduces the packet forwarding latency by around 30%.
URL: https://global.ieice.org/en_transactions/communications/10.1587/transcom.2022EBP3069/_p
Copier
@ARTICLE{e106-b_2_133,
author={Zhiyuan LING, Xiao CHEN, Lei SONG, },
journal={IEICE TRANSACTIONS on Communications},
title={Flow Processing Optimization with Accelerated Flow Actions on High Speed Programmable Data Plane},
year={2023},
volume={E106-B},
number={2},
pages={133-144},
abstract={With the development of network technology, next-generation networks must satisfy many new requirements for network functions and performance. The processing of overlong packet fields is one of the requirements and is also the basis for ID-based routing and content lookup, and packet field addition/deletion mechanisms. The current SDN switches do not provide good support for the processing of overlong fields. In this paper, we propose a series of optimization mechanisms for protocol-oblivious instructions, in which we address the problem of insufficient support for overlong data in existing SDN switches by extending the bit width of instructions and accelerating them using SIMD instruction sets. We also provide an intermediate representation of the protocol-oblivious instruction set to improve the efficiency of storing and reading instruction blocks, and further reduce the execution time of instruction blocks by preprocessing them. The experiments show that our approach improves the performance of overlong data processing by 56%. For instructions involving packet field addition and deletion, the improvement in performance reaches 455%. In normal forwarding scenarios, our solution reduces the packet forwarding latency by around 30%.},
keywords={},
doi={10.1587/transcom.2022EBP3069},
ISSN={1745-1345},
month={February},}
Copier
TY - JOUR
TI - Flow Processing Optimization with Accelerated Flow Actions on High Speed Programmable Data Plane
T2 - IEICE TRANSACTIONS on Communications
SP - 133
EP - 144
AU - Zhiyuan LING
AU - Xiao CHEN
AU - Lei SONG
PY - 2023
DO - 10.1587/transcom.2022EBP3069
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E106-B
IS - 2
JA - IEICE TRANSACTIONS on Communications
Y1 - February 2023
AB - With the development of network technology, next-generation networks must satisfy many new requirements for network functions and performance. The processing of overlong packet fields is one of the requirements and is also the basis for ID-based routing and content lookup, and packet field addition/deletion mechanisms. The current SDN switches do not provide good support for the processing of overlong fields. In this paper, we propose a series of optimization mechanisms for protocol-oblivious instructions, in which we address the problem of insufficient support for overlong data in existing SDN switches by extending the bit width of instructions and accelerating them using SIMD instruction sets. We also provide an intermediate representation of the protocol-oblivious instruction set to improve the efficiency of storing and reading instruction blocks, and further reduce the execution time of instruction blocks by preprocessing them. The experiments show that our approach improves the performance of overlong data processing by 56%. For instructions involving packet field addition and deletion, the improvement in performance reaches 455%. In normal forwarding scenarios, our solution reduces the packet forwarding latency by around 30%.
ER -