Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture

Yeim-Kuan Chang, Tung Yin Chi

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Packet classification is a very important component for today's network architecture. It can help or provide packet forwarding and other network functions. With the development of Internet and the emergence of software-defined networking (SDN), the methods designed for the traditional 5-dimensional rule set is not sufficient to process the current rule set that contains rules of 12 or more dimensions. The main problem is to process the rule sets of 12 or more dimensions in high throughput. To achieve high throughput, we study the implementations on GPU where some use a single hash table and others use Binary Range Tree to process the searching. In 12-dimensional rule sets defined by OpenFlow 1.0, 8 fields are in the format of exact value or wildcard, and so using the single hash table or binary range tree is not efficient. Another problem to implement packet classification on GPU is that we must transfer the input data and results via the PCI-E bus that will incur long bus latency. In this paper, we propose a modified hash table to process the fields that contain only exact value or wildcard, and use the compressing method to reduce memory consumption. On the other hand, we implement the proposed method on APU that uses Heterogeneous System Architecture to skip the bus latency. According to the experimental results on AMD A10-7850 APU, our method can achieve the throughput of 1814 MPPS, and can support the rule sets of more than 12K 12-dimensional rules. The achieved throughput is 10 times of the methods on legacy GPU.

Original languageEnglish
Title of host publicationICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks
PublisherIEEE Computer Society
Pages300-305
Number of pages6
ISBN (Electronic)9781728113395
DOIs
Publication statusPublished - 2019 Jul 1
Event11th International Conference on Ubiquitous and Future Networks, ICUFN 2019 - Zagreb, Croatia
Duration: 2019 Jul 22019 Jul 5

Publication series

NameInternational Conference on Ubiquitous and Future Networks, ICUFN
Volume2019-July
ISSN (Print)2165-8528
ISSN (Electronic)2165-8536

Conference

Conference11th International Conference on Ubiquitous and Future Networks, ICUFN 2019
CountryCroatia
CityZagreb
Period19-07-0219-07-05

Fingerprint

Throughput
Network architecture
Internet
Data storage equipment
Graphics processing unit

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Computer Science Applications
  • Hardware and Architecture

Cite this

Chang, Y-K., & Chi, T. Y. (2019). Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture. In ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks (pp. 300-305). [8806091] (International Conference on Ubiquitous and Future Networks, ICUFN; Vol. 2019-July). IEEE Computer Society. https://doi.org/10.1109/ICUFN.2019.8806091
Chang, Yeim-Kuan ; Chi, Tung Yin. / Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture. ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks. IEEE Computer Society, 2019. pp. 300-305 (International Conference on Ubiquitous and Future Networks, ICUFN).
@inproceedings{1f1b26b2744b451eb1378162bbf7851c,
title = "Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture",
abstract = "Packet classification is a very important component for today's network architecture. It can help or provide packet forwarding and other network functions. With the development of Internet and the emergence of software-defined networking (SDN), the methods designed for the traditional 5-dimensional rule set is not sufficient to process the current rule set that contains rules of 12 or more dimensions. The main problem is to process the rule sets of 12 or more dimensions in high throughput. To achieve high throughput, we study the implementations on GPU where some use a single hash table and others use Binary Range Tree to process the searching. In 12-dimensional rule sets defined by OpenFlow 1.0, 8 fields are in the format of exact value or wildcard, and so using the single hash table or binary range tree is not efficient. Another problem to implement packet classification on GPU is that we must transfer the input data and results via the PCI-E bus that will incur long bus latency. In this paper, we propose a modified hash table to process the fields that contain only exact value or wildcard, and use the compressing method to reduce memory consumption. On the other hand, we implement the proposed method on APU that uses Heterogeneous System Architecture to skip the bus latency. According to the experimental results on AMD A10-7850 APU, our method can achieve the throughput of 1814 MPPS, and can support the rule sets of more than 12K 12-dimensional rules. The achieved throughput is 10 times of the methods on legacy GPU.",
author = "Yeim-Kuan Chang and Chi, {Tung Yin}",
year = "2019",
month = "7",
day = "1",
doi = "10.1109/ICUFN.2019.8806091",
language = "English",
series = "International Conference on Ubiquitous and Future Networks, ICUFN",
publisher = "IEEE Computer Society",
pages = "300--305",
booktitle = "ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks",
address = "United States",

}

Chang, Y-K & Chi, TY 2019, Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture. in ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks., 8806091, International Conference on Ubiquitous and Future Networks, ICUFN, vol. 2019-July, IEEE Computer Society, pp. 300-305, 11th International Conference on Ubiquitous and Future Networks, ICUFN 2019, Zagreb, Croatia, 19-07-02. https://doi.org/10.1109/ICUFN.2019.8806091

Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture. / Chang, Yeim-Kuan; Chi, Tung Yin.

ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks. IEEE Computer Society, 2019. p. 300-305 8806091 (International Conference on Ubiquitous and Future Networks, ICUFN; Vol. 2019-July).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture

AU - Chang, Yeim-Kuan

AU - Chi, Tung Yin

PY - 2019/7/1

Y1 - 2019/7/1

N2 - Packet classification is a very important component for today's network architecture. It can help or provide packet forwarding and other network functions. With the development of Internet and the emergence of software-defined networking (SDN), the methods designed for the traditional 5-dimensional rule set is not sufficient to process the current rule set that contains rules of 12 or more dimensions. The main problem is to process the rule sets of 12 or more dimensions in high throughput. To achieve high throughput, we study the implementations on GPU where some use a single hash table and others use Binary Range Tree to process the searching. In 12-dimensional rule sets defined by OpenFlow 1.0, 8 fields are in the format of exact value or wildcard, and so using the single hash table or binary range tree is not efficient. Another problem to implement packet classification on GPU is that we must transfer the input data and results via the PCI-E bus that will incur long bus latency. In this paper, we propose a modified hash table to process the fields that contain only exact value or wildcard, and use the compressing method to reduce memory consumption. On the other hand, we implement the proposed method on APU that uses Heterogeneous System Architecture to skip the bus latency. According to the experimental results on AMD A10-7850 APU, our method can achieve the throughput of 1814 MPPS, and can support the rule sets of more than 12K 12-dimensional rules. The achieved throughput is 10 times of the methods on legacy GPU.

AB - Packet classification is a very important component for today's network architecture. It can help or provide packet forwarding and other network functions. With the development of Internet and the emergence of software-defined networking (SDN), the methods designed for the traditional 5-dimensional rule set is not sufficient to process the current rule set that contains rules of 12 or more dimensions. The main problem is to process the rule sets of 12 or more dimensions in high throughput. To achieve high throughput, we study the implementations on GPU where some use a single hash table and others use Binary Range Tree to process the searching. In 12-dimensional rule sets defined by OpenFlow 1.0, 8 fields are in the format of exact value or wildcard, and so using the single hash table or binary range tree is not efficient. Another problem to implement packet classification on GPU is that we must transfer the input data and results via the PCI-E bus that will incur long bus latency. In this paper, we propose a modified hash table to process the fields that contain only exact value or wildcard, and use the compressing method to reduce memory consumption. On the other hand, we implement the proposed method on APU that uses Heterogeneous System Architecture to skip the bus latency. According to the experimental results on AMD A10-7850 APU, our method can achieve the throughput of 1814 MPPS, and can support the rule sets of more than 12K 12-dimensional rules. The achieved throughput is 10 times of the methods on legacy GPU.

UR - http://www.scopus.com/inward/record.url?scp=85071855785&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071855785&partnerID=8YFLogxK

U2 - 10.1109/ICUFN.2019.8806091

DO - 10.1109/ICUFN.2019.8806091

M3 - Conference contribution

AN - SCOPUS:85071855785

T3 - International Conference on Ubiquitous and Future Networks, ICUFN

SP - 300

EP - 305

BT - ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks

PB - IEEE Computer Society

ER -

Chang Y-K, Chi TY. Hash-based OpenFlow Packet Classification on Heterogeneous System Architecture. In ICUFN 2019 - 11th International Conference on Ubiquitous and Future Networks. IEEE Computer Society. 2019. p. 300-305. 8806091. (International Conference on Ubiquitous and Future Networks, ICUFN). https://doi.org/10.1109/ICUFN.2019.8806091