gflow: towards gpu-based high- performance table matching in openflow switches author : kun qiu, zhe...

15
GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information Networking (ICOIN), 2015 International Conference on Presenter: Tung-yin Chi Date: 2015/3/25 Department of Computer Science and Information Engineering National Cheng Kung University, Taiwan R.O.C.

Upload: barbara-johnson

Post on 19-Jan-2018

222 views

Category:

Documents


0 download

DESCRIPTION

Data Structure: ItemGraph National Cheng Kung University CSIE Computer & Internet Architecture Lab 3

TRANSCRIPT

Page 1: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

GFlow: Towards GPU-based High-Performance Table Matching in

OpenFlow Switches

Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin WangPublisher : Information Networking (ICOIN), 2015 International Conference on Presenter: Tung-yin ChiDate: 2015/3/25

Department of Computer Science and Information Engineering National Cheng Kung University, Taiwan R.O.C.

Page 2: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Introduction

This paper investigates the acceleration of Software based OpenFlow switches, equipped with commodity off-the-shelf hardware, for high-performance table matching.

In our work, we leverage the power of GPUs to accelerate table matching in software-based OpenFlow switches. We propose GFlow, which can handle OpenFlow table matching in a parallel fashion.

Based on our extensive evaluations, we can see the GFlow is 8 to 10 times faster than existing GPU-based matching algorithm.

National Cheng Kung University CSIE Computer & Internet Architecture Lab

2

Page 3: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Data Structure: ItemGraph

National Cheng Kung University CSIE Computer & Internet Architecture Lab

3

Page 4: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

ItemGraph

National Cheng Kung University CSIE Computer & Internet Architecture Lab

4

Page 5: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

ItemGraph

National Cheng Kung University CSIE Computer & Internet Architecture Lab

5

1

4

2 3

Page 6: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

ItemGraph

National Cheng Kung University CSIE Computer & Internet Architecture Lab

6

1

4

2 3

Page 7: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

ItemGraph

National Cheng Kung University CSIE Computer & Internet Architecture Lab

7

1

4

2 3

Page 8: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Matching on the ItemGraph

National Cheng Kung University CSIE Computer & Internet Architecture Lab

8

Page 9: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

The architecture of GFlow

National Cheng Kung University CSIE Computer & Internet Architecture Lab

9

Page 10: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Parallel matching processing the GPU

National Cheng Kung University CSIE Computer & Internet Architecture Lab

10

Page 11: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Experimental Environment

commodity PC CPU : Intel I7 2600K @3.2GHz

• 4 cores/8 threads Memory : 8GB DDR3-1333 GPU : Nvidia GTX 470 @607MHz

• 448 cores, 1280MB GDDR5 memory@3384MHz OS : Fedora 18 (Kernel version 3.8.11-200) The algorithm in GPU is implemented by

OpenCLNational Cheng Kung University CSIE Computer & Internet Architecture Lab

11

Page 12: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Compared with Existing Approaches

Linear search used by OpenFlow vSwitch Linear search with CPU parallel optimization LightFlow, the GPU parallel optimization GFlow, the work in this paper

National Cheng Kung University CSIE Computer & Internet Architecture Lab

12

Page 13: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Experimental Result

National Cheng Kung University CSIE Computer & Internet Architecture Lab

13

Page 14: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Experimental Result

National Cheng Kung University CSIE Computer & Internet Architecture Lab

14

Page 15: GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information

Experimental Result

National Cheng Kung University CSIE Computer & Internet Architecture Lab

15