2018/4/23 Dynamic Load-balanced Path Optimization in SDN-based Data Center Networks Author: Yuan-Liang Lan , Kuochen Wang and Yi-Huai Hsu Presenter: Yi-Hsien.

Slides:



Advertisements
Similar presentations
Jaringan Komputer Lanjut Packet Switching Network.
Advertisements

In-Band Flow Establishment for End-to-End QoS in RDRN Saravanan Radhakrishnan.
Study of Distance Vector Routing Protocols for Mobile Ad Hoc Networks Yi Lu, Weichao Wang, Bharat Bhargava CERIAS and Department of Computer Sciences Purdue.
1 Algorithms for Bandwidth Efficient Multicast Routing in Multi-channel Multi-radio Wireless Mesh Networks Hoang Lan Nguyen and Uyen Trang Nguyen Presenter:
Implement a QoS Algorithm for Real-Time Applications in the DiffServ-aware MPLS Network Zuo-Po Huang, *Ji-Feng Chiu, Wen-Shyang Hwang and *Ce-Kuen Shieh.
Distributed Quality-of-Service Routing of Best Constrained Shortest Paths. Abdelhamid MELLOUK, Said HOCEINI, Farid BAGUENINE, Mustapha CHEURFA Computers.
LABERIO: Dynamic load- balanced routing in OpenFlow- enabled networks Author:Hui Long, Yao Shen*, Minyi Guo, Feilong Tang IEEE 27th International.
OpenFlow-Based Server Load Balancing GoneWild Author : Richard Wang, Dana Butnariu, Jennifer Rexford Publisher : Hot-ICE'11 Proceedings of the 11th USENIX.
Load-Balancing Routing in Multichannel Hybrid Wireless Networks With Single Network Interface So, J.; Vaidya, N. H.; Vehicular Technology, IEEE Transactions.
Multipath Routing in Wireless Mesh Networks Mobile Adhoc and Sensor Systems (MASS), 2006 IEEE International Conference on Author: Nagesh S. Nandiraju,
StriD 2 FA: Scalable Regular Expression Matching for Deep Packet Inspection Author: Xiaofei Wang, Junchen Jiang, Yi Tang, Bin Liu, and Xiaojun Wang Publisher:
DBS A Bit-level Heuristic Packet Classification Algorithm for High Speed Network Author : Baohua Yang, Xiang Wang, Yibo Xue, Jun Li Publisher : th.
Memory-Efficient Regular Expression Search Using State Merging Author: Michela Becchi, Srihari Cadambi Publisher: INFOCOM th IEEE International.
SwinTop: Optimizing Memory Efficiency of Packet Classification in Network Author: Chen, Chang; Cai, Liangwei; Xiang, Yang; Li, Jun Conference: Communication.
Early Detection of DDoS Attacks against SDN Controllers
Shadow MACs: Scalable Label- switching for Commodity Ethernet Author: Kanak Agarwal, John Carter, Eric Rozner and Colin Dixon Publisher: HotSDN 2014 Presenter:
Updating Designed for Fast IP Lookup Author : Natasa Maksic, Zoran Chicha and Aleksandra Smiljani´c Conference: IEEE High Performance Switching and Routing.
Binary-tree-based high speed packet classification system on FPGA Author: Jingjiao Li*, Yong Chen*, Cholman HO**, Zhenlin Lu* Publisher: 2013 ICOIN Presenter:
A Fast Regular Expression Matching Engine for NIDS Applying Prediction Scheme Author: Lei Jiang, Qiong Dai, Qiu Tang, Jianlong Tan and Binxing Fang Publisher:
Lightweight Traffic-Aware Packet Classification for Continuous Operation Author: Shariful Hasan Shaikot, Min Sik Kim Presenter: Yen-Chun Tseng Date: 2014/11/26.
Investigating the Prefix-level Characteristics A Case Study in an IPv6 Network Department of Computer Science and Information Engineering, National Cheng.
Packet Classification Using Dynamically Generated Decision Trees
GFlow: Towards GPU-based High- Performance Table Matching in OpenFlow Switches Author : Kun Qiu, Zhe Chen, Yang Chen, Jin Zhao, Xin Wang Publisher : Information.
LOP_RE: Range Encoding for Low Power Packet Classification Author: Xin He, Jorgen Peddersen and Sri Parameswaran Conference : IEEE 34th Conference on Local.
Hierarchical Hybrid Search Structure for High Performance Packet Classification Authors : O˜guzhan Erdem, Hoang Le, Viktor K. Prasanna Publisher : INFOCOM,
LightFlow : Speeding Up GPU-based Flow Switching and Facilitating Maintenance of Flow Table Author : Nobutaka Matsumoto and Michiaki Hayashi Conference:
JA-trie: Entropy-Based Packet Classification Author: Gianni Antichi, Christian Callegari, Andrew W. Moore, Stefano Giordano, Enrico Anastasi Conference.
A Multi-dimensional Packet Classification Algorithm Based on Hierarchical All-match B+ Tree Author: Gang Wang, Yaping Lin*, Jinguo Li, Xin Yao Publisher:
Authors: Jiang Xie, Ian F. Akyildiz
Confluent vs. Splittable Flows
Minimizing latency of critical traffic through SDN
2018/5/8 An approach for detecting encrypted insider attacks on OpenFlow SDN Networks Author: Charles V. Neu , Avelino F. Zorzox , Alex M. S. Orozcoy and.
2018/5/13 CoSwitch: A Cooperative Switching Design for Software Defined Data Center Networking Author: Yue ZhangKai ,Zheng, Chengchen Hu, Kai Chen, Yi.
2018/6/5 Centralized Retransmission Management with SDN in Multihop Wireless Access Network Author: Bong-Hwan Oh , Jaiyoung Lee Presenter: Yi-Hsien Wu.
2018/6/26 An Energy-efficient TCAM-based Packet Classification with Decision-tree Mapping Author: Zhao Ruan, Xianfeng Li , Wenjun Li Publisher: 2013.
2018/7/4 QoS-aware Multipathing in datacenters using effective bandwidth estimation and SDN Author: Runxin Wang, Simone Mangiante, Alan Davy, Lei Shi,
Tapping Into The Unutilized Router Processing Power
Statistical Optimal Hash-based Longest Prefix Match
2018/11/19 Source Routing with Protocol-oblivious Forwarding to Enable Efficient e-Health Data Transfer Author: Shengru Li, Daoyun Hu, Wenjian Fang and.
Parallel Processing Priority Trie-based IP Lookup Approach
MR2RP: The Multi-Rate and Multi-Range Routing Protocol for IEEE 802
2018/12/10 Energy Efficient SDN Commodity Switch based Practical Flow Forwarding Method Author: Amer AlGhadhban and Basem Shihada Publisher: 2016 IEEE/IFIP.
2018/12/29 A Novel Approach for Prefix Minimization using Ternary trie (PMTT) for Packet Classification Author: Sanchita Saha Ray, Abhishek Chatterjee,
2019/1/1 High Performance Intrusion Detection Using HTTP-Based Payload Aggregation 2017 IEEE 42nd Conference on Local Computer Networks (LCN) Author: Felix.
Congestion Control in SDN-Enabled Networks
2019/1/3 Exscind: Fast Pattern Matching for Intrusion Detection Using Exclusion and Inclusion Filters Next Generation Web Services Practices (NWeSP) 2011.
Memory-Efficient Regular Expression Search Using State Merging
Virtual TCAM for Data Center Switches
A Small and Fast IP Forwarding Table Using Hashing
A New String Matching Algorithm Based on Logical Indexing
2019/5/2 Using Path Label Routing in Wide Area Software-Defined Networks with OpenFlow ICNP = International Conference on Network Protocols Presenter:Hung-Yen.
Compact DFA Structure for Multiple Regular Expressions Matching
2019/5/3 A De-compositional Approach to Regular Expression Matching for Network Security Applications Author: Eric Norige Alex Liu Presenter: Yi-Hsien.
2019/5/5 A Flexible Wildcard-Pattern Matching Accelerator via Simultaneous Discrete Finite Automata Author: Hsiang-Jen Tsai, Chien-Chih Chen, Yin-Chi Peng,
2019/5/13 A Weighted ECMP Load Balancing Scheme for Data Centers Using P4 Switches Presenter:Hung-Yen Wang Authors:Peng Wang, George Trimponias, Hong Xu,
Pipelined Architecture for Multi-String Matching
SDN-Guard: DoS Attacks Mitigation in SDN Networks
QoS Constrained Path Optimization Algorithm in NFV/SDN Environment
Fast Network Congestion Detection And Avoidance Using P4
OpenSec:Policy-Based Security Using Software-Defined Networking
Congestion Control in SDN-Enabled Networks
A Hybrid IP Lookup Architecture with Fast Updates
2019/7/26 OpenFlow-Enabled User Traffic Profiling in Campus Software Defined Networks Presenter: Wei-Li,Wang Date: 2016/1/4 Author: Taimur Bakhshi and.
An Improved Wu-Manber Multiple Patterns Matching Algorithm
2019/9/14 The Deep Learning Vision for Heterogeneous Network Traffic Control Proposal, Challenges, and Future Perspective Author: Nei Kato, Zubair Md.
2019/9/3 Adaptive Hashing Based Multiple Variable Length Pattern Search Algorithm for Large Data Sets 比對 Simple Pattern 的方法是基於 Hash 並且可以比對不同長度的 Pattern。
A SRAM-based Architecture for Trie-based IP Lookup Using FPGA
2019/10/9 A Weighted ECMP Load Balancing Scheme for Data Centers Using P4 Switches Presenter:Hung-Yen Wang Authors:Jin-Li Ye, Yu-Huang Chu, Chien Chen.
MEET-IP Memory and Energy Efficient TCAM-based IP Lookup
2019/11/12 Efficient Measurement on Programmable Switches Using Probabilistic Recirculation Presenter:Hung-Yen Wang Authors:Ran Ben Basat, Xiaoqi Chen,
Presentation transcript:

2018/4/23 Dynamic Load-balanced Path Optimization in SDN-based Data Center Networks Author: Yuan-Liang Lan , Kuochen Wang and Yi-Huai Hsu Presenter: Yi-Hsien Wu Conference: 2016 10th International Symposium on Communication Systems, Networks and Digital Signal Processing Networking (Cloud net) Date: 2017/2/7 Department of Computer Science and Information Engineering National Cheng Kung University, Taiwan R.O.C. CSIE CIAL Lab 1

Outline Introduction Related Work Algorithm Performance Evaluation Conclusion National Cheng Kung University CSIE Computer & Internet Architecture Lab

Introduction DLPO (Dynamic Load-balanced Path Optimization) 2018/4/23 Introduction DLPO (Dynamic Load-balanced Path Optimization) It is an algorithm which can suit for different SDN-based data center network topologies . The proposed DLPO changes paths of flows during flow transmissions, achieves load balancing among different links, and efficiently resolves the network congestion problem in SDN-based data center networks. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

2018/4/23 Related Work SDN-based path optimization can be classified into static path optimization and dynamic path optimization. Static path optimization : means that the path of a flow cannot be changed during flow transmissions Dynamic path optimization : It can change the path of a flow during flow transmissions National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Related Work Static path optimization DFS worst-fit : 2018/4/23 Related Work Static path optimization Tosmate et al. proposed a depth-first search worst-fit multipath routing algorithm (DFS worst-fit) to find static paths for new-coming flows based on the DFS search algorithm . DFS worst-fit : It selects the one with the largest amount of remaining bandwidth. Its time complexity is O(1) by trying only the link with the largest amount of available bandwidth. It also has a faster speed to find a path than first-fit and best-fit by avoiding excessive backtracking. Since the available bandwidths of links in data center networks may change during flow transmission, static routing cannot efficiently balance link loads and may increase queuing delay in data center networks National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Related Work Dynamic path optimization 2018/4/23 Related Work Dynamic path optimization LABERIO (LoAd-BalancEd Routing wIth OpenFlow) is the first work to introduce an SDN-based dynamic path optimization algorithm to balance network traffic during flow transmissions. At the initialization step of the path selection for a new coming flow, LABERIO chooses a temporary path that the amount of available bandwidth of its bottleneck link is the maximum among all possible paths according to current available bandwidths of links in data center networks. The bottleneck link is defined as the link which has the minimal available bandwidth in a path. LABERIO is highly dependent on network topology and hence LABERIO must be modified if the network topology changes National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Related Work National Cheng Kung University CSIE Computer & Internet Architecture Lab

2018/4/23 Related Work National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Algorithm Path initialization stage 2018/4/23 Algorithm The proposed DLPO is composed of two stages, the path initialization stage and the dynamic path optimization stage. Path initialization stage Find a temporary path according to the available bandwidth of each path's bottleneck link. Among all possible paths between the source host and the destination host, the path whose bottleneck link has the largest available bandwidth will be chosen as a temporary path. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Algorithm Dynamic path optimization stage 2018/4/23 Algorithm Dynamic path optimization stage DLPO will change paths of flows during flow transmissions to balance link loads and resolve the congestion problem in data center networks. The proposed DLPO uses the OpenFlow protocol to retrieve load statistics from switches and detect the load-balance status. If link loads in the data center network are imbalanced, the path optimization algorithm will be triggered to balance link loads. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

2018/4/23 Algorithm We use the simple moving average (SMA) of the variance of link loads as our load-balance detection parameter, which is defined in Eq. (1), to compute the average value of the variance of link loads during a period of time to smooth out short-term fluctuations. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

2018/4/23 Algorithm Besides, we will derive a path optimization trigger threshold δ*. If δ(t) > δ*, the proposed DLPO will be triggered to balance link loads in the date center network. The value of δ* is network topology-dependent and can be determined by experiments under a fixed traffic scenario. The dynamic path optimization stage of the proposed DLPO is composed of two algorithms, which are multi-link DLPO algorithm and single-link DLPO algorithm. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

2018/4/23 Algorithm When the imbalance of link loads in a data center network is detected , the multi-link DLPO algorithm will be triggered first to balance link loads. It will find the top 10% busiest links in the data center network and then change the path of the flow which covers most of the top 10% busiest links to a path whose bottleneck link has the largest available bandwidth. If the link loads are still imbalanced after all the top 10% busiest links have already been handled by the multi-link DLPO algorithm, the single-link DLPO algorithm will be triggered to resolve the congested paths that the multi-link DLPO algorithm cannot handle. The single-link DLPO algorithm will reroute the largest flow on the busiest link to a path whose bottleneck link has the largest available bandwidth so as to resolve the congested path problem efficiently. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

2018/4/23 Algorithm Because flows in a data center network may vary at any time, it is not feasible to change paths of all flows at one time. Thus, the proposed DLPO will only handle one flow at a time and sleep for a sleeping period Ps after changing the path of the flow that is most likely to cause congestion. The proposed DLPO balances the link loads efficiently and resolves the congestion problem, the queuing delay of packets can be reduced and hence the throughput of flows can be improved as well. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Performance Evaluation 2018/4/23 Performance Evaluation Setup : National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Performance Evaluation 2018/4/23 Performance Evaluation We evaluate the performance of the proposed DLPO, LABERIO, and Floodlight in a fully populated topology, as shown in Figure 2. The traffic pattern for simulation is a hotspot traffic pattern , which is prone to cause congestion in certain links. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Performance Evaluation 2018/4/23 Performance Evaluation The relationship between average throughput per flow and the value of δ* is shown in Figure 3. We observed that throughput reaches the maximum value when δ* is set to be 20× 109 byte2. Therefore, we set the triggering threshold δ* to 20× 109 byte2 for the fully populated topology. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Performance Evaluation 2018/4/23 Performance Evaluation The average link bandwidth utilization of the proposed DLPO, LABERIO, and Floodlight are shown in Figure 4. It is obvious that although LABERIO reacts to imbalance quicker during the first 20 seconds, the proposed DLPO can eventually has higher overall bandwidth utilization. Simulation results show that the average bandwidth utilization of the proposed DLPO is 8.3% higher than that of LABERIO. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

Performance Evaluation 2018/4/23 Performance Evaluation Figure 5 shows that DLPO can increase 24.8% of average throughput per flow compared to LABERIO in the fully populated topology. A fully populated topology is very suitable for dynamic path optimization because it provides many alternative links between switches. National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab

2018/4/23 Conclusion We have presented a Dynamic Load balanced Path Optimization (DLPO) algorithm which is suited for different data center network topologies. The proposed DLPO may change paths of flows during flow transmissions, achieve load balancing among different links, and efficiently resolve the congestion problem in data center networks. 近期因為ruleset size快速成長,因此ruleset複雜度造成一般封包分類方法memory表現很差 Swintop是一種將ruleset去分類的方式 National Cheng Kung University CSIE Computer & Internet Architecture Lab CSIE CIAL Lab