Upload
sadi
View
13
Download
0
Embed Size (px)
DESCRIPTION
Load Balancing in SDN Networksusing Metaheuristic Methods
Citation preview
GYTE - Bilgisayar Mühendisliği Bölümü
Bilgisayar Mühendisliği Bölümü
GYTE - Bilgisayar Mühendisliği Bölümü
Load Balancing in SDN Networks
using Metaheuristic Methods
CSE624
Mahmud Rasih ÇELENLİOĞLU
December 2013
GYTE - Bilgisayar Mühendisliği Bölümü
Outline
• Overview of Software Defined Networking
• Related Works
• LB-VNS
• LB-GA
• Conclusion
2
GYTE - Bilgisayar Mühendisliği Bölümü
SDN Architecture
3
Control Plane:
• Logically centralized
• Smart
• Slow
Data Plane:
• Dummy
• Fast
Protocol for Controller – Switch Communication
(e.g. OpenFlow Protocol)
GYTE - Bilgisayar Mühendisliği Bölümü
Software-Defined Networking
• Traffic Engineering
• Network Virtualization
• Virtual Private Networks
Scalability
• Signalling
• Computational Complexity
• Switch Configuration
Reliability
• Controller
4
GYTE - Bilgisayar Mühendisliği Bölümü
Related Works – MPLS
5
GYTE - Bilgisayar Mühendisliği Bölümü
Virtual Path Approach
6
• Create virtual links
• Capacity of VL is proportional to pyhsical link
capacity
• Local Solution
GYTE - Bilgisayar Mühendisliği Bölümü
Objective Function
7
• Max load balancing with min path change
• Path Change Ratio (PCR): How many flows are
shifted to other paths
• Avg Distance Ratio (ADR): Avg distance btw current
loads and ideal loads for all paths
GYTE - Bilgisayar Mühendisliği Bölümü
Details
8
• Distribute N flows to M paths
• Solution Representation
8 20 12 16 32 14 10 8 8 24
2 2 1 3 1 2 1 1 3 3
2 2 2 3 3 2 2 1 1 2
GYTE - Bilgisayar Mühendisliği Bölümü
LB-VNS
9
Neighborhood Structures:
– Smart Insert: Randomly chosen flow is assigned to the path whose
distance is maximum
– Random Insert: Randomly chosen flow is assigned to a randomly
chosen path
– Smart Swap: Randomly chosen two flows are swapped between two
paths with max distances. Distance of one path exceeds the ideal load
and the another one does not
– Random Swap: Randomly chosen two flows are assigned to two
randomly chosen paths
– Fully Random Assignment: All flows are randomly assigned to paths
without any consideration
Local Search: Smart Insert, Smart Swap
Shaking: Fully Random Assignment
GYTE - Bilgisayar Mühendisliği Bölümü
LB-VNS
10
GYTE - Bilgisayar Mühendisliği Bölümü
GA
11
• Population Size: 20
• Tournament Selection:
– First Tournament:
1. Select F number of individuals
2. Pick best among them
– SecondTournament:
1. Discard individuals attended to first tournament
2. Select S number of individuals & Pick best among them
• F = Population Size/3 & S = (Population Size - F)/3path
GYTE - Bilgisayar Mühendisliği Bölümü
GA
12
Crossover:
• Dominant Gene: Flows Size < Th-Min OR Flows Size > Th-Max
• Recessive Gene: Th-Min <= Flows Size <= Th-Max
• Dominant Gene: Stay in the same path with probability of 0.9 otherwise
inherited from one of parents
• Recessive Gene: Always inherit assignment of one of parents
Mutation:
• Shift a flow belongs to overloaded path to underloaded
• Perform after crossover
GYTE - Bilgisayar Mühendisliği Bölümü
LB-GA
13
GYTE - Bilgisayar Mühendisliği Bölümü
LB-GA Experiment
14
• 10 run for each path number [2,5]
• 1000 iteration
GYTE - Bilgisayar Mühendisliği Bölümü
LB-GA Experiment
15
GYTE - Bilgisayar Mühendisliği Bölümü
Conclusion
16
• LB-GA works better than LB-VNS
• LB-VNS is faster
• Performance of LB-VNS can be increased by changing local
search method
GYTE - Bilgisayar Mühendisliği Bölümü
Questions?
Thank You
&
Questions?
17
GYTE - Bilgisayar Mühendisliği Bölümü
Resources
18
1. N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar, L. Peterson, J. Rexford, S. Shenker, and J. Turner, OpenFlow:
Enabling innovation in campus networks, ACM Computer Communication Review, Apr. 2008.
2. "Software-defined networking: The new norm for networks," in ONF White Paper. Open Networking Foundation, 2012.
[Online]. Available: https://www.opennetworking.org/images/stories/downloads/white-papers/wp-sdn-newnorm.pdf
3. "Open networking foundation," https://www.opennetworking.org/, March 2011.
4. Elwalid, A.; Cheng Jin; Low, S.; Widjaja, I., "MATE: MPLS adaptive traffic engineering," INFOCOM 2001. Twentieth Annual
Joint Conference of the IEEE Computer and Communications Societies. Proceedings. IEEE , vol.3, no., pp.1300,1309 vol.3,
2001.
5. Aukia, P.; Kodialam, M.; Koppol, P.V.N.; Lakshman, T. V.; Sarin, H.; Suter, B., "RATES: a server for MPLS traffic engineering,"
Network, IEEE , vol.14, no.2, pp.34,41, Mar/Apr 2000.
6. Haci A. Mantar, "An Admission Control and Traffic Engineering Model for DiffServ-MPLS Networks", Lecture notes in computer
science (LNCS) 4238, pp. 152–161, 2006.
7. Haci A. Mantar, "A scalable QoS routing model over MPLS Networks", Journal of Telecommunication Systems, Vol.34, No. 4,
2007.
8. Suleyman Goger, Haci Ali Mantar, "An SDN-based Energy-Aware Resource Management Management Model for Wide Area
Networks", to appear in Elsevier Computer Networks, 2013.
9. S.Burak Göger, Elif Öztürk, Hacı Ali Mantar, "GMNet: An SDN Based Energy -Aware Routing Model for MPLS Networks",
ISITES 2013, Sakarya, Turkiye.
10. Haci Mantar, Junseok Hwang, Steve Chapin, Ibrahim Okumus ”A Scalable Model for Inter-Bandwidth Broker Resource
Reservation and Provisioning”, IEEE Journal on Selected Communications (JSAC), special issue implementation and analysis of
communication protocol, No.10, December 2004.
11. Nikhil Handigol, Srini Seetharaman, Mario Flajslik, Nick McKeown, Ramesh Johari, Plug-n-Serve: Load-Balancing Web
Traffic using OpenFlow, ACM SIGCOMM Demo, Aug 2009.
12. Yannan Hu; Wendong Wang; Xiangyang Gong; Xirong Que; Shiduan Cheng, "BalanceFlow: Controller load balancing for
OpenFlow networks," Cloud Computing and Intelligent Systems (CCIS), 2012 IEEE 2nd International Conference on , vol.02,
no., pp.780,785, Oct. 30 2012-Nov. 1 2012.
Mohammad Al-Fares, Sivasankar Radhakrishnan, Barath Raghavan, Nelson Huang, and Amin Vahdat. 2010. "Hedera: dynamic
flow scheduling for data center networks." In Proceedings of the 7th USENIX conference on Networked systems design and
implementation (NSDI'10). USENIX Association, Berkeley, CA, USA, 19-19.
Minlan Yu, Jennifer Rexford, Michael J. Freedman, and Jia Wang Scalable flow-based networking with DIFANE, ACM
SIGCOMM, New Delhi, India, August 2010.
Richard Wang, Dana Butnariu, and Jennifer Rexford OpenFlow-based server load balancing gone wild, Workshop on Hot Topics
in Management of Internet, Cloud, and Enterprise Networks and Services (Hot-ICE), Boston, MA, March 2011.
H.E. Egilmez, B. Gorkemli, A.M. Tekalp, S. Civanlar. Scalable Video Streaming Over OpenFlow Networks: An Optimization
Framework For QoS Routing, Proc. IEEE International Conference on Image Processing (ICIP 2011), Brussels, Belgium,
September 2011.
A. Tootoonchian and Y. Ganjali, "Hyperflow: A distributed control plane for openflow," in Proceedings of the 2010 internet network
management conference on Research on enterprise networking. USENIX Association, 2010, pp. 3–3.
Othman, M.M.O.; Okamura, K., "Enhancing Control Model to Ease Off Centralized Control of Flow-Based SDNs," Computer
Software and Applications Conference (COMPSAC), 2013 IEEE 37th Annual , vol., no., pp.467,470, 22-26 July 2013.
GYTE - Bilgisayar Mühendisliği Bölümü
Resources
19
13. Mohammad Al-Fares, Sivasankar Radhakrishnan, Barath Raghavan, Nelson Huang, and Amin Vahdat. 2010. "Hedera:
dynamic flow scheduling for data center networks." In Proceedings of the 7th USENIX conference on Networked systems design
and implementation (NSDI'10). USENIX Association, Berkeley, CA, USA, 19-19.
14. Minlan Yu, Jennifer Rexford, Michael J. Freedman, and Jia Wang Scalable flow-based networking with DIFANE, ACM
SIGCOMM, New Delhi, India, August 2010.
15. Richard Wang, Dana Butnariu, and Jennifer Rexford OpenFlow-based server load balancing gone wild, Workshop on Hot
Topics in Management of Internet, Cloud, and Enterprise Networks and Services (Hot-ICE), Boston, MA, March 2011.
16. H.E. Egilmez, B. Gorkemli, A.M. Tekalp, S. Civanlar. Scalable Video Streaming Over OpenFlow Networks: An Optimization
Framework For QoS Routing, Proc. IEEE International Conference on Image Processing (ICIP 2011), Brussels, Belgium,
September 2011.
17. A. Tootoonchian and Y. Ganjali, "Hyperflow: A distributed control plane for openflow," in Proceedings of the 2010 internet
network management conference on Research on enterprise networking. USENIX Association, 2010, pp. 3–3.
18. Othman, M.M.O.; Okamura, K., "Enhancing Control Model to Ease Off Centralized Control of Flow-Based SDNs," Computer
Software and Applications Conference (COMPSAC), 2013 IEEE 37th Annual , vol., no., pp.467,470, 22-26 July 2013.
19. H.E. Egilmez, S.T. Dane, K.T. Bagci, A.M. Tekalp, “OpenQoS: An OpenFlow Controller Design for Multimedia Delivery with
End-to-End Quality of Service over Software-Defined Networks” Proc. of APSIPA Annual Summit and Conference (APSIPA ASC
2012), Los Angeles, USA, December 2012.
20. 16] Chiaraviglio, L.; Mellia, M.; Neri, F.; "Reducing Power Consumption in Backbone Networks," Communications, 2009.
1001BF-01 Güncelleme Tarihi: 24/01/2012 ICC '09. IEEE International Conference on , vol., no., pp.1-6, 14-18 June 2009
21. Levy, Steven, "Going With the Flow: Google’s Secret Switch to the Next Wave of Networking", Wired, April 17, 2012.
Retrieved 2012-04-17.