Experimenting with ONOS Scalability on Software Defined Network

Experimenting with ONOS Scalability on Software Defined Network By: Mohammed Sameer & Dr. Bhargavi Goswami Organization: Christ University sameersam 9519@gmail. com bhargavihgowsami@gmail. com

Experimenting with ONOS Outline Experimenting with Scalability on Software Defined of Floodlight Controller in Network » Limitations of Current Networks » What is SDN? ONOS SDN Controller » SIMULATION ENVIRONMENT » OBTAINED RESULTS » Conclusion 2

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network Limitations of Current Networks » Lack of programmability and innovation » Enterprise networks are difficult to manage » “New control requirements have arisen”: » How to easily configure huge networks? 3 http: //www. excitingip. net/27/a-basic-enterprise-lan-network-architecture-block-diagram-and-components/

What is SDN Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network Traditional vs SDN 4 Open. Flow/SDN tutorial, Srini Seetharaman, Deutsche Telekom, Silicon Valley Innovation Center

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network ONOS SDN Controller ONOS » Open source » Java based » web GUI ONOS SDN Controller Architecture 5

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network ONOS SDN Controller » Limitations of Onos Controller Deploying on a VM • Lack Of System Resources • Long Waiting time to start a Controller and Stop How to Overcome this ? Solution: Docker 6

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network ONOS Docker Container 7

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network SIMULATION ENVIRONMENT from mininet. topo import Topo class My. Topo( Topo ): "Simple topology example. " def __init__( self ): "Create custom topo. " # Initialize topology Topo. __init__( self ) #Add switches for s 11 Switch in range(1): s 1 Switch = self. add. Switch( 's 1' ) for s 22 witch in range(1): s 2 Switch = self. add. Switch( 's 2' ) self. add. Link( s 1 Switch, s 2 Switch ) for s 33 witch in range(1): s 3 Switch = self. add. Switch( 's 3' ) self. add. Link( s 2 Switch, s 3 Switch ) self. add. Link( s 1 Switch, s 3 Switch ) for s 44 witch in range(1): s 4 Switch = self. add. Switch( 's 4' ) self. add. Link( s 1 Switch, s 4 Switch ) self. add. Link( s 2 Switch, s 4 Switch ) self. add. Link( s 3 Switch, s 4 Switch ) for s 55 witch in range(1): s 5 Switch = self. add. Switch( 's 5' ) self. add. Link( s 1 Switch, s 5 Switch ) self. add. Link( s 2 Switch, s 5 Switch ) self. add. Link( s 3 Switch, s 5 Switch ) self. add. Link( s 4 Switch, s 5 Switch ) Connectivity between nodes in the scenario for s 66 witch in range(1): s 6 Switch = self. add. Switch( 's 6' ) self. add. Link( s 1 Switch, s 6 Switch ) self. add. Link( s 2 Switch, s 6 Switch ) self. add. Link( s 3 Switch, s 6 Switch ) self. add. Link( s 4 Switch, s 6 Switch ) self. add. Link( s 5 Switch, s 6 Switch ) # Add hosts and link to switches for h 1_ in range(0, 30): h 1=self. add. Host('h 1_%s' % (h 1_+1)) self. add. Link( h 1, s 1 Switch ) for h 2_ in range(30, 60): h 2=self. add. Host('h 2_%s' % (h 2_+1)) self. add. Link( s 2 Switch, h 2 ) for h 3_ in range(60, 90): h 3=self. add. Host('h 3_%s' % (h 3_+1)) self. add. Link( s 3 Switch, h 3 ) #for h 4_ in range(30, 40): #h 4=self. add. Host('h 4_%s' % (h 4_+1)) #self. add. Link( s 4 Switch, h 4 ) for h 5_ in range(90, 120): h 5=self. add. Host('h 5_%s' % (h 5_+1)) self. add. Link( s 5 Switch, h 5 ) for h 6_ in range(120, 150): h 6=self. add. Host('h 6_%s' % (h 6_+1)) self. add. Link( s 6 Switch, h 6 ) topos = { 'mytopo': ( lambda: My. Topo() ) } Source code written with Mininet sudo docker run -i -d --name myonos -t sdnhub/onos 8

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network Throughput for scenarios with 100 nodes. Result: Throughput for scenarios with 200 nodes. 9 Throughput for scenarios with 300 nodes.

RYU Controller’s Scalability Experimenting with Scalability Experiment on of. Software Floodlight Controller Defined Networksin Result: Throughput for scenarios with 400 nodes. Throughput for scenarios with 500 nodes. 10 Average Throughput

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network Conclusion With this paper, the authors make attempt to present experimentation with scalability of ONOS Controller in varied networking circumstances. Keeping the focus on throughput, the authors have provided complete details of the experiments executed with minor details of experimental test bed setup, execution of experiment in multiple scenarios commands list. We conclude the paper by providing green signal to the researchers who are working on ONOS controller and wish to move ahead with their research in the domain of ONOS and SDN. The controller behaves stable while varied number of nodes and is able to handle the network blast to some extent, which is observed and discussed during performance analysis. With this paper, authors have opened the way to new bees who would like to explore more on the ONOS controller as the paper includes clear step by step implementation with clear set of commands to explore multiple scenarios on ONOS controller. Authors have provided the implementation structure along with the solutions to the issues faced during the experimentations related to hardware, platform and software. With this paper, authors throw light on the ONOS controller which is better in comparison of existing ones by providing comparison table of controllers. The authors also address programmers, researchers and network administrators working on L 2, L 3 and L 4 layer through this paper who wants to practically explore and experiment further on ONOS Controller in SDN. Further, researchers are experimenting with more such controllers to come with more practical research based experiments in the area of SDN. 11

Experimenting with ONOS Experimenting with Scalability on Software Defined of Floodlight Controller in Network REFERENCES 1) Saleh Asadollahi, Bhargavi Goswami, Mohammed Sameer, (2018). “Ryu controller's scalability experiment on software defined networks ” International Conference on “Current Trends in Advanced Computing”(ICCTAC)` , June 2018. Bangalore, India. IEEE Xplore ISBN: 978 -1 -5386 -4322 -8 2) Gowsami, B. Asadollahi, S. , (2016). “Novel Approach to Improvise Congestion Control over Vehicular Ad Hoc Networks (VANET)” Proceedings of the 10 th INDIACom; International Conference on “Computing for Sustainable Global Development”, March 2016. Delhi, India. IEEE Xplore ISBN: 978 -9 -3805 -4421 -2 3) Asadollahi, S. , Gowsami, B. (2017). Revolution in Existing Network under the Influence of Software Defined Network. Proceedings of the INDIACom 11 th, Delhi, March 1 -3. 2017 IEEE Conference ID: 40353 4) Mc. Keown et al, N. (2008). Open. Flow: Enabling innovation in campus networks. ACM SIGCOMM - Computer Communication Revie, vol. 38, no. 2, p. 69– 74. 5) Smith et al, M. (2014). Op. Flex control protocol, Internet Engineering Task Force, from : http: //tools. ietf. org/html/draft-smith-opflex-00 6) Doria et al, A. (2010). Forwarding and control element separation (For. CES) protocol specification. Internet Engineering Task Forc, from http: //www. ietf. org/r/fc/rfc 5810. txt. 7) Enns, R. , Bjorklund, M. , Schoenwaelder, J. , Bierman, A. (2011). Network configuration protocol (NETCONF). Internet Engineering Task Forc, form http: //www. ietf. org/rfc 6241. txt 8) Song, H. (2013). Protocol-oblivious forwarding: Unleash the power of SDN through a future-proof forwarding plane. Proceedings of ACM SIGCOMM Workshop Hot Topics Softw Defined Netw II. p. 127– 132. 9) Asadollahi, S. , Gowsami, B. (2017). Implementation of SDN using Open. Daylight Controller. Proceeding of An International Conference on Recent Trends in IT Innovations - Tec'afe 2017. ISSN(Online) : 2320 -9801 10) Mc. Cauley, M. (2012). POX, from http: //www. noxrepo. org/ 11) Erickson, D. (2013). The Beacon Open. Flow controller. Proceedings of ACM SIGCOMM Workshop Hot Topocs Software Defined Network II, 13 -18 p, 2013. 12) Gude al, N. (2008). NOX: Towards an operating system for networks. ACM SIGCOMM - Computer Communication Revie. vol. 38, no. 3, pp. 105– 110. 13) Nippon Telegraph and Telephone Corporation, RYU network operating system, 2012, from http: //osrg. github. com/ryu 14) Project Floodlight, Floodlight. (2012). from http: //floodlight. openflowhub. org/ 15) Asadollahi, S. , Gowsami, B. (2017). Software Defined Network, Controller Comparison. Proceedings of Tec'afe 2017, Vol. 5, Special Issue 2, April 2017. ISSN: 2320 -9798. 16) OFNet, from http: //sdninsights. org/ 17) Lantz, B. Heller, and N. Mc. Keown. (2010). A network in a laptop: Rapid prototyping for software-defined network. Proceedings of ACM SIGCOMM Workshop Hot Topics Netw, 19 th. p. 19: 1– 19: 6 18) B. Pfaff and B. Davie. (2013). The Open v. Switch database management protocol. Internet Engineering Task Force, RFC 7047, from http: //www. ietf. org/rfc 7047. txt 19) Gowsami, B. Asadollahi, S. , “Performance evaluation of widely implemented congestion control algorithms over diversified networking situations” Proceedings of International Conference on Computer Science Networks and Information Technology Held on 23 rd-24 th January 2016, Pattaya, Thailand. 12

Thanks! 13