Vertex-centric computation of service function chains in multi-domain networks

Computation of service function chains in multi-domain networks is critical for enabling the wide deployment of network function visualization for providing end-to-end network services. In this paper, we present a vertex-centric distributed orchestration framework for multi-domain networks, in which physical infrastructure information is maintained locally within each domain without infrastructure information sharing between domains, so that the management and control of multi-domain networks can be significantly simplified. We also propose a distributed computing algorithm for mapping a service function chain request in multi-domain networks. Our objective is to find all feasible mappings, from which we can further prune to obtain the optimal solution satisfying different constraints and policies. We implement the proposed algorithm in an open source vertex-centric distributed computing system and simulate two practical network topologies to evaluate the performance, including signaling delay, message overhead, and computation time. Simulation results demonstrate superior efficiency and scalability of the proposed algorithm.

[1]  Raouf Boutaba,et al.  PolyViNE: policy-based virtual network embedding across multiple domains , 2010, VISA '10.

[2]  Sangjin Hong,et al.  Virtual optical network embedding in multi-domain optical networks , 2014, 2014 IEEE Global Communications Conference.

[3]  Aart J. C. Bik,et al.  Pregel: a system for large-scale graph processing , 2010, SIGMOD Conference.

[4]  Riccardo Trivisonno,et al.  Recursive, hierarchical embedding of virtual infrastructure in multi-domain substrates , 2015, Proceedings of the 2015 1st IEEE Conference on Network Softwarization (NetSoft).

[5]  Mario Vento,et al.  An Improved Algorithm for Matching Large Graphs , 2001 .

[6]  Jennifer Widom,et al.  Optimizing Graph Algorithms on Pregel-like Systems , 2014, Proc. VLDB Endow..

[7]  David Dietrich,et al.  Multi-domain virtual network embedding with limited information disclosure , 2013, 2013 IFIP Networking Conference.

[8]  Mark D. Feuer,et al.  Integrated Restoration for Next-Generation IP-Over-Optical Networks , 2011 .

[9]  Holger Karl,et al.  Specifying and placing chains of virtual network functions , 2014, 2014 IEEE 3rd International Conference on Cloud Networking (CloudNet).

[10]  Jeffrey S. Chase,et al.  Embedding virtual topologies in networked clouds , 2011, CFI.

[11]  Jorge Lobo,et al.  Experimental results on the use of genetic algorithms for scaling virtualized network functions , 2015, 2015 IEEE Conference on Network Function Virtualization and Software Defined Network (NFV-SDN).

[12]  Rob Sherwood,et al.  The controller placement problem , 2012, HotSDN@SIGCOMM.

[13]  Seung-Ik Lee,et al.  Resource Management in Service Chaining , 2016 .

[14]  Djamal Zeghlache,et al.  Virtual network provisioning across multiple substrate networks , 2011, Comput. Networks.

[15]  Qiong Zhang,et al.  Effective virtual optical network embedding based on topology aggregation in multi-domain optical networks , 2014, OFC 2014.