Failure recovery using vlan-tag in SDN: High speed with low memory requirement

To achieve high availability, fault tolerance is desirable for most networks, including Software Defined Networks (SDNs). Link and switch failures are commonly observed in SDN, severely affecting the network performance and even making the whole system break down. In this paper, we propose a protection-based recovering method for link failures using vlan-tag, which is fast with low memory consumption. Inheriting from protection mechanism, our method is able to react and recover quickly without the participation of controller(s) by immediately switching to the backup path(s) precomputed when failure occurs. More importantly, it is memory efficient as it calculates the backup based on each link (instead of each flow), and simply assigns a vlan-id for each link, avoiding the high memory (TCAM used for flow table storage is costly) requirements of conventional protection methods. The simulation results expose that our method can recover from link failures within 20 ms (on the average) like other protection approaches, while the restoration needs 50 ms (on the average). Moreover, in experimental environment our method requires 100 times less flow table entries' amount than the protection method based on each flow to store backup paths.

[1]  Nick McKeown,et al.  OpenFlow: enabling innovation in campus networks , 2008, CCRV.

[2]  Piotr Cholda,et al.  Network Recovery, Protection and Restoration of Optical, SONET-SDH, IP, and MPLS [Book Review] , 2005, IEEE Communications Magazine.

[3]  Christian Esteve Rothenberg,et al.  SlickFlow: Resilient source routing in Data Center Networks unlocked by OpenFlow , 2013, 38th Annual IEEE Conference on Local Computer Networks.

[4]  Min Zhu,et al.  B4: experience with a globally-deployed software defined wan , 2013, SIGCOMM.

[5]  Wei Zhang,et al.  When Software Defined Networks Meet Fault Tolerance: A Survey , 2015, ICA3PP.

[6]  Vyas Sekar,et al.  The middlebox manifesto: enabling innovation in middlebox deployment , 2011, HotNets-X.

[7]  Fernando M. V. Ramos,et al.  Software-Defined Networking: A Comprehensive Survey , 2014, Proceedings of the IEEE.

[8]  Sujata Banerjee,et al.  DevoFlow: scaling flow management for high-performance networks , 2011, SIGCOMM.

[9]  Thyaga Nandagopal,et al.  Coping with link failures in centralized control plane architectures , 2010, 2010 Second International Conference on COMmunication Systems and NETworks (COMSNETS 2010).

[10]  Michael J. Freedman,et al.  Ravana: controller fault-tolerance in software-defined networking , 2015, SOSR.

[11]  Didier Colle,et al.  OpenFlow: Meeting carrier-grade recovery requirements , 2013, Comput. Commun..

[12]  Piero Castoldi,et al.  OpenFlow-based segment protection in Ethernet networks , 2013, IEEE/OSA Journal of Optical Communications and Networking.

[13]  Piet Demeester,et al.  Network Recovery: Protection and Restoration of Optical, SONET-SDH, IP, and MPLS , 2004 .

[14]  Nick Feamster,et al.  CORONET: Fault tolerance for Software Defined Networks , 2012, 2012 20th IEEE International Conference on Network Protocols (ICNP).

[15]  Deborah Brungard,et al.  Requirements of an MPLS Transport Profile , 2009, RFC.

[16]  Marco Canini,et al.  Automatic failure recovery for software-defined networks , 2013, HotSDN '13.

[17]  Fernando M. V. Ramos,et al.  On the Feasibility of a Consistent and Fault-Tolerant Data Store for SDNs , 2013, 2013 Second European Workshop on Software Defined Networks.

[18]  Weifa Liang,et al.  Cost Minimization for Rule Caching in Software Defined Networking , 2016, IEEE Transactions on Parallel and Distributed Systems.

[19]  Marco Canini,et al.  FatTire: declarative fault tolerance for software-defined networks , 2013, HotSDN '13.

[20]  Fernando M. V. Ramos,et al.  On the Design of Practical Fault-Tolerant SDN Controllers , 2014, 2014 Third European Workshop on Software Defined Networks.