Abstract

Programmable data plane hardware creates a possibility to solve network-related problems. Ensuring fault tolerance of link failures is a fundamental network issue. Link failure recovery mechanisms are widely used in traditional and software-defined networks. The proactive failure recovery mechanism usually requires a backup path to be installed in the switch in advance. When the link fails, the switch can quickly switch to the backup path to continue sending packets. However, storing a large number of backup paths consumes a lot of switch storage. In this article, we analyze why implementing traditional proactive failure recovery mechanism introduces huge switch storage overhead, and discuss the flexibility and limitations of the programmable data plane. Then, we present P4Neighbor, a proactive link failure recovery framework based on the programmable data plane. P4Neighbor encapsulates backup paths into the header of a packet when a link failed and leverages this information to achieve link failure recovery. By storing only the backup paths of the neighbor switches, P4Neighbor requires little switch storage to store backup paths. Besides, P4Neighbor also takes complex link failure situations into consideration, which makes the network's fault tolerance slightly increase. Experimental results show that compared with the traditional failure recovery mechanism, P4Neighbor achieves a reduction rate of 57.9%-84.5% in terms of stored switch entries. Meanwhile, P4Neighbor also has a higher failure recovery ratio than traditional proactive link failure recovery mechanisms.

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.