The Deepest Vertex First Reboot: Rebooting Network Edge Switches in a Campus Network

Abstract

Network edge switches require to be rebooted for some reasons, e.g., updating a firmware, configuring a special behavior and so on. This paper proposes the deepest vertex first reboot that can reboot many network switches with less overhead downtime. The deepest vertex first reboot can reduce downtime overhead by vertex contraction. The deepest vertex first reboot then enables to reboot all switches by only 13-s overhead where the under-estimated ideal minimum downtime is 75 s in an actual campus network where there are more than 300 network equipment.