60 likes | 190 Views
This document outlines the requirements and solutions for TESI segment protection, focusing on end-to-end support and the scaling challenges encountered when multiple Technical Ethernet Service Instances (TESIs) share the same links. It discusses the integration of CCM for connectivity state monitoring and proposes a cycle-based notification mechanism to optimize fault reporting. The approach maintains compliance with existing standards, offering a scalable solution for segment protection while ensuring reliability. Additionally, it emphasizes the need for simplified mechanisms to address connectivity state effectively.
E N D
Segment Protection Scaling Bob Sultan (bsultan@huawei.com) Deng Zhusheng
TESI Segment Protection Requirement F B J A D H L C G K • PBB-TE supports end-to-end TESI protection • Requirement to protect a TESI ‘segment’ • See Vinod Kumar Service Provider Requirements
TESI Segment Protection F B J A D H L C G K • Segment endpoints are MEPs • CCM detects connectivity state of W/P • Switch achieved by FDB update at endpoints • Simple scheme based on existing 802.1ag • End-to-end TESI protection/CCM as usual
Segment Protection Scaling Issue F B J A D H L C G K • Many TESIs can use same set of links for W/P • Requires W/P CCMs per TESI segment • Can be scaling problem in some cases
Scaling Segment Protection with ‘Cycle’ X cycle F B J A D H L C G K • Observe that W+P segments always form a cycle • Observe that many paths may use the same cycle • Replace CCM per segment (many) with one fault notification per cycle • Please note: this use of the word ‘cycle’ has absolutely nothing to do with ‘p-cycles’. A cycle is just a closed loop of links.
Summary • We support CCM-based segment protection • We propose cycle-based notification for scaling purposes for cases in which many segments use the same cycle (cycle is provisioned) • Notification mechanism could re-use G.8032 or ring/cycle mechanism developed in 802.1 (see Norm Finn proposal).