An Effective Recovery Mechanism for Fault Tolerant MPLS
碩士 === 輔仁大學 === 資訊工程學系 === 93 === Multi-Protocol Label Switching(MPLS) is developed network standard by IETF.It's the popular technology to come true Internet of wide band;It's purpose to provide more flexable, a highly scalable and high-performance IP forwarding technology.However,Any net...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2005
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/50510485219776839255 |
| Summary: | 碩士 === 輔仁大學 === 資訊工程學系 === 93 === Multi-Protocol Label Switching(MPLS) is developed network standard by IETF.It's the popular technology to come true Internet of wide band;It's purpose to provide more flexable, a highly scalable and high-performance IP forwarding technology.However,Any network technology always face one question that it do routing optimized spontaneously and keep on work when network break down.Now some specialists have provided “Traffic Engineering” concept to solve its.Traffic Engineering talk about Fault-Tolerance subject so I want to search it in my thesis.We study many related RFC drafts and papers about MPLS, and analyze the difference between different recovery mechanisms.Some methods only consider to reroute backup LSP presently after one LSP occurs error.We discuss issues being occurred in actual network operation. Issue involve: causes network congestion seriously due to select the bad path from "protection switch method" and causes long service interrupted time if using "dynamic rerouting method".For our study, we integrate the advantages of existing recovery mechanisms to propose a new efficient recovery mechanisms.Using "shortest augmenting path algorithm" calculate all no-bandwidth-reserved paths in LSRs of work path beforce fail for keeping work soon.Upstream LSR after fail use FIS message to inform ingress LSR.Ingress LSR use "Dijkstra algorithm" depend on used bandwidth at present to find a best、shortest path after receiving FIS message.We use NS2 network simulator to evaluate these performance.Finally according to simulations, we evaluate our method to have better performances than the other methods.
|
|---|