1 / 1

Towards Quantification of IP Network Reliability

Towards Quantification of IP Network Reliability. Hao Wang, Alex Gerber, Albert Greenberg, Jia Wang, Yang R. Yang AT&T Labs Research Microsoft Research Yale University. Internet: A Critical Infrastructure. Model & Methodology. The Messages. VoIP Video Conferencing Online Trading

Download Presentation

Towards Quantification of IP Network Reliability

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Towards Quantification of IP Network Reliability Hao Wang, Alex Gerber, Albert Greenberg, Jia Wang, Yang R. Yang AT&T Labs Research Microsoft Research Yale University Internet: A Critical Infrastructure Model & Methodology The Messages • VoIP • Video Conferencing • Online Trading • Online gaming • E-Commerce • Evaluation of the effects of various factors on IP network reliability • Abilene has decent yet insufficient reliability • >70% higher than two 9’s • <60% higher than five 9’s • IP layer reliability techniques help • IGP re-convergence vs. the others • No single reliability techniques prevails • As failures last longer (high failure -> low failure) • IGP histogram improves • All others histograms degrades IP Network Reliability • Quantification of Reliability: # of Nine’s Future Work • IP network topology • Abilene router-level as of October, 2006 • Traffic demands • Abilene Oct. 4 2006 – Oct. 31, 2006 • Failure models • Generate synthetic failures • Parametric model • Mean time-to-failure (MTTF) • Mean time-to-repair (MTTR) • Failure ratio (severity) level (FRL) • Evaluations on more (typical) IP networks • Incorporating more practical factors • Edge router reliability • Interdomain (BGP) route oscillation • Questions around IP Network Reliability • Estimations vary from ~99% to ~99.999% • Methodology of estimations unclear • Effects of traffic engineering (TE) techniques on reliability • TE optimizes for network specific metrics • MLU, network costs, etc • Reliability perceived by end-users • Connectivity, loss rate, delay, etc • What is the effect of TE on reliability? • Availability vs. Reliability • Mere connectivity is not enough - performance matters • Low failure example: maintenance • High failure example: aging equipment Major Results Objective: Quantification of Reliability • In terms of aggregated statistics of O-D pair reliability level • Question #1: What is the reliability level of IP networks? • Defined using Service Level Agreement (SLA) • Connectivity • Delay • Loss rate • Reliability Level = Time SLA satisfied / Total time • How many nines can they achieve? • Question #2: How effective are the IP layer techniques in improving the reliability of IP networks? • IGP (CSPF) re-convergence • IGP (CSPF) fast rerouting • TE with fast rerouting • TE with CSPF fast rerouting • Do they make a difference?

More Related