1 / 7

Impact of Layering and Faults on Availability and its End-to-End Implications

This study examines the effects of layering and faults on system availability and end-to-end implications, focusing on response time, fault handling, and end-to-end checks. It explores how different layers impact system performance, fault tolerance, and overall reliability in computer systems.

lisav
Download Presentation

Impact of Layering and Faults on Availability and its End-to-End Implications

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. Impact of Layering and Faults on Availabilityand its End-to-End Implications Ricardo Bianchini, Richard Martin, Thu Nguyen Department of Computer Science Rutgers University Computer Science, Rutgers University

  2. Application Language VM OS VM Virtual FS FS Virtual Disk SCSI IDE Layers on Layers on Layers Operating System Socket TCP IP Enet WiFi Computer Science, Rutgers University

  3. Clients IP Network Load Manager Web Server Web Server Web Server DataBase DataBase Global Layering Tier 1 Tier 2 Tier 3 Computer Science, Rutgers University

  4. Availability • Correct response within accepted time bound • Fraction correct/90th percentile response time • Two components: • Correctness • Time bound • To get to 99.999 nines (5 min unavailability/year) • Not a lot of time to mess around Computer Science, Rutgers University

  5. Clients IP Network Load Manager Web Server Web Server Web Server DataBase DataBase 2 Second Reponse time Tier 1 Tier 2 Tier 3 Computer Science, Rutgers University

  6. Application Language VM OS VM Virtual FS FS Virtual Disk SCSI Impact of Faults and Layers • Each layer built independently • Intermediate layers hide exceptional conditions • Buffer and continue • Retry N • Crashes • Punt • Hard to build availabile systems • React quickly enough? • Diagnosis difficult • Prediction try { do_stuff(); }catch(e) { freak_out(); } Fault Computer Science, Rutgers University

  7. End-to-End Implications • Traditional Philosophy • End-to-end checks necessary for correctness, • Intermediate checks only a performance enhancement • Availability oriented: • End-to-end checks necessary for correctness • Intermediate checks and timely cross-layer propagation of information necessary • How to achieve these properties while maintaining layering? Computer Science, Rutgers University

More Related