Dissecting Video Server Selection Strategies in the YouTube CDN
Top Cited Papers
- 1 June 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 248-257
- https://doi.org/10.1109/icdcs.2011.43
Abstract
In this paper, we conduct a detailed study of the YouTube CDN with a view to understanding the mechanisms and policies used to determine which data centers users download video from. Our analysis is conducted using week-long datasets simultaneously collected from the edge of five networks - two university campuses and three ISP networks - located in three different countries. We employ state-of-the-art delay-based geolocation techniques to find the geographical location of YouTube servers. A unique aspect of our work is that we perform our analysis on groups of related YouTube flows. This enables us to infer key aspects of the system design that would be difficult to glean by considering individual flows in isolation. Our results reveal that while the RTT between users and data centers plays a role in the video server selection process, a variety of other factors may influence this selection including load-balancing, diurnal effects, variations across DNS servers within a network, limited availability of rarely accessed video, and the need to alleviate hot-spots that may arise due to popular video content.Keywords
This publication has 10 references indexed in Scilit:
- Understanding data-center driven content distributionPublished by Association for Computing Machinery (ACM) ,2010
- YouTube traffic dynamics and its interplay with a tier-1 ISPPublished by Association for Computing Machinery (ACM) ,2010
- Internet inter-domain trafficPublished by Association for Computing Machinery (ACM) ,2010
- Challenging statistical classification for operational usagePublished by Association for Computing Machinery (ACM) ,2009
- Characteristics of YouTube network traffic at a campus network – Measurements, models, and implicationsComputer Networks, 2009
- Analyzing video services in Web 2.0Published by Association for Computing Machinery (ACM) ,2008
- I tube, you tube, everybody tubesPublished by Association for Computing Machinery (ACM) ,2007
- Constraint-Based Geolocation of Internet HostsIEEE/ACM Transactions on Networking, 2006
- Measuring IP and TCP behavior on edge nodes with TstatComputer Networks, 2005
- Scale-free behavior of the Internet global performanceZeitschrift für Physik B Condensed Matter, 2003