Mohammad Al-Fares
ABSTRACT
Fast-loading web pages are key for a positive user experience. Unfortunately, a large number of users suffer from page load times of many seconds, especially for pages with many embedded objects. Most of this time is spent fetching the page and its objects over the Internet.
This paper investigates the impact of optimizations that improve the delivery of content from edge servers at the Yahoo! Content Delivery Network (CDN) to the end users. To this end, we analyze packet traces of 12.3M TCP connections originating from users across the world and terminating at the Yahoo! CDN. Using these traces, we characterize key user-connection metrics at the network, transport, and the application layers. We observe high Round Trip Times (RTTs) and inflated number of round trips per page download (RTT multipliers). Due to inefficiencies in TCP's slow start and the HTTP protocol, we found several opportunities to reduce the RTT multiplier, e.g. increasing TCP's Initial Congestion Window (ICW), using TCP Appropriate Byte Counting (ABC), and using HTTP pipelining.
Using live workloads, we experimentally study the micro effects of these optimizations on network connectivity, e.g. packet loss rate. To evaluate the macro effects of these optimizations on the overall page load time, we use realistic synthetic workloads in a closed laboratory environment. We find that compounding HTTP pipelining with increasing the ICW size can lead to reduction in page load times by up to 80%. We also find that no one configuration fits all users, e.g. increasing the TCP ICW to a certain size may help some users while hurting others.
- Global web stats. http://www.w3counter.com/globalstats.php, 2010.Google Scholar
- SPDY: An Experimental Protocol for a Faster Web. http://dev.chromium.org/spdy/spdy-whitepaper, 2010.Google Scholar
- Allman, M. Tcp byte counting refinements. SIGCOMM Comput. Commun. Rev. 29 (July 1999), 14--22. Google ScholarDigital Library
- Allman, M. A web server's view of the transport layer. SIGCOMM Comput. Commun. Rev. 30 (October 2000). Google ScholarDigital Library
- Allman, M. TCP Congestion Control with Appropriate Byte Counting (ABC). RFC 3465, IETF, 2003. Google ScholarDigital Library
- Allman, M. tcpsplit. http://www.icir.org/mallman/software/tcpsplit/, 2010.Google Scholar
- Allman, M., Floyd, S., and Partridge, C. Increasing TCP's Initial Window. RFC 3390, IETF, 2002. Google ScholarDigital Library
- Arbor Networks. 2009 Internet Observatory Report. http://www.nanog.org/meetings/nanog47/presentations/Monday/Labovitz\_ObserveReport\_N47\_Mon.pdf, 2010.Google Scholar
- Balakrishnan, H., Rahul, H. S., and Seshan, S. An integrated congestion management architecture for internet hosts. In Proceedings of SIGCOMM'99 (New York, NY, USA, 1999), ACM. Google ScholarDigital Library
- Chu, J., Dukkipati, N., Cheng, Y., and Mathis, M. Increasing TCP's Initial Window. http://tools.ietf.org/html/draft-ietf-tcpm-initcwnd-01, 2011.Google Scholar
- Dukkipati, N., Refice, T., Cheng, Y., Chu, J., Herbert, T., Agarwal, A., Jain, A., and Sutin, N. An argument for increasing tcp's initial congestion window. SIGCOMM Comput. Commun. Rev. 40 (June 2010). Google ScholarDigital Library
- Fielding, R., Gettys, J., Mogul, J., Frystyk, H., Masinter, L., Leach, P., and Berners-Lee, T. Hypertext Transfer Protocol -- HTTP/1.1, 1999.Google Scholar
- Freedman, M. J., Freudenthal, E., and Mazières, D. Democratizing content publication with coral. In Proceedings of NSDI'04 (Berkeley, CA, USA, 2004), USENIX Association. Google ScholarDigital Library
- Hopkins, A. Optimizing Page Load Time. http://www.die.net/musings/page_load_time/, 2010.Google Scholar
- Krishnamurthy, B., and Wang, J. On network-aware clustering of web clients. In Proceedings of SIGCOMM'00 (New York, NY, USA, 2000), ACM. Google ScholarDigital Library
- Krishnan, R., Madhyastha, H. V., Srinivasan, S., Jain, S., Krishnamurthy, A., Anderson, T., and Gao, J. Moving beyond end-to-end path information to optimize cdn performance. In Proceedings of IMC'09 (New York, NY, USA, 2009), ACM. Google ScholarDigital Library
- Leighton, T. Improving Performance on the Internet. Commun. ACM 52 (February 2009). Google ScholarDigital Library
- Nottingham, M. Making HTTP Pipelining Usable on the Open Web. http://tools.ietf.org/html/draft-nottingham-http-pipeline-00, 2010.Google Scholar
- Nygren, E., Sitaraman, R. K., and Sun, J. The akamai network: a platform for high-performance internet applications. SIGOPS Oper. Syst. Rev. 44 (August 2010). Google ScholarDigital Library
- Olshefski, D., and Nieh, J. Understanding the management of client perceived response time. In Proceedings of SIGMETRICS'06/Performance'06 (New York, NY, USA, 2006), ACM. Google ScholarDigital Library
- Ostermann, S. tcptrace. http://www.tcptrace.org/, 2010.Google Scholar
- Padmanabhan, V. N., and Katz, R. H. TCP Fast Start: A Technique For Speeding Up Web Transfers. In IEEE Globecom (1998).Google Scholar
- Qian, F., Gerber, A., Mao, Z. M., Sen, S., Spatscheck, O., and Willinger, W. Tcp revisited: a fresh look at tcp in the wild. In Proceedings of IMC'09 (New York, NY, USA, 2009), ACM. Google ScholarDigital Library
- R. Stewart, E. Stream Control Transmission Protocol. RFC 4960, IETF, 2007.Google Scholar
- Souders, S. High-performance web sites. Commun. ACM 51 (December 2008). Google ScholarDigital Library
Index Terms
- Overclocking the Yahoo!: CDN for faster web page loads
Recommendations
Anycast-aware transport for content delivery networks
WWW '09: Proceedings of the 18th international conference on World wide webAnycast-based content delivery networks (CDNs) have many properties that make them ideal for the large scale distribution of content on the Internet. However, because routing changes can result in a change of the endpoint that terminates the TCP session,...
Research papers: A study of traffic from the perspective of a large pure IPv6 ISP
Our understanding of IPv6 traffic cannot keep up with the growth of IPv6 traffic. Unraveling the characteristics of traffic is essential for network scale expansion, network technology selection, network management and security enhancement. In this ...
End-User Mapping: Next Generation Request Routing for Content Delivery
SIGCOMM '15: Proceedings of the 2015 ACM Conference on Special Interest Group on Data CommunicationContent Delivery Networks (CDNs) deliver much of the world's web, video, and application content on the Internet today. A key component of a CDN is the mapping system that uses the DNS protocol to route each client's request to a ``proximal'' server ...
Comments