HTTP/2 vs. HTTP/1.1: A performance analysis
The new HTTP/2 protocol is mostly focused on web performance, but what kind of performance improvement can you expect by switching to HTTP/2? Ragnar Lnn explains the key improvements within HTTP/2 and demos a new application developers can use to see how their applications will function on the new protocol.
Talk Title | HTTP/2 vs. HTTP/1.1: A performance analysis |
Speakers | Ragnar Lonn (Load Impact) |
Conference | Fluent |
Conf Tag | The Web Platform in Practice |
Location | San Francisco, California |
Date | March 8-10, 2016 |
URL | Talk Page |
Slides | Talk Slides |
Video | |
HTTP/2 is happening—most major browser and server makers are working on support for HTTP/2, and some, like Firefox, already have it enabled by default. The protocol is backward-compatible with HTTP/1.1 and will soon become ubiquitous when new versions of clients and servers switch it on. But the web development community is still largely ignorant about HTTP/2 and what it means for their applications. Ragnar Lönn presents the findings of a study where he measured the performance impact different types of applications would typically get if they switch their data transport from HTTP/1.1 to HTTP/2. In the study, Ragnar emulated different HTTP application types while communicating with application backends over HTTP/1.1 and HTTP/2, in order to find out how much performance and user experience differed between the two protocols. Ragnar set up a test bed where he could, under controlled conditions, simulate different use cases in terms of the number of concurrently used objects and object size/weight in order to cover everything from the “REST API” scenario to the “classic web page” scenario. He also simulated different network characteristics to capture the user-experience impact of different types of Internet connections—particularly mobile connections that experience high network delay. As an extra bonus, Ragnar will demo “HTTP/2 vs. HTTP/1.1: A Performance Analysis,” a free-to-use web application he developed that will directly compare a website’s performance over HTTP/2 to its current performance using HTTP/1.1.