Google Chrome Speed Tracer what does Request Timing and Response Timing actually measure?

Question

I'm testing out the Google Chrome Speed Tracer on a few common web pages and taking a look through the results. One thing I'm not sure I understand is what the "Request Timing" and "Response Timing" properties of resources are actually measuring.

Initially I thought Request Timing must measure the time from a request for a resource being sent and when that request arrived at the server. However, I then wondered how the Speed Tracer would actually have any way of measuring this. Furthermore, the Response Timing that I'm getting for resources tends to be far less than the Request Timing (e.g. 500ms request, 1ms response), which is a little bit suss.

So is anyone able to explain exactly what Request Timing and Response Timing are measuring?

Answer 1

The Request Timing is how long the request took before data began to be received from the server. Response Timing is how long that data took to be sent back to the client until completion.

Answer 2

I've fooled around with Speed Tracer and my GWT application until I had an idea of what these times could mean (Google doesn't seem to love writing documentation, so I can only guess). My interpretation:

Request Timing: Transfer and processing (server-side) of the request

Response Timing: Transfer and processing (client-side) of the response

The request timing includes everything that happens on the server and the response timing everything that happens on the client, including XmlHttpRequest callback code (as well as object serialization in the case of GWT).

Hope that helps.