I have a ruby on rails website that is hosted on unicorn webserver behind a nginx reverse proxy. I want to evaluate if using spdy will enhance my performance under this setup.
Specifically, I need the directions on the following:
- does using reverse proxy nullify the gains from using spdy?
- what performance/load testing tool can i use to evaluate gains from using spdy?
- is there a way to force nginx to only talk spdy?
In regards to the second question i found that neoload and loadrunner 12 does support spdy.
With neoload i'm having trouble getting it to recognize that my server is using spdy despite chrome://net-internals/#spdy and http://spdycheck.org/ telling me that it IS using spdy (I'm using spdy 3.1, when i used spdy 2 neoload did recognized it).
With loadrunner i don't have yet found a way to be sure that it is testing spdy. I'm using the spdy_url function however the documentation does not tell me if spdy_url assures me that spdy is being used underneath. One way to be sure would be to configure nginx to only talk spdy (third question).