Closed iislucas closed 9 years ago
So, we now have echo-server-stresser.sh
.
@iislucas what else did you want to do in terms of performance testing for the echo server? Perhaps we should broaden this issue's title to include the SOCKS server benchmarking which we've already started?
OK, let's close this issue in its current form. Going forward I would like some continuous test framework so we know what impact is on performance. Something that can be run from a Gruntfile. Let's chat with mLab folk, and figure out what the best metrics are. I'm thinking something like a matrix of throughput/latency vs number of connections & qty of data. This would be for both socks-rtc and also echo-server. Something to understand on an objective way our JS performance, and carry out tests in the longer run, e.g. impact of CHURN on it.
We should have performance tests for our networking code.
e.g. for a simple test for echo-server: see https://github.com/uProxy/uproxy/issues/439