not faster that http-proxy in benchmark #9

Open
dominictarr opened this Issue Nov 8, 2011 · 3 comments

Comments

Projects
None yet
2 participants

hey, I ran your benchmark, and was sorry to find that I did not see it was faster than http-proxy

also, the variance on 100 requests isn't enough to be very sure, it's better to do 10,000.

and better yet to apply some statistics...

any way, here are my results on -n 10000 -c 100, I got similar results with -n 1000 -c 10

https://gist.github.com/1347231

http-proxy higher average requests per second than bouncy, although it's worth noting that bouncy has a lower standard deviation.

can you post your results?

just read your code,

I gotta say, I quite like your api, and the code is way more elegant than http-proxy.

oh, hang on.

I see that your claim what that 0.0.4 was faster.

that is correct, it's much faster.
but bouncy 1 is slower than http-proxy

Owner

substack commented Nov 8, 2011

Yep it got slower when I moved to parsley. Parsley could use some performance work in buffering messages more.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment