I tested each against the invoice URL for the stats I included, but I've had similar results from the customer and product lists. It should be the same JSON data every time. I didn't want to post the actual URI for fear that my server would suddenly get pounded by an 'ab -c 20 -n 100' flood. Computer geeks tend to be curious types... *smirk*
I didn't notice that Racket's total bytes transferred was so far off; I wonder about that too. I wonder if it represents failed requests. Racket seems to be able to keep up with Perl/Starman with -c 1; both embarass Ruby/Unicorn, I'm sad to say. It's only when the number of concurrent requests get large that Perl begins to trounce Racket. I suspect it might have more to do with Starman vs Racket's server than Perl vs Racket. Glad to know it's not that my first Racket program was terrible. :) TK On Thu, Apr 12, 2012 at 06:30:35AM -0600, Ryan Culpepper wrote: > > What URL(s?) are you testing? It's hard to analyze the numbers below > without knowing what they're measuring. > <snip> > Are you sure the "Failed requests" are really failures? A > Stackoverflow answer suggests that these might be the result of a > nondeterministic response length. (See > http://stackoverflow.com/questions/579450.) > > It's odd that the total transferred for the Racket benchmark is > almost an order of magnitude less than the totals for Perl and Ruby. > It would help to see the actual URLs used in the tests to make sure > this isn't an apples-to-oranges comparison. > > It would also be useful to run the benchmarks with "-c 1" to measure > the pure sequential performance. I'm not familiar with the > frameworks in question, but a brief scan suggests that Starman and > Unicorn might be forking multiple OS-level workers, which would lead > to more parallelism on multicore systems. Your Racket setup is > limited to a single core. > ____________________ Racket Users list: http://lists.racket-lang.org/users
