Joe Armstrong's report of testing the performance of web servers YAWS and Apache under heavy load.
I like that the experiment is simple, and the description of it reflects this. Very clean. Very neat.
February 12, 2003 11:55 AM
| TrackBack
I find this pretty strange, to me it does not seem to be too practical. Apache falls over. Yaws continues, but only at ~10 bytes/sec per connection.
So the 20 kbyte test pages will be downloaded in approximately 40 minutes.
Just as a reference, the BBC news website (80K) would load in ~2 hours, with no pictures.
A short email message is around 2 kbytes (with headers, but no html tags to render it nicely) -> 8 minutes.
Still curious about erlang, tho :)
Correct me if my calculations are wrong :)
What about it is simple? I would love for someone to actually "repeat" this "experiment" - it is not do-able. Largely, this is because very little derail is available as to how this "test" was actually performed. The scientific-backing to thee results noted does not exist.
Comment by: Erik Onnen May 31,2008