3 ms·
The performance analysis is pretty bad. Comparing synchronous I/O performance against asynchronous I/O based solely in terms of bits per second transferred onl
by otterley 16y ago
The performance analysis is pretty bad. Comparing synchronous I/O performance against asynchronous I/O based solely in terms of bits per second transferred only tells a small part of the story. It doesn't show how well the server scales as more and more connections are added, which absolutely essential information. It would have been far better if the author had plotted the connection rate and response time against the transfer size per connection.