You’re doing it wrong
June 19th, 2010
Poul-Henning Kamp in acmqueue:
On a modern multi-issue CPU, running at some gigahertz clock frequency, the worst-case loss is almost 10 million instructions per VM page fault. If you are running with a rotating disk, the number is more like 100 million instructions.
What good is an O(log2(n)) algorithm if those operations cause page faults and slow disk operations? For most relevant datasets an O(n) or even an O(n^2) algorithm, which avoids page faults, will run circles around it.
Besides being an interesting article, this is the first time I read about Varnish. I know lighttpd, but haven’t heard about Varnish before . Anyone ran into that before? Maybe even used it?