MySQL benchmarks

[Ivan Voras]

Maxim Sobolev wrote:
> The difference between linux and everything else in those tests is so 
> significant even on UP, so that I suspect that something wrong with the 
> approach.

The only way to prove it is to make a better-controlled benchmark. But I
wouldn't bet on it - from things seen on this and performance@ lists,
the numbers, somewhat sadly, seem about right.

Of course, I might just be overly pessimistic :)