Simon Riggs wrote:
> - yes, the random sampling is random - please read the code and comments
>
> - yes, I would expect the results you get. If you sample 5% of rows and
> each block has on average at least 20 rows, then we should expect the
> majority of blocks to be hit.
and it seems from the benchmark posted to this list that random is
_very_ expensive (probably because the random reads are spread out so
well, that we do alot of I/O instead of just logical I/O from some cache)
regards,
Lukas