Statistics used by optimizer.

[Rakshit Trivedi]

Hi,

I wanted to know about the use of statistics by MYSQL optimizer. According to what I have followed, for estimating the number of rows to be read from a table , the optimizer probes the indexes and calculates the values. Is it true that for different constant values for a given query, the optimizer probes and accesses the indexes to estimate the rows? Or does it only use the already stored statistics of the table?

And one more question. While browsing on summaries and statistics used by MYSQL, I came across the following post:
http://forge.mysql.com/worklog/task.php?id=4816

Here, the reasons for not preferring the histograms are presented. I just wanted to know how the use of histograms leads the optimizer not to use right instructions mentioned there? And how not using histograms is beneficial to optimizer?

Thanks.