Several count fetchs, scalability?

Thursday 05 February 2004 4:53:06 am

Hi

I want to make an statistical page about various information about the users of my site. In order to do this I will need to make around 10 tree_count fetches (i.e: fetch('content',tree_count) on various attributes.

Example:
Show percent of male users

fetch total number of users
fetch number of user with male attribute="m"

Calculate male %.

My question is two folded:

1: Would it be faster to first get all users using an standard fetch (non count), and then loop through them and count the various relevant attributes settings(6-8). And in the end perform the neccesary calculations. Or is it faster to have several fetch counts on various attributes?

2: Any thoughts about how these solutions would scale when the number of user go up, I know that the system caches, but when the number of user goes up, so does the frequncy of user editing themselfs, i.e. cache expires faster.

Nice if someone had any thoughts or experience to share on this one.

By the way I really like the new layout for the documentation of the various functions. Once it is finished it will greatly reduce development time and allow more functional EZ sites I belive.

Cheers
Tore