several months ago, after i set up a squid cache at my work location i got to looking around and realized that while there are a number of scripts that perform various kinds of queries against the squid access.log, there are few tools to enable the kind of longer-term analysis that could provide input for tuning the cache. so i wrote a little script for parsing the log and inserting the results into a mysql table. at that point i realized that the elements in the table are not generally well-suited for statistical analyses, so i started thinking about proxy variables and other transforms. ultimately, i launched into writing a script that pulled records from the mysql table, flagged certain type of dates and did a couple of other things, and inserted the results in a postgresql table. the scripts i use, the sql statements used to create the tables, and a couple of pages of general information can be found here. (sorry about the ads ... it's free, right?)
this is obviously a work in progress, but anyone who wants to use any part of it is welcome to, subject to the standard disclaimers.
|
|---|