Weblog

Weblog is a suite of programs used to analyze a web server's access_log files on a regular basis (typically weekly). It was created because most existing web log analysis programs tend to provide too much details, and are often oriented for the use of people maintaining the web server rather than that of its authors. Weblog is meant to be run by web administrators for their web authors who are interested in tracking how often their pages are viewed.

Weblog has been in use since February 1996 on the UC Santa Cruz central web servers (www, and www2). Weblog started as a combination of awk and sed scripts tied together by a C-shell script, but now consists of several Perl programs.

weblog_analyze: Collects the data from log files (in either NCSA or W3C formats), saves the hits for each URL, and generates summary, report, and error HTML files. It is typically run on a weekly basis as a cron job after the logs have been rotated.
weblog_report: Reads the database created by weblog_analyze, and can track arbitrary set of pages over arbitrary time periods, by presenting a spreadsheet-like matrix of pages and weekly statistics. This program is usually run as a cgi-bin script.
weblog_rotate: Rotates, compresses, and removes older log files. This program is usually invoked on a regular basis (e.g. weekly) by crontab. It is typically run on a weekly basis as a cron job before splitting and analyzing the logs.
weblog_split: Split the logs into virtual hosts, and resolves numeric IP addresses. It is typically run on a weekly basis as a cron job after the logs have been split, but before being analyzed.

The start of an FAQ about weblog is available.