Busy Internet archives generate large logs for each access method bein
g used. These raw log files can be difficult to process and to search.
This paper describes a system for reading these growing logs, a combi
ned log file format into which they are re-written and a system that a
utomates this building and integration for multiple access methods. Au
tomated summarizing of the information is also provided giving statist
ics on accesses by user, site, path-name and date/time amongst others.