pjchilds@imforei.apana.org.au writes:
>In article <4q406f$gli@al.imforei.apana.org.au> you wrote:
>
>: I have a small problem with the logfile analysis scripts. After I finally
>: found out that the scripts didnŝt like the httpd-ish logfiles I got them
>
> Yes.. that caught me too :)
You can give the -h option to have the scripts eat the httpd format
(I think it still works).
>: to give me some statistics on Client Usage and others. The only thing
>: that still doesn't work is the Top 25 listing.
>
> I find that initially this worked, then as the logs grew a little
> (we are only a small site) that perl started core dumping
> (perl 5.002 under FreeBSD 2.1-stable)
>
> I haven't had time to check it out, but i think the perl uses
> associative arrays it builds on the urls, and that these just
> get way too big with big logs..
This is why the access-extract-urls.pl uses dbm databases for
the assoc arrays. It opens them in /var/tmp/...
Duane W.
Received on Tue Jun 18 1996 - 11:06:48 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:32:31 MST