searching log files for fun and profit.

rick taught me new and exciting ways to search through apache log files to find where the hits to my site have been coming from. If I were in charge of the server, it'd be easy. But since I exist here only through the grace of god I try to tread lightly.

After punching in an exceptionally long, heavily pipe'd command,(the actual command goes as follows: cat /usr/local/apache/logs/access_log | grep chico | awk '{ print $1 }' | tail -200 | sort -u | more ) i get a string of results. From that I can use some guess work to figure out who's who. For example:

cache-mtc-ac02.proxy.aol.com

There's a short list of people who could be attached to that name. Another entry would be:

crawler11.googlebot.com

that's the google fairy paying me a visit. The interesting links follow. Part of the host name has been removed, just in case it's a visitor from a machine that has a static host name.

9-29 More fun.

This site has had five unique (non-search spider) hits over five days, averaging a whopping one person per day. A really talented geek could automate those figures.

9-30

10-6

New Fun and games. Slashdot has an article on it's main page today concerning the way google is handling some search patterns. I've added a simple, basically blank page with some correct keyword meta tags to see if i can generate a spike in the number of hits to this site. And to annoy a percentage of slashdot readers and Google Whackers.

The Google fairy happened to stop by here just before I uploaded the page, so the traffic may take a while to come it, if it comes at all. I still need a non-intrusive way of tracking the loads of just that particular page. Never mind. I can make grep do that.

Dear Diary; Today I learned that you can use grep for phrases other than just your user name which will still return the results you want, So long as the phrase you use is unique to your site name.

home