squidpeek 1.4

About Squidpeek

This is a quick and dirty script for getting cool per-URL statistics and sparklines out of your Squid access logs.

Because it’s per-URL, it’s most appropriate for use with gateways (aka “reverse proxies”).

For a sample of its output, see http://mnot.github.com/squidpeek/.

Requirements and Installation

Squidpeek needs Python 2.5 or greater; see http://python.org/.

You’ll also need the Python Imaging Library (PIL); see http://www.pythonware.com/products/pil/.

The easy way to install is with pip;

pip install squidpeek

Obviously, you also need some Squid logs; see http://squid-cache.org/. Other programs that generate Squid logs (e.g., Traffic Server) may or may not work, depending on how faithfully they follow the format’s semantics.

Finally, you’ll need a browser that can handle data: URLs. Note that less-than-recent versions of IE can’t do this.

Using Squidpeek

The command-line interface is like this:

% squidpeek.py [-q] [-n num] logfile
    -d Debug parse errors
    -q use the query string as part of the URI
    -n [num] show the top num URLs (default: 100)

Typically, you’d use squidpeek in a cron job, like this:

# run once an hour; assumes logs are rotated right beforehand
2 * * * * root squidpeek /var/log/squid/access_log

Support and Contributions

Please feel free to file issues above; this code doesn’t see a lot of maintenance, so I can’t promise a quick response.

All input is welcome, particularly code contributions via a Github pull request.

Maintainer: Mark Nottingham mnot@mnot.net