Surftrackr

Squid - Dansguardian log analizer

	Skill level: Advanced
	The instructions on this page may require deviations from standard procedures. A good understanding of linux and Koozali SME Server is recommended.

Description

Surftrackr is a log file viewer for Squid and Dan's Guardian. It makes it easy for you to monitor Web usage, including the type of information accessed, the sites visited by your web users, and the amount of information downloaded. Best of all, Surftrackr is free and runs in any web browser.

This is an alternative for Sarg.

Read more about Surftrackr.

PLEASE READ EVERYTHING BEFORE CONTINUE

This contribution is not a traditional adaptation, so you should know exactly what you are doing. The sticking points are: Surftracker is based on the Django framework, that despite the fact that it works correctly under python 2.3 (which come with SME 7.x), but do not get this right Surftrackr. Therefore it is necessary to upgrade to Python 2.4. But all precautions were taken so that this does not affect the natural python settings of SME 7.x Another important point is the use of mod_python. If you already use mod_python for another application, you must leave here this installation. mod_python is compiled under python 2.3, and our customized version is compiled under python 2.4, so an error will occur. You have two choices: Uninstall your mod_python and install this customized version (be careful when using Trac or another application that makes use of mod_python). Use FastCGI with Django instead of mod_python as described here: http://docs.djangoproject.com/en/dev/howto/deployment/fastcgi

Installation

Before install rpms, you must create a new ibay named surftrackr. Then run the following commands:

db accounts setprop surftrackr AllowOverride all
db accounts setprop surftrackr FollowSymLinks enabled
signal-event ibay-modify surftrackr

Note:

Currently I have no time to test apache rewrite rules for install under /opt directory, because Surftrackr require a FQDN main domain. See the entire Surftrackr site and blog because have a lot of info about this. May be you can write the rewrite rules and we can install under /opt dir without a FQDN. I think it is posible if you have the know how with python code because I am a very very newbie with python. If you can do this, share with us

Create the database logs.

mysqladmin create logs --default-character-set=utf8
mysql
GRANT ALL PRIVILEGES ON `logs` . * TO 'logsu'@'localhost' IDENTIFIED BY 'logsp' WITH GRANT OPTION ;
quit

First download and install python 2.4 (skip this step if you have already Pootle python 2.4 in your system):

wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/pootle/python2.4/rpms/alternatives-0.2.0rc3-9.noarch.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/pootle/python2.4/rpms/python24-2.4.2-10.el4.pyv.i386.rpm

yum localinstall *.rpm

Then run:

/usr/sbin/alternatives-helper --remove python24
mv /usr/bin/python.alternatives_save /usr/bin/python
mv /usr/share/man/man1/python.1.gz.alternatives_save /usr/share/man/man1/python.1.gz

Now download the following rpms into another clean directory:

wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/Django-0.97_pre-1.noarch.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/mod_python-3.1.3-5.1.custom.i386.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/MySQL-python-1.2.2-1.i386.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/pygooglechart-0.2.2-1.noarch.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/pyparsing-1.5.1-1.noarch.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/python-dateutil-1.4.1-1.noarch.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/setuptools-0.6c9-1.noarch.rpm
wget http://mirror.contribs.org/smeserver/contribs/nhall/sme7/contribs/surftrackr/rpms/surftrackr-20080326-1.noarch.rpm
wget http://mirror.contribs.org/smeserver/releases/7/smecontribs/i386/RPMS/smeserver-mod_python-0.1-1.el4.sme.noarch.rpm

Install:

yum localinstall *.rpm

Into surftrackr html directory open the file settings.py and edit the following variables:

• LOG_FILE - If you have danguardian, change to /var/log/dansguardian/access.log and set DG logfileformat = 3 (squid format)
• DISABLE_AUTH - Look at http://surftrackr.net/blog/view/28/multi-user-surftrackr/ but leave as False
• LIVE_LOGGING and DATA_FROM_LOG - Look as Surftracker site, but leave as True and False respectivelly
• TIME_ZONE - It is important because set the local time to reports
• MEDIA_URL - Modify this line to your IP or FQDN
• SECRET_KEY - Change this key. DO NOT CHANGE THE LONGITUDE

Open the file .htaccess and comment require valid user and uncomment require user admin to allow only admin access.

Usage

You can use to LIVE logs or to proccess existing logs. The default allow to live logs in real time!

cd /home/e-smith/files/ibays/surftrackr/html/utils
python2.4 surftrackr-livelog.py

Then go at your IP or FQDN with a browser and go to LIVE page and click at GO to see what happens.

If you want to proccess existing logs change LIVE_LOGGING to False and DATA_FROM_LOG to True

Then run

cd /home/e-smith/files/ibays/surftrackr/html/media
python2.4 logfiles.py

You can set a cron job to do this at 5 or 10 minutes interval. The first time run this manually because Surftrackr will process a lot of old logs info.

It is VERY IMPORTANT to run the above script with python2.4 and NOT with python. Also it is very important to change directory to the script before runing. Do NOT run as python2.4 /home/e-smith/files/ibays/surftrackr/html/media/logfiles.py. Change to dir before run.

As this [doc http://surftrackr.net/blog/view/28/multi-user-surftrackr/] say, every user logged at Surftrackr is created at Django DB. So read all ST documents and blog to know all the features. Because this is not fully documented, I advice you to read all blog post entries from the first to last and download the PDF docs.