This is an attempt at a guide and a few helpful pointers for those who don't care to use the prepackaged NAV for Debian.
Make sure you have the non-free archive present in your
deb http://ftp.no.debian.org/debian/ lenny main contrib non-free deb-src http://ftp.no.debian.org/debian/ lenny main contrib non-free deb http://security.debian.org/ lenny/updates main contrib non-free deb-src http://security.debian.org/ lenny/updates main contrib non-free
Install NAV's runtime dependencies:
sudo aptitude install perl python sun-java6-jdk apache2-mpm-prefork \ cricket postgresql-8.3 libapache2-mod-python \ libpg-java libsnmp1.4-java python-pysnmp2 logrotate \ python-cheetah python-xmpp python-django \ python-egenix-mxdatetime python-gammu python-ipy \ python-ldap python-psycopg2 python-pyrad \ python-pysnmp-se python-twisted python-twisted-snmp \ python-rrd rrdtool samba-common
Be aware that NAV doesn't run well under the kaffe Java VM, which is why you should confirm that Sun's Java JDK is installed (
sun-java6-jdk) and is the active alternative for the
sudo update-alternatives --config java sudo update-alternatives --config javac
Next, NAV's build time dependencies:
sudo aptitude install ant autoconf automake libsnmp1.4-java python python-cheetah sun-java5-jdk make
tar xvzf nav-3.7.0.tar.gz cd nav-3.7.0 ./configure CRICKETDIR=/usr/share/cricket export CLASSPATH=/usr/share/java/libsnmp-1.4.jar make sudo make install
Note: avoid using -j when calling make. It will break the build, since make tries to run many ant processes with interdependencies at the same time.
ident sameuser into
md5 for local users connecting through UNIX sockets (except the postgres user). The file should look something like this:
local all postgres ident sameuser local all all md5 host all all 127.0.0.1/32 md5 host all all ::1/128 md5
Then reload PostgreSQL to make the changes effective:
sudo invoke-rc.d postgresql-8.3 reload
While still in the NAV source code directory,
doc/sql. The following set of commands assume you have sudo access to run commands as the
postgres system user. If not, please become the
postgres user in some other way. The procedure has been adapted from
To create the PostgreSQL user and database:
# The following command will prompt for a password for the new database user. # Pick one of your own choosing. sudo -u postgres ./createdb.sh
Then, take the password you chose for the database user and put this as the value to the
userpw_nav option in
We do this because we don't install the libraries to the systemwide Python library directories by default. We do however want them available for use systemwide:
sudo bash -c 'cat << EOF >> /etc/python2.5/sitecustomize.py import sys __navpath = "/usr/local/nav/lib/python" if __navpath not in sys.path: sys.path.append(__navpath) EOF '
cd /usr/local/nav/lib/java sudo ln -s /usr/share/java/postgresql.jar sudo ln -s /usr/share/java/libsnmp-1.4.jar
Put this in
ServerName nav.example.com ServerAdmin email@example.com DirectoryIndex index.html index.html index.php index.cgi index.py AccessFileName .htaccess AddDefaultCharset UTF-8 AddHandler cgi-script .cgi DocumentRoot /usr/local/nav/apache/webroot <Directory /> Options FollowSymLinks AllowOverride None </Directory> <Directory "/usr/local/nav/apache/webroot"> Options Indexes FollowSymLinks ExecCGI AllowOverride All Order allow,deny Allow from all </Directory> LogLevel warn <Location /> # Enable debugging on errors, if you like PythonDebug on </Location>
Then run the following commands:
sudo a2dissite default sudo a2ensite default-nav sudo a2enmod mod_python sudo invoke-rc.d apache2 force-reload
Create groups and users, and set proper permissions:
sudo addgroup --system nav sudo adduser --system --no-create-home --home /usr/local/nav \ --shell /bin/sh --ingroup nav navcron; sudo addgroup navcron dialout cd /usr/local/nav/var sudo chown -R navcron:nav .
/usr/local/nav/etc/nav.conf, set your
DJANGO_DEBUG=True if you want complete error information when stuff goes awry in the web interface.
Optionally make the nav process control command more available:
sudo ln -s /usr/local/nav/bin/nav /usr/sbin/nav
Cricket should already be installed as it is a part of the runtime dependencies.
/etc/cron.d/cricket or comment out the active cron job line in the file.
Create a directory to place the nav cricket-config:
sudo mkdir /usr/local/nav/etc/cricket-config
Change owner to navcron:nav
sudo chown navcron:nav /usr/local/nav/etc/cricket-config
Edit the /etc/cricket/cricket-conf.pl and set the configroot variable. At the same time, fix the logging. Cricket defaults to /var/log/cricket as logdir. This directory is created on install, but the owner is cricket. Either create your own logdir and set the correct path ($gLogDir) in /etc/cricket/cricket-conf.pl or change ownership of /var/log/cricket to navcron.
$gConfigRoot = "/usr/local/nav/etc/cricket-config" $gLogDir = your/log/dir
Copy /usr/local/nav/doc/cricket/cricket/subtree-sets to /etc/cricket/
cd /etc/cricket sudo cp /usr/local/nav/doc/cricket/cricket/subtree-sets .
As navcron, copy recursively /usr/local/nav/doc/cricket/cricket-config/ to your cricket-config
sudo -u navcron cp -r /usr/local/nav/doc/cricket/cricket-config /usr/local/nav/etc/
Create a directory to store Cricket's data files, and give
navcron ownership of that directory.
sudo mkdir /usr/local/nav/var/cricket-data sudo chown navcron /usr/local/nav/var/cricket-data
Go to the cricket-config directory and edit the “Defaults” file. Replace the Datadir with the correct path to cricket-data. Don't forget the %auto-target-path%
Target --default-- dataDir = /usr/local/nav/var/cricket-data/%auto-target-path%
Compile the cricket-config as navcron to create the cricket-database.
sudo -u navcron cricket-compile
mcc.py to create config from the database. The first time, remember that the database should contain data before doing this. Also, it's a good idea to start
ipdevpoll and let it run a good couple of minutes before making the cricket-config as it then will have gathered and stored information about devices and interfaces in the database.
cd /usr/local/nav/bin sudo -u navcron ./mcc.py
Now, start Cricket:
cd /usr/local/nav/bin sudo ./nav start cricket
Create the cricket directory in apache/webroot/
cd /usr/local/nav/apache/webroot sudo mkdir cricket
Make a symlink to the cricket cgi-scripts and copy cricket.css from source.
cd /usr/local/nav/apache/webroot/cricket sudo ln -s /usr/lib/cgi-bin/cricket/grapher.cgi sudo ln -s /usr/lib/cgi-bin/cricket/mini-graph.cgi sudo ln -s grapher.cgi index.cgi sudo ln -s /usr/share/cricket/images sudo cp /usr/local/nav/doc/cricket/public_html/cricket.css .
Create .htaccess-file to make sure mod_python doesn't handle this directory.
sudo bash -c 'echo "SetHandler none" > .htaccess'