Turbostart quide

Today, the fastest way of a Nagios install is using OMD. If you rather want to build the most current version of check_mk, read on.

This guide assumes that you have experiance with Nagios und Linux and that you are impatient (like me sometimes). It doesn't explain why something is done and does not give examples. Please read the article Getting Started if you need a more detailed introduction.

Preparations

Before you can setup check_mk on your Nagios server please make sure that you've made the following preparations:

• You have a fully functional Nagios up and running.
• Your can issue commands to Nagios via its WWW interface.
• You've installed Python in version 2.3, 2.4, 2.5 or 2.6.
• You have xinetd installed.

How to install and setup check_mk

1. Install check_mk

Download the most recent tarball of check_mk from the projects homepage. Unpack it somewhere. Become root, change to the unpacked directory and run:

root@linux# ./setup.sh

You can change the installations paths if you want. A a value is autodetected then you can safely assume that it is correct. Changing such a value breaks something.

2. Install the agent on localhost

The agent is setup by installing it as service for xinetd. Assuming the default installation paths this is done with the following commands:

root@linux# cd /usr/share/check_mk/agents
root@linux# cp check_mk_agent.linux /usr/bin/check_mk_agent
root@linux# cp xinetd.conf /etc/xinetd.d/check_mk
root@linux# /etc/init.d/xinetd restart

3. Setup monitoring for localhost

Edit /etc/check_mk/main.mk and add localhost to all_hosts:

all_hosts = [ "localhost" ]

1.1.9i2 Starting with this innovation release the inventory types become mandatory. Use any of tcp,ping,snmp.

all_hosts = [ "localhost|tcp" ]

After that remove localhost from your Nagios configuration if present. Otherwise Nagios won't start because of a duplicate host entry.

Now you can inventorize your new host. This will detect a couple of checks and automatically configure them:

root@linux# check_mk -I tcp

After that you can call check_mk with the option -O. That will create Nagios configuration files for check_mk's hosts and services, create precompiled host checks and reload Nagios.

root@linux# check_mk -O

After this Nagios should show the new checks for localhost in its web interface.

Troubleshooting

If something goes wrong, please check:

• Did setup.sh really autodetect all values needed for Nagios integration? If not, please enter them manually. You can repeat setup.sh at any time.
• Does your Nagios correctly react to external commands?
• Is your agent on localhost correctly installed? A telnet localhost 6556 should show the agents output.
• Was your Nagios process running while doing setup.sh?
• Does your Nagios correctly read in the two files check_mk_templates.cfg and check_mk_objects.cfg?

You might also want to have a look into the article "Getting Started", which is more detailed than this one.