nagios fail over

nagios-failover

script to failover nagios servers if and when their down – automtically done via bash script that will require some ground work to be done in order to achieve the end result of each nagios server looking out for one another.

I have also done a few other things

cd /etc/;

ln -s /usr/local/nagios/etc ./nagios

cd /etc/nagios;

ln -s /usr/local/nagios/libexec ./plugins

Requirements.

Actual nagios cfg files –

CMS (configuration Management System – puppet,chef or cfengine)

I have used puppet and used the recursive folder plugin to push all the actual cfg files to nagios nodes:

$objects=”/usr/local/nagios/etc/objects”
$company_objects=”$objects/company”

file { “$company_objects”:
ensure => directory, # so make this a directory
recurse => true, # enable recursive directory management
purge => true, # purge all unmanaged junk
force => true, # also purge subdirs and links etc.
owner => “nagios”,
group => “nagios”,
mode => 0644, # this mode will also apply to files from the source directory
# puppet will automatically set +x for directories
source => “puppet:///modules/nagios/company”,
}

If you do not have cms in place you could configure this script to use existing unison to synchronise configuration across the nodes. simply by adding company folder as a space seperated list to sync_folder variable.

Unison configuration :

You could use the unison protocol and add /usr/local/nagios/etc/objects/$company as a folder to also be synchronised and bypass all this CMS requirements. You would need to move :

commands.cfg
admin.cfg
timeperiods.cfg
templates.cfg
to : /usr/local/nagios/etc/objects/$company and update the reference in nagios.cfg accordingly

Within the script update and move the sync_path variable below $company variable: SYNC_PATH=”/opt/nagios-sync /usr/local/nagios/etc/objects/$company”

———————————————————

so inside: files/company I have currently two data centres: in each of these data centres

files/company/datacentre1
– files/company/datacentre1/prod
— files/company/datacentre1/prod/hosts
— files/company/datacentre1/prod/services

and so on

on my nagios 3.5 servers I have defined the path to each datacentre:

Nagios server 1

# Datacentre1 servers – services
cfg_dir=/etc/nagios/objects/company/datacentre1

Nagios server 2

# Datacentre1 servers – services
cfg_dir=/etc/nagios/objects/company/datacentre2

Nagios will then read all the files within the sub folders of each datacentre recursivly on each nagios host… (this simplify definition of each config file etc and makes it a lot easier to script this solution.

So once you have puppet pushing out the configuration which in short all nagios hosts have all the configrations but only load up the relevant datacentre folder for what it is supposed to monitor:

unison protocol to synchronise amongst nagios hosts

yum install unison or apt-get install unison

A shared folder amongst all nagios servers

In the script I have defined /opt/nagios-sync

mkdir /opt/nagios-sync
mkdir /opt/nagios-sync/config_backup
touch /opt/nagios-sync/status.log
chown -R nagios:nagois /opt/nagios-sync

Nagios user with sudoers access to restart nagios and access files in /opt/nagois-sync sudo -i

visudo

nagios ALL = NOPASSWD: /etc/init.d/nagios

:wq

ssh-keygen and ssh-copy-id across all nagios hosts so nagios can ssh from any nagios host to any other without a password as the nagios user.

in /etc/nagios/nagios.cfg on all hosts you will need to add: under the main configuration loading up current data centres

# AUTOMATION ADD HERE

That should be all that is needed to get it going now refer to the script