Adagios installation on NethServer 7.x

nrauso · December 15, 2016, 2:53pm

Hi all!
Good news are on the way: you can now run Adagios even on NS 7!

Just a couple of notes about this procedure:

You have to use Nagios 3.5.1 core yet (we’ll take the rpm from one of these mirrors): that’s for a known limitation of Adagios mk-liveststatus implementation (take a look here);
The procedure is still experimental so there’s a chance it will change!

Now let’s go on!

Source:
https://github.com/opinkerfi/adagios/wiki/Adagios-on-CentOS-7-with-Nagios-4 (notice the “Under construction, not ready yet” in the title! ;))

Install additional repo:

yum localinstall http://opensource.is/repo/ok-release.rpm -y
Change releasever ref into ok repo url (mandatory or installation fails):

sed -i ‘s/rhel$releasever/rhel7/’ /etc/yum.repos.d/ok.repo
Procede with installation:

yum -y --enablerepo=ok* install pnp4nagios nagios-plugins*
git adagios okconfig acl python-setuptools mk-livestatus pynag
ftp://bo.mirror.garr.it/1/centos/7.3.1611/cloud/x86_64/openstack-newton/common/nagios-3.5.1-1.el7.x86_64.rpm
ftp://bo.mirror.garr.it/1/centos/7.3.1611/cloud/x86_64/openstack-newton/common/nagios-common-3.5.1-1.el7.x86_64.rpm
Don’t want Nagios to log into syslog (too verbose!), so disable it:

sed -i ‘s/use_syslog=1/use_syslog=0/’ /etc/nagios/nagios.cfg
Fit localhost check_http command to NS Apache configuration:

touch /var/www/html/apacheok
sed -i ‘s/check_http/check_http!-u /apacheok/’ /etc/nagios/objects/localhost.cfg
Choose nagiosadmin password:

htpasswd -c /etc/nagios/passwd nagiosadmin
Make sure Adagios can write to nagios configuration files, and that it is a valid git repo so we have audit trail:

cd /etc/nagios/
git init
git config user.name "yourname"
git config user.email "yourname@yourdomain.com"
git add *
git commit -m “Initial commit”
Make sure Nagios group will always have write access to the configuration files:

chown -R nagios /etc/nagios/* /etc/nagios/.git
setfacl -R -m group:nagios:rwx /etc/nagios/
setfacl -R -m d:group:nagios:rwx /etc/nagios/
By default objects created by Adagios will go to /etc/nagios/adagios so make sure that this directory exists and nagios.cfg contains a reference to this directory:

mkdir -p /etc/nagios/adagios
pynag config --append cfg_dir=/etc/nagios/adagios
The status view relies on broker modules livestatus and pnp4nagios, so let’s configure nagios.cfg to use those:

pynag config --append "broker_module=/usr/lib64/nagios/brokers/npcdmod.o config_file=/etc/pnp4nagios/npcd.cfg"
pynag config --append "broker_module=/usr/lib64/mk-livestatus/livestatus.o /var/spool/nagios/cmd/livestatus"
pynag config --set “process_performance_data=1”
Add nagios to apache group so it has permissions to pnp4nagios session files:

usermod -G apache nagios
Define new services according to NethServer default (see developer manual):

config set nagios service status enabled
mkdir -p /etc/e-smith/db/configuration/defaults/nagios
echo “service” > /etc/e-smith/db/configuration/defaults/nagios/type
echo “enabled” > /etc/e-smith/db/configuration/defaults/nagios/status

config set npcd service status enabled
mkdir -p /etc/e-smith/db/configuration/defaults/npcd
echo “service” > /etc/e-smith/db/configuration/defaults/npcd/type
echo “enabled” > /etc/e-smith/db/configuration/defaults/npcd/status

config set nrpe service access private TCPPort 5666 status enabled
mkdir -p /etc/e-smith/db/configuration/defaults/nrpe
echo “service” > /etc/e-smith/db/configuration/defaults/nrpe/type
echo “enabled” > /etc/e-smith/db/configuration/defaults/nrpe/status
echo “private” > /etc/e-smith/db/configuration/defaults/nrpe/access
echo “5666” > /etc/e-smith/db/configuration/defaults/nrpe/TCPPort
We need to fix a known bug (see here):

sed -i “3iALLOWED_HOSTS = [’*’]\n” /etc/adagios/adagios.conf
Enable, start and signal the new services:

systemctl enable nagios
systemctl enable npcd
systemctl enable nrpe

systemctl start nagios
systemctl start npcd
systemctl start nrpe
systemctl restart httpd

signal-event runlevel-adjust
Add the following line at the end of Adagios sudoers fragment file:

cat << EOT > /etc/sudoers.d/nagios-adagios

Nagios daemon control on NethServer 7

Defaults:nagios !requiretty
nagios ALL = (root) NOPASSWD: /usr/bin/systemctl
EOT
Fix permission in logrotate fragment:

sed -i “2isu nagios nagios” /etc/logrotate.d/okconfig
To prevent unforseen nagios package update, add:

mkdir -p /etc/e-smith/templates-custom/etc/yum.conf
echo -e “#\n# 25exclusions\n#\nexclude=nagios nagios-common\n” > /etc/e-smith/templates-custom/etc/yum.conf/25exclude
expand-template /etc/yum.conf

Done! Now you can access Adagios form the default url:

https://server.domain/adagios

using nagiosadmin credentials.

Test it, test it, test it and let me know!

alefattorini · December 15, 2016, 3:22pm

Guess that @mario_spang @ironsky @zapdog @Artem_Voylenko @edi @tramamo @m.traeumner @vores @jaapvdv would be more than happy to help you

Anyway Kudos to @nrauso great shot!

robb · December 15, 2016, 10:31pm

great work @nrauso!!
this is absolutely a more advanced install. not the default create database + dbuser, create apache vhost, download the applucation and unpack in vhost directory, start the installer, fill in some basic stuff and done.

can you explain why adagios uses a git-like local repo?

nrauso · December 16, 2016, 8:48am

Hi @robb!
Happy you like the procedure: I’ll work to make it better!

Concerning Git, Adagios uses it only to keep track of configuration changes: it allows you to log any activity and easily revert any change (you could find the panel to manage git commits under the cog icon on the top-right corner of the gui -> “Object History” menu).

Actually Git configuration is optional.

I could write it down on the procedure, what do you think?

m.traeumner · December 16, 2016, 9:32am

Hi Nicola,
thanks for your great job. I have to set up a new virtual mashine to test it. I’ll try to do it next week.

nrauso · December 16, 2016, 10:06am

Wow!
Thanks @m.traeumner: let me know!

giacomo · December 16, 2016, 2:54pm

Great news and excellent howto!

m.traeumner · December 20, 2016, 1:23pm

Hi Nicola,
the installation went fine. After configuring the ssh-request (no standard ssh port) it works without any errors.
Thanks for your great work.

m.traeumner · December 20, 2016, 2:41pm

Hi to all,
if you have installed nagios and adagios on Nethser 7 with this tutorial and you don’t have the standard ssh port it shows a critical error.

To get rid of it you only have to add an argument with the portnumber and restart the service.

go to the adagios webinterface
choose Configure at the upper menu
choose Configuration at the left side
by Object Configuration switch from Host to Services
click on ssh and choose monitoring
by Argument 1 you can add your port option like “-p 2222” without quotation mark
after that I had to restart nagios from shell “systemctl restart nagios”

I think for nagios on Nethserver 6.8 it’s the same, perhaps somebody can try it.

nrauso · December 20, 2016, 2:46pm

Nice shot @m.traeumner!
And yes, on NS 6.8 it’s the same because it’s the standard ssh check config.

m.traeumner · February 2, 2017, 1:53pm

After update to 7.4 and now final I can’t use nagios anymore.

[root@GroupwareBackup nagios]# systemctl status nagios -l
● nagios.service - Nagios Network Monitoring
Loaded: loaded (/usr/lib/systemd/system/nagios.service; enabled; vendor prese t: disabled)
Active: failed (Result: exit-code) since Thu 2017-02-02 14:45:27 CET; 1min 5s ago
Docs: Documentation | Nagios Open Source
Process: 26089 ExecStartPre=/usr/sbin/nagios -v /etc/nagios/nagios.cfg (code=e xited, status=1/FAILURE)

Feb 02 14:45:27 GroupwareBackup.jonas.de systemd[1]: Starting Nagios Network Mon itoring…
Feb 02 14:45:27 GroupwareBackup.jonas.de systemd[1]: nagios.service: control pro cess exited, code=exited status=1
Feb 02 14:45:27 GroupwareBackup.jonas.de systemd[1]: Failed to start Nagios Netw ork Monitoring.
Feb 02 14:45:27 GroupwareBackup.jonas.de systemd[1]: Unit nagios.service entered failed state.
Feb 02 14:45:27 GroupwareBackup.jonas.de systemd[1]: nagios.service failed.
[root@GroupwareBackup nagios]#

nrauso · February 3, 2017, 8:08am

Oh my!
I’ll take a look at it asap!

nrauso · February 7, 2017, 10:22am

Hi @m.traeumner!
I ran some test but haven’t been able to reproduce your problem.
Could be a matter of configuration?
Could you post here the output of the command:

nagios -v /etc/nagios/nagios.cfg

and attach /etc/adagios/adagios.conf file?

m.traeumner · February 7, 2017, 1:41pm

Hi Nicola,
thanks for your test. I’ve forgotton that I’ve changed something. I have tried to install a nagios win client at my win pc and bind it with the nagios at nethserver. I’ve done it with these steps.
After renaming /etc/nagios/okconfig/PC20141103-01.cfg PC20141103-01.bak

nagios -v /etc/nagios/nagios.cfg

runs without errors:

Nagios Core 4.0.8
Copyright (c) 2009-present Nagios Core Development Team and Community Contributors
Copyright (c) 1999-2009 Ethan Galstad
Last Modified: 08-12-2014
License: GPL

Website: http://www.nagios.org
Reading configuration data…
Read main config file okay…
Read object config files okay…

Running pre-flight check on configuration data…

Checking objects…
Checked 8 services.
Checked 1 hosts.
Checked 2 host groups.
Checked 1 service groups.
Checked 1 contacts.
Checked 2 contact groups.
Checked 189 commands.
Checked 5 time periods.
Checked 0 host escalations.
Checked 0 service escalations.
Checking for circular paths…
Checked 1 hosts
Checked 0 service dependencies
Checked 0 host dependencies
Checked 5 timeperiods
Checking global event handlers…
Checking obsessive compulsive processor commands…
Checking misc settings…

Total Warnings: 0
Total Errors: 0

Things look okay - No serious problems were detected during the pre-flight check

I think the second hostgroup is my problem.

Here is my

/etc/adagios/adagios.conf

file:

# -- coding: utf-8 --

# nagios_config - Path to your nagios configuration file
# If set, adagios will use this file to manage your object
# definitions. If set to None, adagios will search most common
# paths like /etc/nagios/nagios.cfg for it
#nagios_config = “/etc/nagios/nagios.cfg”
nagios_config = “/etc/nagios/nagios.cfg”

# This should be the URL to your nagios server. If set, then
# Adagios can link you directly from configuration to live object.
# If you don’t need this feature. Set nagios_url to ‘’ or None
#nagios_url = “http://mynagiosserver/nagios”
nagios_url = “/nagios”

# destination_directory - Where adagios will store new objects
# by default. You should make sure that this directory exists
# and nagios.cfg has a cfg_dir=/etc/nagios/adagios/
destination_directory = “/etc/nagios/adagios/”

# livestatus_path - Path to livestatus socket
livestatus_path = None
# enable_githandler - If set to true, and your /etc/nagios/ directory
# is a git repository. adagios will automatically commit changes when
# they are made.
enable_githandler = True

# enable_loghandler - If set to true, all changes adagios makes to
# object definitions will be logged to a file.
enable_loghandler = False

# auto_reload - If set to True, then nagios is reloaded automatically on
# changes. Not really recommended unless you are sure you want to fire
# a nagios reload after every single change.
auto_reload = False

# nagios_service - Service name used to start/stop nagios
# Adagios uses this to reload nagios after configuration changes
nagios_service = “nagios”

# nagios_init_script - Path to init script used to start/stop nagios
# Adagios uses this to reload nagios after configuration changes
nagios_init_script=“”

# nagios_binary - Path to your nagios binary
# Adagios uses this to run “nagios -v /etc/nagios/nagios.cfg”
nagios_binary=“/usr/sbin/nagios”

# escape_html_tags - Escape html tags in status view
# If this is enabled, html code in plugin outputs is escaped.
# If you set this to False, then plugins will be able to inject
# HTML code into adagios web interface
escape_html_tags = True

# warn_if_selinux_is_active - Check (and warn user) if selinux is running
# You should only turn this off if you have configured your selinux policy
# and know what you are doing
warn_if_selinux_is_active=True

# pnp_filepath - Path to pnp4nagios index.php
# This is used for integrated graphs into the status view
# if you do not use pnp or do not need the status view of adagios
# simply change this to any other file that exists
pnp_filepath=“/usr/share/nagios/html/pnp4nagios/index.php”

# contrib_dir - Path to user contributed views
contrib_dir = “/var/lib/adagios/contrib/”

# serverside_includes - Path to user contributed includes
# The files located in this directory can be used to include
# user contributed html into the header and footer of every page respectively
serverside_includes = “/etc/adagios/ssi”

# title_prefix - Title of every page is prefixed with this.
# You can customize this if you have multiple adagios instances
# and you want to distinguish the tabs in browser.
title_prefix = “Adagios -”

# enable_pnp4nagios - Enable pnp4nagios support (True by default).
# Disabling this removes the pnp mouse-over graphs as well as the graphs
# tab from the status views.
enable_pnp4nagios = True

# enable_graphite - Enable graphite integration with graphite
# If set to True, status_detail view of hosts and services
# Will show related graphs.
enable_graphite = False

# graphite_url - URL to your graphite instance
# If specified, and enable_graphite is set to True,
# connect to this graphite instance
graphite_url = “http://localhost:9091”

# graphite_querystring - Customize the querystring that is sent to
# graphite’s /render? method. {host} {service} and {metric} will all
# be replaced with actual values while {host_} {service_} and {metric_}
# will apply common escaping techniques, i.e.: example.com becomes example_com
graphite_querystring = “target={host_}.{service_}.{metric_}&width=500&height=200&from={from_}d&lineMode=connected&title={title}&target={host_}.{service_}.{metric_}warn&target={host}.{service_}.{metric_}_crit”

# graphite_title - Put this string at the top of graphite graphs
graphite_title = “{host} - {service} - {metric}”

# Include configuration options from these config files
include=“/etc/adagios/conf.d/*.conf”

# Automaticly generated secret_key
django_secret_key = ‘fjh3ladr!hdq1kekebsps%(-lp!624eoa3q477xzxmqoh2pnyx’

nrauso · February 7, 2017, 3:10pm

Hi @m.traeumner!
That’s the culprit:

Nagios Core 4.0.8

As said in the how-to:

You have to use Nagios 3.5.1 core yet (we’ll take the rpm from one of these mirrors): that’s for a known limitation of Adagios mk-liveststatus implementation

The mk-livestatus error is generated by a known incompatibility with the Nagios version you adopted.

m.traeumner · February 8, 2017, 12:12pm

I can’t remind that I’ve done an update but now I downgraded it

yum localinstall http://opensource.is/repo/ok-release.rpm -y

sed -i ‘s/rhel$releasever/rhel7/’ /etc/yum.repos.d/ok.repo

yum downgrade pnp4nagios nagios-plugins*
git adagios okconfig acl python-setuptools mk-livestatus
ftp://bo.mirror.garr.it/1/centos/7.3.1611/cloud/x86_64/openstack-newton/common/nagios-3.5.1-1.el7.x86_64.rpm

restored my nagios.cfg backup and it works fine again.

Thanks for your support

m.traeumner · February 8, 2017, 12:27pm

Know I now where nagios 4.0.8 comes from, after downgrading to 3.5.1 the Software-Center shows the update to 4.0.8. Is there a way to block it without installing every update in the terminal?

giacomo · February 11, 2017, 6:37pm

Yes, take a look at this:

nrauso · February 16, 2017, 6:01pm

Hi @m.traeumner!

I added some how-to stuff to prevent nagios package update: check it out!

m.traeumner · February 17, 2017, 8:28am

Hi Nicola and Giacomo,
thanks for your answers, the exclusion of nagios update works. I did it with manually editing the template
nano /etc/e-smith… and after deleting it I tried with Nicolas way. Both works fine.
@nrauso
Did you change anymore in your howto?

Adagios installation on NethServer 7.x

Nagios daemon control on NethServer 7