An extremely simple RESTful API for manipulating Nagios scheduled downtime.
Ruby Puppet
Permalink
Failed to load latest commit information.
lib Using Mocha without telling RSpec explicitly about it causes issues w… Nov 22, 2011
puppet/sinagios Add a Puppet module to deploy Sinagios. Nov 24, 2011
rpmfiles Incorrect operation on uninstall. Feb 23, 2012
spec Last set of tests for the client. Jan 1, 2012
.gitignore
.rvmrc
.travis.yml
COPYING Relicence under GPLv2+ Nov 10, 2011
Gemfile Bugs in RSpec 2.6 and needing to maintain compatibility with 1.8.5 me… Dec 29, 2011
Gemfile.lock Bugs in RSpec 2.6 and needing to maintain compatibility with 1.8.5 me… Dec 29, 2011
README.markdown Add config file support to the client and tests. Need to bump RSpec v… Dec 14, 2011
Rakefile Remove duplication between gemfile and rakefile Dec 28, 2011
sinagios.rb 201 is a more appropriate code for a successfully created entity. Nov 28, 2011
sinagios_client.rb More tests. Jan 1, 2012

README.markdown

Overview

Sinagios is a Sinatra app which provides an extremely basic RESTful API for manipulating Nagios service and host downtime.

Requirements

These are the versions I develop with. It will probably work more or less the same on similar, but not equal, major versions.

Mandatory

You need these to run Sinagios.

  • Nagios 2.x or later (see notes below)
  • ruby 1.8.5 or later
  • sinatra 1.2.7
  • rack 1.2.4
  • json 1.5.3
  • thin 1.2.11

Optional

These are used mainly for testing or packaging.

  • rake 0.8.7
  • rspec 2.6.0
  • rack-test 0.6.1
  • fpm (see notes below)

Nagios Version

In Nagios version 2.x the downtime data was maintained in a separate file to the status information. Sinagios presently only looks inside a single file to get information on downtime so if you are using Nagios 2.x please set the status_file entry in the configuration to point to your downtime file (usually something like /var/log/nagios/downtime.dat).

Installation

Assumptions

  • Apache httpd is installed and in the nagios group.
  • If you want security, authentication etc you will proxypass sinagios through httpd (or your web frontend of choice).
  • Nagios is installed and running and the external command interface is enabled.

Manual

  • git clone this repository
  • install the above mandatory gems
  • mkdir /etc/sinagios and copy rpmfiles/{sinagios.conf,config.ru} into it
  • customise /etc/sinagios/sinagios.conf to match your command and status file locations
  • start with 'rake rackup'

Preferred

I only have methods implemented for building RPM packages. Sorry - don't hate me.

Right now fpm (used for packaging) does not have the required features in the mainline code tree (specifically, using a filelist and being able to specify configuration files). You can take a look at my fork and merge the patches from the config_files and inputs_file branches to get a working version. Or just bug me for an RPM.

  • Build the sinagios package using 'rake package'
  • Build the dependent gems using 'rake package_gems'
  • Install the sinagios package and mandatory gem packages.
  • Start sinagios using 'service sinagios start'
  • Optionally use the Puppet module to deploy and maintain the application.

Configuration

Sinagios will determine the location of the Nagios files from:

  • Anything passed to Nagios#new (code internals)
  • Config file specified by environment variable SINAGIOS_CONFIG
  • Config file /etc/sinagios/sinagios.conf
  • Defaults:
    • Command file: /var/spool/nagios/cmd/nagios.cmd
    • Status file: /var/log/nagios/status.dat

If you are using the configuration file in /etc/sinagios/sinagios.conf or a location provided by the environment variable SINAGIOS_CONFIG you can override either or both of these parameters in the YAML format like the following:


cmd_file: /path/to/nagios.cmd status_file: /path/to/status.dat

Please see the notes in the Requirements section about the location of the status file depending on your version of Nagios.

Operations

  • All structured output is in JSON.
  • The API is versioned in case of future interface changes.

You can perform basically three operations:

List hosts and all current downtime IDs

GET /v1/downtime

Delete all downtime for a host

DELETE /v1/downtime/:name

Schedule downtime for a host and all its services

POST /v1/downtime/:name

Parameters (all are required):

  • duration (downtime window expressed in whole seconds)
  • author (name of downtime author, matching [^a-zA-Z0-9-.#\s])
  • comment (a comment for the downtime, matching [^a-zA-Z0-9-.#\s])

Health Check

GET /v1/health

This bonus fourth operation facilitates some basic health checking of the app. It will just ensure it has correct access to the status and command files and return an HTTP 200 and body text of 'OK' if everything is well.

If not, it will attempt to return the exception message and an HTTP 500 code.

Roadmap

Aside from tidy ups, effectively this 1.0.0 version is feature complete. If there are features you'd like to see implemented, please file an issue.

Continuous Integration

You can see the current build status for Sinagios here:

Copyright

This program and libraries are copyright 2011 Oliver Hookins, and licensed under the GNU General Public License Version 2 or higher.