Platform checks

Introduction

Allow to perform checks on a PaaS.

This is built for the Nasqueron servers infrastructure, and will be useful on modern platform installations like microservices, Docker and proxies/load balancers.

Each check is compatible Nagios / NRPE.

Configuration

Checks can be defined in /etc/monitoring/checks.yml.

The format of this file is:

checks:
  <check_type>:
    <key>: <value>

You can also use /usr/local/etc/monitoring/checks.yml as path, or .checks.yml in the current working directory.

Checks reference

check_http_200

Perform a HTTP request to a specific URL. The check success if HTTP code is 200.

Available check types

• check_http_200: for sites you test at the back-end level
• check_http_200_alive: for sites returning "ALIVE" as body content
• check_http_200_proxy: flag site as verified at proxy level, not directly at the back-end
• check_http_200_alive_proxy: combine proxy and alive flags

Configuration example

checks:
  check_http_200:
    acme: http://localhost:41080/health
    cachet: http://localhost:39080/api/v1/ping
    
  check_http_200_proxy:
    openfire: https://xmpp.nasqueron.org/login.jsp
    
  check_http_200_alive_proxy:
    phabricator: https://devcentral.nasqueron.org/status

Run the check

From there, you've two ways to run the check:

• one check for all URLs: call check_http_200 without argument
• one check, one URL: call check_http_200 <service name>, for example check_http_200 openfire

You can also create a check calling check_http_200 without argument, and it will test every site.

check_container_present

Run the check

With argument, check if the specified Docker container is running: check_container_present foo

Without argument, compare the list of containers present with the expected one.

The configuration is only required if you use it without argument.

Available check types

• check_docker_containers: a list of expected containers

Configuration example

checks:
  check_docker_containers:
    - foo
    - bar

Requirements

A Docker engine with CLI restructured, ie Docker 1.13+, is needed:

• To list the containers it uses docker container ls.
• To get more info on a container down, it uses docker container inspect.

Not features

This check isn't intended to detect containers run with other engines like systemd-nspawn.

check_software_version

Run the check

With argument, check if the specified software is up-to-date.

Without argument, get a list of software from the configuration.

If the software can't be found, an UNKNOWN exit code is returned.

Available check types

• check_software_version: a list of software to keep up-to-date

Configuration example

checks:
  check_software_version:
    - consul
    - nomad
    - terraform
    - vagrant
    - vault

Supported software

• HashiCorp CLI products with a "version" subcommand, like Consul, Vault, Vagrant or TerraForm.

Return values of checks

The checks use the standard Nagios/NRPE exit codes:

Exit code	Description
0	OK
1	Warning
2	Critical
3	Unknown, for example a check config issue

As such they're compatible with Nagios, Incinga, Sensu, Shinken, etc.