This site is deprecated in favor of demo.prometheus.io. Repository with current code is available at github.com/prometheus/demo-site.
All code stored here worked flawlessly as of 11 Jan 2020 and is left for historical purposes.
This repository provides an integration testing suite for our ansible roles as well as a demo site for grafana, prometheus, alertmanager and node_exporter (possibly more in the future).
Site is provisioned with ansible running every day and on almost all commits to master branch. Everything is fully automated with travis ci pipeline. If you want to check ansible-playbook output, go to last build.
Have a look at the configuration file group_vars/all/vars.
All applications should be running on their default ports.
| App name | Address (HTTP) | Address (HTTPS) | Status | Uptime |
|---|---|---|---|---|
| node_exporter | demo.cloudalchemy.org:9100 | node.demo.cloudalchemy.org |  |
 |
| snmp_exporter | demo.cloudalchemy.org:9116 | snmp.demo.cloudalchemy.org |  |
 |
| blackbox_exporter | demo.cloudalchemy.org:9115 | blackbox.demo.cloudalchemy.org |  |
 |
| prometheus | demo.cloudalchemy.org:9090 | prometheus.demo.cloudalchemy.org |  |
 |
| alertmanager | demo.cloudalchemy.org:9093 | alertmanager.demo.cloudalchemy.org |  |
 |
| grafana | demo.cloudalchemy.org:3000 | grafana.demo.cloudalchemy.org |  |
 |
This repository consists of two playbooks:
- site.yml - which deploys basic prometheus/grafana stack without additional http proxies and with software listening on default ports
- extras.yml - installs python, caddy http proxy, and setups random_exporter. This will allow HTTPS connections to services like prometheus. Before running it, golang is required to be installed on deployer machine.
Such setup causes that mose of services can be accessed in two ways. As an example, prometheus can be accessed via:
- http://demo.cloudalchemy.org:9090 - default way
- https://prometheus.demo.cloudalchemy.org - workaround which in backgroud communicates with prometheus via insecure, "default" channel mentioned above
This workaround was needed to solve issue #13 and still provide a playbook which could be used by everyone - site.yml.
You can easily run such setup yourself without much knowledge how any part of this works. You just need to do two things:
First of all you need to configure your inventory, ours is located in hosts file. Here you set up your target hosts by changing value of ansible_host variable. Also here you can exclude parts of this demo site, so if you don't need our website, you just remove this part:
[web]
demo
Accordingly you can exclude grafana, prometheus, or influxdb.
For security measures we encrypted some of our passwords, but it is easy to use yours! You can do it by replacing a file located at group_vars/all/vault with following content:
vault_grafana_password: <<INSERT_YOUR_GRAFANA_PASSWORD>>
vault_influxdb_password: <<INSERT_YOUR_INFLUXDB_PASSWORD>>
You need to specify both even if you don't use grafana nor influxdb. You can look over group_vars/all/vault to find why.
# Download roles
ansible-galaxy install -r roles/requirements.yml
# Run playbook
ansible-playbook site.yml
# or when using vault encrypted variables
ansible-playbook --vault-id @prompt site.yml