Monitor Ansible Automation Platform using Prometheus, Node Exporter, and Grafana

Ansible Automation Platform integration

Now we are going to integrate Ansible Automation Platform with Prometheus. With this integration, Prometheus will consume the API metrics context /api/v2/metrics. For this, we will create a user in Ansible Automation Platform who will be responsible for generating the token and providing this integration.

1. In Ansible Automation Platform , in the left side menu, go to Access → Users → Add.
2. Define a name and username that facilitates identification; here we will use aap-metrics.
3. Set a password and confirm.

4. In User Type, select System Auditor and Save (Figure 3). 

   

5. Now log out and connect with the user you just created. Go to Access → Users → click on the username → Tokens → Add.
6. Set a Description and Scope to Read.

7. Save and copy the generated token (Figure 4). 

   

Now on the monitoring server, let's add a new Job to prometheus.yml, modify the bearer_token generated by the aap-metrics user, and add the Ansible Automation Platform endpoint in targets:

cat <<EOF >> /etc/prometheus/prometheus.yml

  - job_name: 'aap-controller'
    # api metrics path
    metrics_path: /api/v2/metrics
    scrape_interval: 5s
    scheme: https
    tls_config:  
      insecure_skip_verify: true
    # aap-metrics user token  
    bearer_token: xxxxxxxxxxxxxxxxx 
    static_configs:
    # AAP Endpoint
    - targets: ['aap.lab.laraujo.com.br']      
EOF


# Restart the prometheus service
systemctl restart prometheus

To validate your jobs, go to Status → Targets in the Prometheus console. You should have 3 targets created, as shown in Figure 5. 

Grafana dashboard

Now let's import the dashboard into Grafana and start monitoring our Ansible Automation Platform. Download and save aap-metrics.json file in the provisioning directory:

# access the proviosining/dashboards directory
cd /etc/grafana/provisioning/dashboards/

# create a directory called AAP
mkdir -v AAP

# download
curl -L https://raw.githubusercontent.com/leoaaraujo/articles/master/aap-nodeexporter-grafana/files/aap-metrics.json -o AAP/aap-metrics.json

# change owner
chown :grafana AAP/aap-metrics.json

Now let's create an aap-metrics.yaml file that will allow the automatic provisioning of our dashboard in Grafana:

cat <<EOF > /etc/grafana/provisioning/dashboards/aap-metrics.yaml
apiVersion: 1
providers:
  - name: AAP-Metrics
    folder: AAP
    type: file
    options:
      # Path where our aap-metrics.json file is saved
      path:
        /etc/grafana/provisioning/dashboards/AAP
EOF

# change owner
chown :grafana /etc/grafana/provisioning/dashboards/aap-metrics.yaml

# restart the grafana-server service
systemctl restart grafana-server.service

To validate that our dashboard was provisioned correctly, in the Grafana console, click Dashboards in the left-hand menu. We will have our dashboard provisioned in a folder called AAP (Figure 6). 

To view the dashboard, click on the dashboard name (Figure 7 and 8).

Conclusion

The Prometheus, Node Exporter, and Grafana stack offers a complete view for monitoring Ansible Automation Platform. Using the integration with the Ansible Automation Platform API, you can have a view of the day-to-day use of resources such as jobs, templates, and failed or successful executions. Node Exporter lets you create a visualization of capacity and consumption of computational resources.