The ansible-runner-service process uses https, so to there are a couple of additional options needed to get Prometheus to scrape the data.
Here's an example (used during testing);
- job_name: "ansible-runner-service"
scrape_interval: 5s
scheme: https
tls_config:
insecure_skip_verify: true
static_configs:
- targets: ["rh460p:5001"]
obviously change the target!
This directory has an example dashboard called ansible-runner-service-metrics.json
which illustrates the kind of insights the
/metrics endpoint can provide. Import it into you grafana instance and you should see something like;
The dashboard uses a template variable to represent the host so you could use the same dashboard to report against multiple instances of the ansible-runner-service. As you can see above, the dashboard also provides CPU busy information. To see this data, you need to have node_exporter (0.16 or above) installed too.