prometheus: use 1 as 'success' value, not 0

related to: #301 (comment)

Because:
* The Base Prometheus scraper used the `systemctl` error code to
  indicate if a service is active: `0` == OK; otherwise NOT OK
* This is not intuitive, especially when extending the monitoring to
  other status, like the network connectivity in this pull request.
* It's better to use `1` as OK and `0` as NOT OK, which also makes
  visualization in Grafana easier.

This commit:
* changes logged values for network connectivity, `1` meaning OK
* changes logged values for systemd units, `1` meaning OK
* adjusts Grafana dashboard to use new values
* extends Grafana dashboard with diskspace usage over time
  (long overdue)