Prometheus
Metrics monitoring service based on Prometheus, Grafana and several agents.
graph TD
P[fa:fa-fire-alt Prometheus] -->|pull| G(fa:fa-cubes gitea)
P -->|pull| C(fa:fa-cubes Confluence)
P -->|pull| NC(fa:fa-cubes Nextcloud)
P -->|pull| NE(fa:fa-cubes Node exporter)
P -->|pull| PG{fa:fa-compress-arrows-alt push gateway}
subgraph per PAC push scripts
FS(fa:fa-terminal directory tree usage) -->|push| PG
N(fa:fa-terminal network traffic) -->|push| PG
Q(fa:fa-terminal filesystem quota) -->|push| PG
end
Agents
- Node exporter: provides Machine and OS specific metrics
- Nextcloud
- Gitea
- Confluence
- Push Gateway: supplied by a handful of self developed providers as shell scripts
Authentication
Grafana is coupled on the LDAP service. It knows of three groups of users:
- admins: members of the
admin_grafanaLDAP group - editors: members of the
ECG_HUB_it - guests (without access at all): everybody else
Guests are excluded due to security concerns, because the metrics allow a detailed view on the ECG services performance data.