Observability reference architecture
Prometheus metrics platform
Scalable Prometheus collection with long-term storage and recording rules for SLOs.
Design principles
- Scrape close to the source
- Recording rules for expensive queries
- Remote write for retention
- Alert rules in source control
Components
| Component | Role |
|---|---|
| Prometheus | Collection and alert evaluation |
| Alertmanager | Routing and silencing |
| Long-term store | Thanos, Mimir or managed service |
| Exporters | Metrics from infrastructure |
More Observability architectures
Dynatrace full-stack monitoring
OneAgent and Kubernetes operator deployment for end-to-end tracing and AI-assisted root cause analysis.
View architectureGrafana dashboards & SLOs
One visualization layer over metrics, logs and traces, with dashboards defined as code.
View architectureAmazon CloudWatch at scale
Cross-account CloudWatch observability with centralized dashboards, alarms and log analytics.
View architectureAdapt this to your environment
We'll help you apply this pattern to your constraints and existing estate.

