Observability in Identity Platform

The Canonical Identity Platform provides integration with the Canonical Observability Stack (COS). This integration ensures that administrators can monitor the health, performance, and security of the identity infrastructure through standardized metrics, automated alerting, and pre-configured dashboards.

Key Observability Features

The observability suite across all Identity Platform charms focuses on three core pillars:

• Health Monitoring (Prometheus): Tracking unit availability (the up metric) to ensure the service is reachable.

• Log Aggregation (Loki): Centralizing logs to identify high-severity events (Error, Critical, Fatal) across the stack.

• Visualization (Grafana): Providing out-of-the-box dashboards to visualize traffic patterns, error rates, and system uptime.

Component Reference Guides

Each component of the Identity Platform provides its own set of specific metrics and alerts. Explore the detailed technical references for each charm below:

• Observability setup in Charmed Hydra
• Observability setup in Charmed Kratos
• Observability setup in Charmed Identity Platform Login UI
• Observability setup in Charmed OpenFGA

Common Alert Severities

Across the platform, alerts are categorized to help your operations team prioritize responses:

• Fatal: Total loss of service; requires immediate intervention.

• Critical: Service is active but at high risk of failure (e.g., only one unit remaining).

• Error: Specific application functions are failing, but the service remains partially available.

• Warning: Anomalies detected (e.g., low-frequency high-severity logs) that do not yet impact service.

Next Steps

To begin monitoring your deployment, ensure your charms are integrated with the COS bundle:

juju integrate <charm-name>:metrics-endpoint prometheus:metrics-endpoint
juju integrate <charm-name>:grafana-dashboard grafana:grafana-dashboard
juju integrate <charm-name>:log-proxy loki:logging