Observability dashboards

Monozu deploys two Azure Monitor dashboards with Grafana in rg-monozu-shared-{env}:

Dashboard	Name	Focus
Application & database	`dash-mz-app-{env}`	HTTP, worker jobs, tenants, SQL, errors, logs
Infrastructure	`dash-mz-infra-{env}`	Container Apps, Azure SQL, jobs queue

Legacy dash-monozu-obs-{env} remains until manually removed from Azure after you accept the new dashboards.

Telemetry path

See Telemetry architecture for the full pipeline: App → OTLP → central Collector → four Application Insights components → LAW → Grafana.

flowchart LR
    ACA[Container Apps] --> Collector[OTel Collector]
    Collector --> AI[Application Insights]
    AI --> LAW[Log Analytics]
    LAW --> App[dash-mz-app-env]
    LAW --> Infra[dash-mz-infra-env]
    ACA -->|Azure Monitor Metrics| Infra
    Collector -->|Azure Monitor Metrics| Infra

RBAC

Monitoring Reader on:

rg-monozu-shared-{env}
rg-monozu-cloud-{env}
rg-monozu-ingest-{env}
rg-monozu-manage-{env}

This is Azure RBAC for portal dashboards — not tenant JWT permissions inside the product.

Service names

Container App	`AppRoleName` in LAW
Cloud API	`monozu-cloud-backend`
Worker	`monozu-cloud-worker`
Ingestor	`monozu-ingestor`
Management	`monozu-management-backend`
OTel Collector	(no App Insights role — routes only)

Edit and deploy

JSON sources: Infrastructure/observability/grafana/monozu-app.dashboard.json and monozu-infra.dashboard.json
Validate: python3 Infrastructure/observability/grafana/validate_dashboards.py <paths>
Deploy infrastructure Bicep — app dashboard from shared.bicep, infra dashboard from main.bicep (after apps)

Troubleshooting

See Observability troubleshooting for empty panels, sampling, datasource UID, and queue metric freshness.