Server configuration

The server is configured through appsettings.json or environment variables. In an environment variable, : becomes __, so Auth:ReadToken is Auth__ReadToken.

1. Database
2. Authentication
3. Retention
4. Alerts
5. Issue trackers
6. Observability of the server itself
7. Demo mode
8. Dashboard

Database

"Database": { "Provider": "Sqlite" },     // Sqlite | PostgreSql | SqlServer
"ConnectionStrings": { "Events": "Data Source=avizoro.db" }

• SQLite is the default and needs no setup. The folder of the database file is created if it’s missing.
• PostgreSQL and SQL Server need ConnectionStrings:Events. For PostgreSQL, a postgres://user:pass@host:5432/db URL, as handed out by Render, Railway or Heroku, is also accepted, and so is DATABASE_URL.
• The schema is migrated on startup.

Authentication

"Auth": { "ReadToken": "avizoro_admin_<long random string>" }

• The admin token can read everything and manage projects, alerts and symbols. Send it as Authorization: Bearer <token>.
• Each project also has a read key (created or rotated on the Projects page) that only sees that project.
• With no token set, reads are open in Development and blocked in Production.
• Ingest (POST /events) is open by design: the project key in the DSN is public. By default only registered projects are accepted:

"Ingest": { "RequireKnownProject": true },
"RateLimit": { "Ingest": { "Enabled": true, "PermitLimit": 300, "WindowSeconds": 60 } }

Rate limiting is per client IP. A throttled SDK gets 429 and retries later, so no events are lost.

Run the server behind HTTPS in production.

Retention

Off by default, so a new install never deletes anything:

"Retention": { "Enabled": true, "Days": 90, "SweepHours": 6 }

Alerts

"Alerts": {
  "DashboardUrl": "https://avizoro.example.com",          // for "open issue" links
  "Smtp": { "Host": "smtp.example.com", "Port": 587, "EnableSsl": true,
            "User": "", "Password": "", "From": "alerts@example.com" }
}

Rules are created in the dashboard. SMTP is only needed for email rules. See Alerts.

Issue trackers

"Integrations": {
  "GitHub": { "Repository": "owner/repo", "Token": "" },
  "Jira": { "BaseUrl": "https://you.atlassian.net", "Email": "", "ApiToken": "",
            "ProjectKey": "APP", "IssueType": "Bug" }
}

• GitHub with only Repository set: the dashboard opens a prefilled new-issue form. Add a fine-grained token with Issues: write to create the issue directly.
• Jira Cloud needs all five values.

Observability of the server itself

"Telemetry": {
  "Prometheus": { "Enabled": true, "RequireAuth": true },
  "Otlp": { "Endpoint": "http://otel-collector:4317" }
}

• GET /health/live: the process is up.
• GET /health/ready: the database can be reached. Use this for load balancer and orchestrator probes.
• GET /metrics: Prometheus metrics: events ingested and rejected, ingest duration, issue changes, session updates, spans, and alerts delivered (avizoro.*), plus ASP.NET Core and .NET runtime metrics. It needs the admin token unless RequireAuth is false.
• With Otlp:Endpoint set, metrics and traces are also pushed over OTLP.

Demo mode

"Demo": { "Seed": true, "ReadOnly": true }

• Seed fills an empty database with two weeks of sample data for a project called demo-shop.
• ReadOnly rejects every write, including ingest, with 403, so the instance can be shared publicly with a known token.

Dashboard

The dashboard is a static Blazor WebAssembly app. It needs to know where the API is:

• From source: wwwroot/appsettings.json → ApiBaseUrl.
• Docker image: the API_BASE_URL environment variable, the server URL as the browser reaches it.

The server allows any origin in CORS by default, so the dashboard and Blazor apps can be served from another host.