ingestion

Ingestion Service

The ingestion service is the gRPC entry point for test event collection. It receives test execution events from reporters and processes them.

Architecture

The ingestion service is designed to be stateless and horizontally scalable. In the target distributed architecture, it will:

1. Accept gRPC calls from reporters (Playwright, pytest, etc.)
2. Validate protobuf payloads
3. Publish validated events to NATS JetStream
4. Handle backpressure and transient errors

Current State

Currently, the ingestion service runs as a standalone gRPC server that can operate:

• Without database: Pure ingestion mode (future: will publish to NATS)
• With database: Direct persistence mode (backward compatible with monolithic setup)

Running

Standalone (no database)

./bin/ingestion
# or
make build-ingestion && ./bin/ingestion

Default port: 50051

Custom port

PORT=6000 ./bin/ingestion
# or
./bin/ingestion -port 6000

Environment Variables

Future Enhancements

• NATS JetStream publisher integration
• Metrics endpoint (Prometheus)
• Health check endpoint
• OpenTelemetry tracing
• TLS support