Files

Run linters on applied template / Python 3.13 lint and build (push) Successful in 1m40s

Details

Changes:
- put ObservabilityMiddleware before ExceptionHandlerMiddleware to avoid repetative code
- add application startup and last metrics update metrics along with CPU usage metric and threads count
- move host and port to new uvicorn section at config along with new reload and forwarded_allow_ips
- add request_id and remove trace_id/span_id generation if tracing is disabled
- move logging logic from utils to observability
- pass trace_id/span_id in HEX form

2026-01-03 16:29:58 +03:00

configs

Version 0.4.0

2026-01-03 16:29:58 +03:00

docker-compose.yaml.jinja

Version 0.4.0

2026-01-03 16:29:58 +03:00

Dockerfile.jinja

Initial commit

2025-11-29 22:13:34 +03:00

Makefile

Version 0.3.0

2025-12-06 19:41:33 +03:00

README.md

Initial commit

2025-11-29 22:13:34 +03:00

README.md

deploy example

This is a complete deployment example with following services:

postgres database (with initialization and rootless user):
- 5432 port exposure commented
api (with data migrations) with 8080 port exposure:
- configured by configs/api.yaml
- exposes 8080 port
- 9090 metrics port not exposed
prometheus to collect metrics
- configured by configs/prometheus.yaml
- exposes 9090 port with prometheus UI (optional)
grafana as a powerful UI for metrics visualization
- exposes 3000 port, default user:password is admin:admin
- prometheus metrics are available at http://prometheus:9090
jaeger to collect and show traces
- exposes UI at port 16686
otel (OpenTELemetry) agent working as proxy for jaeger
- configured by configs/otel.yaml
- does not expose 4317/4318 ports as containers use it inside the internal network