Nginx round-robin entry point for 2+ stateless API replicas with Docker DNS discovery
shared PostgreSQL idempotency and Valkey sessions across replicas
separate migration job, graceful shutdown, failover and rolling update scripts
Prometheus/Grafana monitoring for API replicas, Nginx, PostgreSQL and Valkey
Gitea CI plus manually approved exact-SHA deployment workflow
Verification
make check
go test ./...
go build ./cmd/...
public smoke and PostgreSQL/Valkey integration checks
persistence after restarting API and both stores
two API failover runs with recovery in 1.35–1.44 seconds
rolling update under 60 seconds of Go load: 134894 successful requests, 0 errors
Grade 5 track
Track 1 compares 1/2/4 replicas under fixed and growing resource budgets across 60 measured points. Raw data, median aggregates, the RPS chart, pool tuning result and limitations are documented in REPORT.md and evidence/.
## Summary
- Nginx round-robin entry point for 2+ stateless API replicas with Docker DNS discovery
- shared PostgreSQL idempotency and Valkey sessions across replicas
- separate migration job, graceful shutdown, failover and rolling update scripts
- Prometheus/Grafana monitoring for API replicas, Nginx, PostgreSQL and Valkey
- Gitea CI plus manually approved exact-SHA deployment workflow
## Verification
- `make check`
- `go test ./...`
- `go build ./cmd/...`
- public smoke and PostgreSQL/Valkey integration checks
- persistence after restarting API and both stores
- two API failover runs with recovery in 1.35–1.44 seconds
- rolling update under 60 seconds of Go load: 134894 successful requests, 0 errors
## Grade 5 track
Track 1 compares 1/2/4 replicas under fixed and growing resource budgets across 60 measured points. Raw data, median aggregates, the RPS chart, pool tuning result and limitations are documented in `REPORT.md` and `evidence/`.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Summary
Verification
make checkgo test ./...go build ./cmd/...Grade 5 track
Track 1 compares 1/2/4 replicas under fixed and growing resource budgets across 60 measured points. Raw data, median aggregates, the RPS chart, pool tuning result and limitations are documented in
REPORT.mdandevidence/.92e3d58548toff810a44c1ff810a44c1to0297f4d8c00297f4d8c0to7623187e1a7623187e1ato7cdb7ce0bf7cdb7ce0bfto18d51db435