Infrastructure
Observability
Health checks, logging, Sentry, and monitoring gaps.
Observability
ByteBite observability is application-level today. Centralized infrastructure monitoring is not yet deployed.
Health checks
API endpoints
| Endpoint | Purpose | Tenant lookup |
|---|---|---|
GET /health | Liveness — {"status":"ok","service":"bytebite-api"} | Skipped |
GET /health/db | Database connectivity — ok or degraded | Skipped |
Accessible on direct API hosts and same-origin /api/health through Traefik. On staging those Traefik routes require HTTP BasicAuth; container healthchecks below do not.
Container healthchecks
Defined in Dockerfiles (HEALTHCHECK directive):
| Service | Check |
|---|---|
| API | fetch http://127.0.0.1:3001/health |
| Storefront | fetch http://127.0.0.1:3000/ |
| Dashboard | wget http://127.0.0.1:8080/ |
| Docs | fetch http://127.0.0.1:3002/ |
Local dev additionally healthchecks PostgreSQL (pg_isready) and MinIO (mc ready).
Storefront depends_on: api: condition: service_healthy in compose.deploy.yml.
Error monitoring: Sentry
| Aspect | Detail |
|---|---|
| Package | packages/observability |
| Configuration | SENTRY_DSN environment variable |
| Local dev | Optional — initSentry is a no-op without DSN |
| Staging | Configured via Dokploy (optional) |
| Production | Required when deployed (planned) |
Never document DSN values.
Logging
| Aspect | Detail |
|---|---|
| Library | pino (packages/observability) |
| Format | JSON in production, pretty-printed in development |
| Level | LOG_LEVEL env var (default info in staging compose) |
| Centralized aggregation | Not deployed |
Platform observability
| Capability | Status |
|---|---|
| Docker container status | Available via docker ps / Dokploy UI |
| Traefik access/error logs | Container logs on APP VM |
| Dokploy deployment logs | Dokploy UI |
| PostgreSQL backup service logs | journalctl -u bytebite-postgres-backup.service |
| PostgreSQL maintenance service logs | journalctl -u bytebite-postgres-backup-maintenance.service |
| Restic repository integrity checks | Weekly via maintenance timer |
| Centralized logging (Loki, ELK, etc.) | Not deployed |
| Alerting / paging | Not deployed |
| Uptime monitoring (external) | Not deployed |
| Infrastructure metrics (Prometheus, etc.) | Not deployed |
| Backup failure alerting | Not deployed |
| Restore drill alerting | Not deployed |
Operational checks
After infrastructure changes, run the Validation smoke test suite. For application-level health, see Operations overview.
Related
- Validation — smoke tests
- Troubleshooting — symptom diagnosis
- Known gaps — monitoring gaps