docs(infra): document Compose profile + down/up symmetry gotcha
CI / commits (pull_request) Successful in 1m41s
CI / scan (pull_request) Successful in 1m50s
CI / check (pull_request) Successful in 1m51s
CI / a11y (pull_request) Successful in 1m8s
CI / perf (pull_request) Successful in 2m4s

`docker compose down -v` left pgweb and jaeger running (under
`--profile dbtools` / `--profile observability` respectively), with
no obvious diagnostic. Reason: Compose only operates on services
whose profile is currently active — if you brought them up with a
flag, you must `down` with the same flag, otherwise they are
invisible to that command.

Document the gotcha in `infra/README.md` "Local-dev stack" →
"Operational tips", with the two pragmatic resolutions: pass the
same flags on each command, or set `COMPOSE_PROFILES` once and
let it propagate.
This commit is contained in:
Julien Gautier
2026-05-09 21:16:42 +02:00
parent b2d7fe3fd7
commit c5b2c1fc48
+12
View File
@@ -172,6 +172,18 @@ All ports are overridable via `.env` if the host machine has conflicts.
### Operational tips ### Operational tips
- **Persistence** — state lives in named Docker volumes (`apf-portal-postgres-data`, `apf-portal-redis-data`). Survives `docker compose down`. Use `docker compose -f dev.compose.yml down -v` to wipe (also wipes the audit-roles bootstrap, which re-runs on the next fresh boot). - **Persistence** — state lives in named Docker volumes (`apf-portal-postgres-data`, `apf-portal-redis-data`). Survives `docker compose down`. Use `docker compose -f dev.compose.yml down -v` to wipe (also wipes the audit-roles bootstrap, which re-runs on the next fresh boot).
- **Profiles must match on `down` as on `up`** — `docker compose down` only acts on services whose profile is currently active. If you brought the stack up with `--profile dbtools --profile observability`, those same flags are required on `down` to also stop pgweb and Jaeger; otherwise they keep running until the next reboot. Two pragmatic patterns:
```bash
# Either pass the same flags on each command:
docker compose -f dev.compose.yml --profile dbtools --profile observability down -v
# Or set COMPOSE_PROFILES once (e.g. in your shell or in infra/local/.env)
# so every `up` / `down` / `ps` sees the right scope:
export COMPOSE_PROFILES=dbtools,observability
docker compose -f dev.compose.yml down -v
```
- **Bootstrap re-run** — the SQL in `local/init/postgres/` only runs on a **fresh** Postgres data volume. To replay after editing the file, `down -v` (loses all dev data) or run the SQL manually with `docker compose exec postgres psql -U portal -d portal_dev -f /docker-entrypoint-initdb.d/01-init.sql`. - **Bootstrap re-run** — the SQL in `local/init/postgres/` only runs on a **fresh** Postgres data volume. To replay after editing the file, `down -v` (loses all dev data) or run the SQL manually with `docker compose exec postgres psql -U portal -d portal_dev -f /docker-entrypoint-initdb.d/01-init.sql`.
- **Logs** — `docker compose -f dev.compose.yml logs -f <service>` to follow a single service. `otel-collector` is the loudest — its `debug` exporter prints every span / metric / log it receives. - **Logs** — `docker compose -f dev.compose.yml logs -f <service>` to follow a single service. `otel-collector` is the loudest — its `debug` exporter prints every span / metric / log it receives.
- **Image upgrades** — same policy as the runner image (deliberate, not via `:latest`). Renovate's docker-compose manager will surface bumps automatically once the dashboard rule allows them. - **Image upgrades** — same policy as the runner image (deliberate, not via `:latest`). Renovate's docker-compose manager will surface bumps automatically once the dashboard rule allows them.