2026-09-28 15:51:13 +02:00
|
|
|
# BUGS.md — open findings
|
|
|
|
|
|
2026-11-16 15:10:10 +01:00
|
|
|
One finding is still open (B-93), narrowed to what applies to the Django stack. A finding is
|
|
|
|
|
removed from this file once it is fixed, and `git log --all --grep 'B-nn'` is the record of how it
|
|
|
|
|
was closed. Each entry gets its own commit with its own regression test, and the `B-nn` marker
|
|
|
|
|
goes in a comment next to the fix. The full history of every B-nn is in
|
|
|
|
|
[docs/storico-bug.md](docs/storico-bug.md); the owner of this one in the plan is roadmap step C6.
|
2026-09-28 15:51:13 +02:00
|
|
|
|
|
|
|
|
## Medium
|
|
|
|
|
|
|
|
|
|
### B-93 — Deploys and rollbacks
|
|
|
|
|
|
2026-11-16 15:10:10 +01:00
|
|
|
On the Django stack (`docker-compose.yml`):
|
2026-09-28 15:51:13 +02:00
|
|
|
|
2026-11-16 15:10:10 +01:00
|
|
|
- every image is tagged `plm-lottery-web:local`, so there is no per-release tag and an older image
|
|
|
|
|
cannot be put back;
|
|
|
|
|
- no `stop_grace_period`: a worker stopped in the middle of a step gets Docker's default timeout;
|
|
|
|
|
- `SITE_ADDRESS` passed on the command line reverts to `localhost` on the next plain
|
|
|
|
|
`docker compose up`;
|
|
|
|
|
- there is no documented pre-deploy backup and no downgrade procedure, and an older image cannot
|
|
|
|
|
start on a database already migrated forward.
|
|
|
|
|
|
|
|
|
|
(The old stack's part — SSE streams blocking a graceful shutdown, task cancellation — goes away
|
|
|
|
|
with roadmap D1.)
|
|
|
|
|
|
|
|
|
|
Fix: tag images per release; `stop_grace_period: 30s` and let the worker finish its step on
|
|
|
|
|
SIGTERM; `SITE_ADDRESS` in `.env` read as `${SITE_ADDRESS:?}`; document the backup before a deploy
|
|
|
|
|
and the downgrade (restore the backup, start the previous tag).
|