Gancio refuses to start when its DB has tables but the data volume has no
config.json ("Non empty db! Please move your current db elsewhere than retry"),
which produces an infinite restart loop. This hit production tenants bnkops
and trbh (>1200 restart cycles each) — proximate cause was a missing
config.json in changemakerlite_gancio-data with the DB fully populated.
Add gancio-config-init alpine sidecar that runs on every `up`:
- no-op when config.json exists
- regenerates from .env when missing (1000:1000 ownership)
- gancio service now depends on its service_completed_successfully
Also harden verify_gancio_config in upgrade.sh to error loudly when
multiple gancio-data volumes match (silent head -1 could pick the wrong
one after a compose project rename).
bunker-admin
a82e95946b