0001_initial was referencing events.0011_dashboard_indexes which no
longer exists as a file on disk (the DB has it applied but the file
was removed). Updated dependency to 0010_merge_20260324_1443 which
is the latest events migration file present, resolving the
NodeNotFoundError on management commands.
Users have been migrated from eventify-django SQLite to eventify-backend
PostgreSQL. The temporary users_db workaround is no longer needed:
- settings.py: removed users_db SQLite secondary database config
- views.py: removed _user_db()/_user_qs() helpers; user views now query
the default PostgreSQL directly with plain User.objects.filter()
- docker-compose.yml: SQLite read-only volume mount removed
All 27 users (25 non-superuser customers) now live in PostgreSQL.
The admin_api was querying eventify-backend's empty PostgreSQL. Real users
live in eventify-django's SQLite (db.sqlite3 on host). Fix:
- settings.py: auto-adds 'users_db' database config when users_db.sqlite3
is mounted into the container (read-only volume in docker-compose)
- views.py: _user_db() helper selects the correct database alias;
_user_qs() defers 'partner' field (absent from older SQLite schema)
- UserMetricsView, UserListView, UserDetailView, UserStatusView all use
_user_qs() so they query the 25 real registered customers
- UserListView and UserMetricsView now filter is_superuser=False so only
end-user accounts appear in the admin Users page (not admin/staff)
- _serialize_user now returns avatarUrl from profile_picture field so the
grid view renders profile images instead of broken img tags
- RegisterForm and WebRegisterForm now set is_customer=True and
role='customer' on save so future registrants are correctly classified
