fix: Ensure deploy script runs once and is part of entrypoint

The deploy.sh script is now re-added to the entrypoint.sh script
to ensure it runs only during first container startup.

A flag file (/flags/.deployed) is now created after a successful deployment.
The deploy.sh script checks for this flag and will not re-run
deployment steps unless FORCE_DEPLOY is set to true. This prevents
unnecessary re-runs of migrations, collectstatic, etc., on subsequent
container starts within the same deployment.

Corrected permissions for `/app/.cursor-server` and created a `/flags`
directory with appropriate permissions in the `Dockerfile`. Added
ENV DJANGO_SETTINGS_MODULE with default value to `Dockerfile`.
This commit is contained in:
badblocks 2025-05-23 16:31:49 -07:00
parent c87d73435b
commit d4948e7cd3
No known key found for this signature in database
3 changed files with 21 additions and 8 deletions

View file

@ -1,5 +1,10 @@
#!/bin/bash
if [[ -f /flags/.deployed && "$FORCE_DEPLOY" != "true" ]]; then
echo "*** Previously deployed successfully."
exit 0
fi
echo "*** Running makemigrations --check to make sure migrations are up to date..."
django-admin makemigrations --noinput --check 2>&1 || exit 1
@ -12,4 +17,7 @@ django-admin clear_cache 2>&1
echo "*** Running collectstatic..."
django-admin collectstatic -c --no-input 2>&1
echo "*** Marking as deployed..."
touch /flags/.deployed
echo "*** Deployed successfully!"