Restart nextcloud on change in setup-apps

The setup-apps script is stored in a configmap -- created by helm based on the apps helm value. This script is mounted inside the main nextcloud pod. We run the script using kubectl exec from a separate Job on helm post-install and post-upgrade.

This works, but if

1. setup-apps changes, for example because of a change in the apps value, and
2. the nc-nextcloud deployment doesn't change

then the deployment still has the old configmap mounted, and so the old version of the script gets run by the post-upgrade script.

Perhaps we can solve this by including a hash of the apps value as an annotation on the deployment, but this is tricky because that's created by the upstream nextcloud chart, so we'd have to set it in the values.yaml in our chart, so this only works if that particular helm value gets templated by the nextcloud chart 🙏