When performing a snapshot restore we now (for the last ~7 months) overwrite the security alphabet in the rest-security.conf with the old value.
Unfortunately, we restore the stage files and regenerate the stage auth token before we do this, which causes the stage auth token to be incorrectly generated.
The stage auth token needs to be regenerated after the security files are restored, or the security files need to be restored before we call the restore_stage_files here: https://github.com/cloudify-cosmo/cloudify-manager/blob/master/workflows/cloudify_system_workflows/snapshots/snapshot_restore.py#L384
Steps to reproduce:
1. Restore a snapshot onto a manager with a different rest-security.conf to the old one (not just another copy of the same image)
2. Verify that stage stops working because /opt/cloudify-stage/resources/admin_token has an invalid security token.