How can I resume provisioning after step error?

When one of the provisioning steps fail, I would like to retry (as the logging message suggests - “correct the error and try again”).

This is however not possible. After running “vagrant provision” again or “vagrant reload --provision”, the script either complains about “guest specific” stuff that it can’t do right now, because the machine isn’t ready for it or it just isn’t able to connect and goes on connecting/retrying).

I’m on Vagrant 2.3.4, Windows 10, trying to install generic/fedora37.

I noticed folks on the internet suggested to use Ansible provisioning instead of the default so I went ahead and rewrote my provisioning to an ansible playbook, but to no avail (still the same failure).

Is there a good way to achieve this (being able to resume provisioning)?

PS: I’ve also seen advice like “remove the vbguest plugin”. I didn’t try that because…well…I would like to use that plugin…