We have detected instances provisioning issues on AMS1.
Your instance may not be able to start and can fallback to being "archived" in the Console.
The issue was related to our worker and our compute API.
It has been fixed. Everything is back to normal.
Issue has been escalated to local team