I was wondering how cloud providers like AWS, GCP, and Azure handle hardware issues that could affect virtual machines.