Easier said than done, since a hypervisor is yet another moving part that can fail. People were running Crowdstrike on AWS (which in most (all?) cases involves

Easier said than done, since a hypervisor is yet another moving part that can fail. People were running Crowdstrike on AWS (which in most (all?) cases involves a hypervisor), and recovery was still painful for them.