Enable to extend 35 days maintenance window in case of failure
Today, we can leverage "Planned Maintenance" to avoid any impact on VMs. But, the maintenance may fail if it cannot find appropriate resource to re-deploy a VM. It would be great if the 35 days maintenance window can be reset in that circumstance. In some mission-critical system, it's hard to secure sufficient time frequently for maintenance (let's say once a month). Thus, extending the window is needed in case of failure.