Bump this from 10 minutes (60 * 10 seconds) to 30 minutes (180 * 10
seconds) to ensure there is sufficient time for (relatively) large VMs
to migrate with (relatively) slow networking.
This playbook will perform a oneshot upgrade of the systems in the
cluster, including performing a clean and safe reboot of the node(s) if
required (either due to services needing a restart, or the kernel
changing). It runs in serial=1 and only reboots if needed.