When trying to understand how the WMCO works, I have found the following annoying or not working as intended : 

1. When using VSphere and IPI, I was unable to use another username than "Administrator" (which is an issue when working with other than EN-US languages). It is somehow weird as the BYOH based on a configmap allows to use another username.

2. When using VSphere and IPI, as a new windows node is being bootstrapped, when the hostname is set, the node is issued a restart command but the WMCO controller still follows that command with a few other provisionning command before crashing because of the connection loss. It would be better to have a "wait" or properly handle the host reboot instead of crashing and bootsrapping again.

3. It would be great to be able to inject custom containerd, kubelet and other components custom configurations instead of having them generated and overwritten if hash does not match. The objectives would be to enable registry auth on the containerd side for example. It might also help solve other issues in specific cases without having to change the operator code.

OpenShift 4.12.8 and WMCO 7.0.1

Every time

1. Follow the WMCO instructions in a Vsphere/Vcenter environnment
2. Add a "--debugLogging" arg to the wmco container (in the deployment included in the yaml operator) 
3. Follow the standard Vsphere provisionning instructions and see the above points unfold during bootstrap. 

The node restarts after hostname change and the WMCO controller restarts provisionning.

Have the node restart but the WMCO controller wait for the reboot and continue boostrapping or, with a timeout, bootstrap another node.