So let me start this off with a scenario, all your VMs are down, but don’t worry you have HA setup and the VMs are starting back up on another host, however, your other VM running AD has started last so all of that time you spent waiting for it to come up your other hosts were not running properly. VM startup order is the solution here. You simply create a VM/host group on the cluster level, this would encompass the set of VMs you want to begin powering on, then VM/host rules decides the order in which each group is powered on. One group cannot begin powering on these VMs until the dependency rules are met for the first group which can be found in HA settings. For example, there is resources allocated, when resources have been allocated to the “Domain Controller” group for example the next group can begin powering on, then there’s powered on, so you’ll need to wait for group 1 to power on before group 2 can, then probably what’s best is heartbeat received, so the VM needs to be up and sending heartbeats confirming the OS is working properly. Technically there’s also app heartbeats that you can enable so only proceed once you know a specific app is working. You will need to select VM to VM in the host rules for startup order, once group A is up proceed with group B, the other options you see there are VM to host, which affects which host a group of VMs should run on and allow you to keep a group of VMs off certain hosts, you will need to create a specific host group first though, then there’s default of keeping VMs together which just starts the VMs normally no startup order or specific host it need to avoid or be run on, finally there’s keep VMs separate in which case you can define a set of VMs say a Domain Controller for each host that will prevent these VMs from running on the same host this rule will keep them separated, I will try to circle back and go back into detail about host rules. The main story here was VM startup order.