What’s the best way to start all of my stacks sequentially to avoid overwhelming resources on host reboot?

My system (disk especially) gets bogged down when all of my stacks attempt to deploy at once. I set almost all of my stacks to restart:no (which pains me) and created a startup procedure that enables me to set the startup sequence of stacks.

The first problem I am facing with this is whether I use Start Stack or Deploy Stack – they sometimes fail (seemingly because the stack may already be started in some scenarios).

What's the standard way to handle this?

One constraint, err… preference, that I have is that my DNS/DHCP container gets priority to get the network back online quickly.