I have an AKS cluster of four nodes that was working (4/4), but I wanted to scale it up to handle my workload better. I doubled the number of nodes, but the new VMs were all in