Initial version of NDv4+cycle+anf+nhc deployment

[garvct]

Complete E2E deployment of an NDv4 cluster leveraging Cyclecloud

The NDv4 cluster will consist of

• NDv4 compute nodes running Ubunt-hpc 18.04, Max GPU app clock freqencies are set, GPU persistent mode is enabled, local NVMe SSD's are mounted.
• Cyclecloud 8.2.2 with SLURM 2.6.4(pmix support) is installed and the pmix libraries are installed on NDv4 nodes.
• Automatic recovery from a reboot enabled (e.g NVMe SSD will be remounted and GPU clock freq reset)
• Premium SSD's used for OS disks, with larger capacity (60GB)
• Accelerated networking is enabled on NDv4
• User home directories are mounted on Azure netapp files
• Extensive automatic pre-job healthchecks are enabled (Unhealthy nodes will be put into a DRAIN state)
• A separate login node (separate from the scheduler is deployed)
• Cyclecloud autoscaling is disabled.
• Windows server (winbox) is deployed to access the Cyclecloud portal via RDP