ARM template auto-shutdown and cleanup helper scripts

[bpkroth]

trafficstars

Often times we will run with teardown: false and tweak experiments for different variations. This means that after a while we may have stale resources running and eating resources in the cloud (Azure).

To fix that I propose:

1. [ ] Simply enable an auto-shutdown policy that somehow (*) detects idle resources and shutsdown the VM. That could help with compute costs, but not allocation costs.
   • [ ] (*) Not sure if the native version does this or if it's just a schedule, in which case we'd have to write something to run on the VM
   • [ ] Could also configure the ARM templates to do cascading deletion of certain resources, but probably not all (e.g., shared storage, shared vnets, shared db, etc.)
2. [ ] We need a script that could look at a shared database instance for all experiment definitions, and can determine which resources may no longer be necessary, and propose deletions.
   • [ ] We might need to add additional metadata to help track that.