First Draft: Orchestration for node placement

Potentially research the following:
- PIpelining
- other wms/graph algorithms 

Find a way for how to identify:
- Nodes that can run on the same executor
- Have all their inputs ready (for a subgraph/community)
- use shared resources

