In this meta issue, we can highlight the main entry points and access patterns for the genesis file. This will ensure that any work to change the current behavior/functionality not breaking (if possible). There are many enhancements (e.g. remove the genesis from memory) that we plan to do and the information in this meta issue will make it easier to perform the work to develop these enhancements by surfacing the context around the genesis file usage in the project.

These are some of the things we need to get more insights into:

• When it is accessed via RPC, are all fields used or just some? If only a subset is needed, can we provide users
  with means to return only the requested part?
• When the node starts, it loads the entire genesis file into memory, is this needed for the entire lifecycle of a node operation ?
• Is it possible to load the genesis information on demand ? And free it from memory when it's not needed ?