Week 9 recap

Pydantic

• Provides
  • Type enforcement
  • Type casting
  • Rich objects (e.g. attribute based)
  • JSON objects
• Types come from
  • Python (str, bool, etc)
  • Typing (List, Optional, etc)
  • Pydantic itself
• Enums
  • Can create your own ENUM (or choices)
• Nesting can happen with Pydantic

DiffSync

• Models
  • Define what the data will look like
  • Based on Pydantic
• Model Key attributes
  • _modelname*: Defines the type of the model; used to identify common models between different systems.
  • _identifiers*: List of instance field names used as primary keys for this object.
  • _shortname: List of instance field names to use for a shorter name. Unnecessary if matches _identifiers.
  • _attributes: List of non-identifier instance field names for this object; used to identify the fields in common between data models for different systems.
  • _children: Dict of {: } indicating which fields store references to child data model instances.
• Adapters
  • Define how to load the data
  • Performs the CRUD operations

Nornir

• Comes with a SimpleInventory that allows yaml diagrams
• Three Components
  • hosts: device info, connection details, user defined data, groups
  • groups: logical grouping of devices, data inheritance, scale
  • defaults: define data that applies to all hosts as a baseline
• Simple Filtering
  • nr.filter(role="leaf")
  • nr.filter(role="border", region="EU")
• Advanced Filtering
  • nr.filter(F(region="EU") | F(region="US"))
  • nr.filter(F(role="border") & ~F(region="US"))