Every aspect of Grai can be serialized to yaml. This allows you to manage your entire data lineage as code, backup your lineage in version control, or create and edit nodes/edges in code. You can find explicit yaml definitions for all Grai objects in the grai-schemas library. Documentation for the schemas can be found here.

Web App

The YAML File integration is used to upload a single YAML file, which needs to have the correct Grai format.

Fields

CLI

You can create, edit, and delete lineage in yaml through the CLI. Let's say you wanted to add an is_really_important flag to the metadata on the customer_id column of the customers table. In that case we might write a yaml definition like this

type: Node
version: v1
spec:
  name: "customers.customer_id"
  namespace: default
  is_active: true
  metadata:
    grai:
      node_type: Column
    is_really_important: true

We could then update the server with this new information through the CLI

grai apply <node_file>

You can do the same with edges

version: v1
type: Edge
spec:
  source:
    name: public.lineage_node.id
    namespace: default
  destination:
    name: node2
    namespace: test
  is_active: true
  metadata:
    grai:
      edge_type: "Edge"
    stuff: goes_here