Recipe Graphs

Automation recipes are directed acyclic graphs (DAGs) with discrete steps that can either do data work or provision infrastructure.

Each recipe has at least one graph, defined in the recipe's variations.json file, that represents its node structure. The Graphs tab on the Recipes page lists all of the graphs configured for a recipe and the variations that use each graph.

Open a graph

To view a graph, open a kitchen and select a recipe from the list. Click a variation name. The Graph tab opens by default.

Visual example in Automation

Parts of a graph

Node

The components of a graph associated with the steps that do work in a recipe are called nodes. Graphs are built from the nodes available in a recipe. A graph requires at least one node to be valid. Nodes come in many types and configurations.

See Recipe Nodes for more information.

Edges

The nodes in a graph are connected by edges, although they are not required. Edges always connect two nodes and are directional—outbound from one node and inbound to another node. Edges can connect a single node to multiple nodes.

Conditions

Flow-control decision points that can determine which downstream node paths to process.

See Conditional Nodes for more information.

Graph size

• Single-node: graphs can be as small as a single node with no edges.
• Multi-node: graphs can be designed to be very complex, containing steps for infrastructure configuration, the preparation of data asset inputs, data integrations, analyses, generation of data outputs, and visualization. They can consist of dozens or even hundreds of nodes and edges.
  • Multi-node graphs can also contain floating nodes—nodes that do not share any edges with the other nodes in the graph.

Graph structure

• Series graph: simple graphs run all nodes sequentially in a series.

• Parallel graph: graphs can run more than one path of nodes concurrently.

See Graph Processing for how series and parallel graphs process nodes during an order run.