Parameters

Using Parameters in DAGs

Parameters allow you to make your workflows dynamic and reusable by injecting values at runtime.

Defining Parameters in DAGs

Here’s an example DAG with parameter definitions:

apiVersion: kontroler.greedykomodo/v1alpha1
kind: DAG
metadata:
  name: dag-sample-event
spec:
  parameters:
    - name: first
      defaultFromSecret: secret-name    # Default value from a Kubernetes secret
    - name: second
      defaultValue: value               # Static default value
  task:
    - name: "random"
      image: "alpine:latest"
      command: ["sh", "-c"]
      args: ["if [ $((RANDOM%2)) -eq 0 ]; then echo $second; else exit 1; fi"]
      parameters:
        - first                        # Declare which parameters this task uses
        - second                       

Parameter Types

1. Secret References:

   parameters:
     - name: first
       defaultFromSecret: secret-name

   • Securely fetch values from Kubernetes secrets
   • Useful for credentials and sensitive data

2. Static Values:

   parameters:
     - name: second
       defaultValue: value

   • Hardcoded default values
   • Can be overridden in DagRuns

Using Parameters in Tasks

Tasks must declare which parameters they use:

task:
  - name: "task-name"
    parameters:
      - second    # List parameters used by this task

Overriding Parameters in DagRuns

Create a DagRun to execute a DAG with custom parameter values:

apiVersion: kontroler.greedykomodo/v1alpha1
kind: DagRun
metadata:
  name: dagrun-sample5
spec:
  dagName: dag-sample1
  parameters:
    - name: first
      fromSecret: secret-name-new    # Override with different secret
    - name: second
      value: value_new               # Override with new value

Parameter Override Types

1. Secret Override:

   parameters:
     - name: first
       fromSecret: secret-name-new

2. Direct Value Override:

   parameters:
     - name: second
       value: value_new

Best Practices

1. Security:

   • Use secrets for sensitive information
   • Avoid hardcoding credentials in DAG definitions

2. Naming:

   • Use clear, descriptive parameter names
   • Follow consistent naming conventions

3. Task Declaration:

   • Always declare parameters used by tasks
   • Only expose necessary parameters to tasks

4. Default Values:

   • Provide sensible defaults when possible
   • Document expected parameter formats

Parameter Scoping

• Parameters are available to tasks that explicitly declare them
• Each task can access only its declared parameters
• Parameter values are immutable during execution