Workflow

Source: src/AWS/MWAAServerless/Workflow.ts

An Amazon Managed Workflows for Apache Airflow Serverless workflow — a serverless Airflow DAG defined by a YAML file in S3 and executed by AWS-managed, multi-tenant Airflow infrastructure without provisioning an environment.

Each update to the definition or configuration creates a new workflow version; MWAA Serverless keeps only the latest version actively scheduled.

Creating a Workflow

Basic Workflow

import * as MWAAServerless from "alchemy/AWS/MWAAServerless";
import * as IAM from "alchemy/AWS/IAM";

const role = yield* IAM.Role("WorkflowRole", {
  assumeRolePolicyDocument: {
    Version: "2012-10-17",
    Statement: [{
      Effect: "Allow",
      Principal: { Service: "airflow-serverless.amazonaws.com" },
      Action: ["sts:AssumeRole"],
    }],
  },
});

const workflow = yield* MWAAServerless.Workflow("Etl", {
  definitionS3Location: {
    bucket: "my-dag-bucket",
    objectKey: "workflows/etl.yaml",
  },
  roleArn: role.roleArn,
});

Workflow with Logging and Tags

const workflow = yield* MWAAServerless.Workflow("Etl", {
  definitionS3Location: {
    bucket: "my-dag-bucket",
    objectKey: "workflows/etl.yaml",
  },
  roleArn: role.roleArn,
  description: "nightly ETL",
  loggingConfiguration: { logGroupName: "/mwaa-serverless/etl" },
  tags: { team: "data" },
});