Updating an Amazon MWAA environment - Amazon Managed Workflows for Apache Airflow

Updating an Amazon MWAA environment

The following section describes how you can update your environment preferences and settings, such as the environment class, the number of Apache Airflow schedulers, your environment's maintenance window, as well as the path for your DAGs, requirements.txt, and plugins.zip resources in Amazon S3.

Updating your environment using the Amazon MWAA console

You can update your environment on the Amazon MWAA console at any time.

To configure the environment

  1. Open the Environments page on the Amazon MWAA console.

  2. Choose an environment from the list.

  3. Choose Edit.

  4. (Optional) On the Specify details page, you can do the following:

    1. In the Environment details section, for Weekly maintenance window start (UTC) option, you can choose a day of the week and the time in 24-hour format (HH:mm) from the dropdown lists to specify when Amazon MWAA will perform scheduled maintenance operations for your environment.

    2. In the DAG code in Amazon S3 section, you can change your environment's Amazon S3 bucket, specify a different DAGs folder in your bucket, and change the location or version of your plugins.zip and requirements.txt files.

  5. Choose Next.

  6. (Optional) On the Configure advanced settings page, in the Networking section, can do the following:

    1. Web server access — Change your environment's web server access mode. Choose Private network to restrict access to the Apache Airflow web server only to your Amazon VPC. Choose Public access if you want to be able to access the web server using secure login over the internet. For more information, see Apache Airflow access modes.

    2. Security group(s) — Create a new security group, and add or remove existing ones to configure the inbound and outbound rules for your environment.

  7. (Optional) In the Environment class section, you can do the following:

    1. Change the class type to configure your environment's scheduler, worker, and web server vCPU capacity. For more information about environment classes, see Amazon MWAA environment class.

    2. For Maximum worker count, specify a number between 1 and 25 to set the maximum number of workers that Amazon MWAA can scale up to. When a worker reaches its maximum task capacity, Amazon MWAA adds additional workers until the total number of workers reaches the number you set here.

    3. For Minimum worker count, specify a number less than Maximum worker count to set the minimum number of workers that Amazon MWAA maintains to run tasks.

    4. For Scheduler count, specify the number of schedulers you want to run for your environment.

      • Apache Airflow v2 — Accepts between values between 2 and 5.

      • Apache Airflow v1 — Accepts 1.

  8. (Optional) In the Monitoring section, under Airflow logging configuration you can activate or deactivate CloudWatch Logs for various Apache Airflow log types, and choose the Log level for each log type from the corresponding dropdown lists. For more information on Apache Airflow logs, see Viewing Airflow logs in Amazon CloudWatch.

  9. (Optional) In the Airflow configuration options section, you can add or remove custom configuration values to override the Apache Airflow defaults. For more information on Apache Airflow configuration options, see Apache Airflow configuration options.

  10. (Optional) In the Permissions section, under Execution role you can choose the IAM role used by your environment to perform various actions such as access your DAGs and resources in Amazon S3, and send logs to CloudWatch.

  11. Choose Next.

  12. On the Review and save page, review the changes you've made to your environment.

  13. Choose Save.


When you update your environment, it can take between 10 to 30 minutes for changes to take effect.

Troubleshooting environment updates

  • If you update your environment to a different environment class (such as changing an mw1.medium to an mw1.small), and the request to update your environment fails, the status changes to the UPDATE_FAILED state, the environment is rolled back, and is billed according to, the previous stable version of the environment. For more information, see I tried changing the environment class but the update failed.