AWS Systems Manager now supports taking action for Automation step failure by allowing branching to another step within the Automation Document. This feature allows customers to perform graceful exits, send notifications, and a variety of clean up actions when an automation step fails. For example, when an Amazon Machine Image (AMI) update automation step fails midway, the customer can now call for automation to take a snapshot, close the EC2 instance, notify admin, and exit gracefully. Previously, there were only two choices, “Continue” and “Abort” upon failure of a step.

Systems Manager Automation is an AWS-hosted feature for securely automating workflows through schema-based runbooks called Automation Documents. Common use cases include start/stop instances with approvals, AMI building, RDS snapshot, rescue EC2 instance, and safe patching. The on failure conditional feature is available in all AWS Regions.

For more details on this feature, visit our Documentation. To learn more about AWS Systems Manager, visit our Product Page.