Demo AWS Fault Injection Simulator

• Create an IAM role that FIS can assume to perform actions on your behalf.
• Create an experiment template that targets one or more resources and defines the fault to inject.
• Start the experiment, monitor progress, and review results (including CloudTrail and FIS logs).

​

Step 1 — Create the FIS IAM role

1. In the AWS Console, open IAM and create a role that FIS can assume.
2. The role must include a trust policy allowing the service principal fis.amazonaws.com to assume it and must grant the permissions required by the experiment (for example, permissions to stop EC2 instances).
3. Attach least-privilege permissions required for the actions you will run (for example, a policy that allows ec2:StopInstances on the target instance(s) and iam:PassRole if needed).

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": { "Service": "fis.amazonaws.com" },
      "Action": "sts:AssumeRole"
    }
  ]
}

​

Step 2 — Create an experiment template

1. Open the FIS console and choose “Create experiment template.”
2. Define one or more targets using resource identifiers (for example, EC2 instance IDs or tags that match the instances you want to test).
3. Add one or more actions (for example, aws:ec2:stop-instances) and set the action parameters:
   • Region (for example, us-east-2)
   • Duration or timeout, if applicable
   • Any additional action inputs required by the action
4. Associate the experiment with the IAM role you created in Step 1.
5. Review and save the template. You can reuse templates for repeatable testing.

​

Step 3 — Start the experiment and observe

arn:aws:ec2:us-east-2:635573991785:instance/i-0d96cc656ab9a987c

​

Observability and post-experiment checks

• Use the Timeline view in FIS to monitor remaining duration and progress of actions while the experiment runs.
• Review the Targets and Actions sections to confirm which resources were affected and whether each action completed successfully.
• Check Log events in FIS and consult CloudTrail for a full audit trail of API calls made by FIS during the experiment.
• Validate your monitoring, alerting, and automated recovery mechanisms to ensure they behave as expected during the injected failure.

​

Links and references

• AWS Fault Injection Simulator (FIS) docs: https://docs.aws.amazon.com/fis/latest/userguide/
• AWS CloudTrail: https://docs.aws.amazon.com/awscloudtrail/latest/userguide/
• IAM documentation: https://docs.aws.amazon.com/IAM/latest/UserGuide/
• EC2 documentation: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/