Run an experiment

Preview

This product is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms. Pre-GA products are available "as is" and might have limited support. For more information, see the launch stage descriptions.

This page explains how to run an experiment in Fault Injection Testing.

Running an experiment involves using an existing template to create an experiment. The experiment does a dry run where it automatically verifies the target resources and permissions needed. Once the experiment is ready, you can proceed with the fault injection.

Before proceeding, you must have the roles/faulttesting.operator role.

To run an experiment:

1. Select Experiment templates from the left-hand navigation and choose a template.
   • (Optional) Enter a custom name for this experiment. If you don't enter an experiment name, it will use the auto-generated one. Example: "template_name_Sep_3_2025".
   • (Optional) Enter a description for this experiment.

2. Select Start Experiment.

   A details page showing the progress of the experiment appears, and the dry run of the experiment starts.

   When the dry run completes:

   • The target resources appear on the page.
   • The status indicates Prepared, meaning that you can start the fault injection. If an experiment is left in the Prepared state for more than 10 days, it is automatically canceled.

3. Select Start injection. The fault injection starts on the target resources, and you can see the progress.