Experiments for AWS FIS
AWS FIS enables you to perform fault injection experiments on your AWS workloads. To get started, create an experiment template. After you create an experiment template, you can use it to start an experiment.
An experiment is finished when one of the following occurs:
-
All actions in the template completed successfully.
-
A stop condition is triggered.
-
An action cannot be completed because of an error. For example, if the target cannot be found.
-
The experiment is stopped manually.
You cannot resume a stopped or failed experiment. You also cannot rerun a completed experiment. However, you can start a new experiment from the same experiment template. You can optionally update the experiment template before you specify it again in a new experiment.
Tasks
Start an experiment
You start an experiment from an experiment template. For more information, see Start an experiment from a template.
You can schedule your experiments as a one-time task or recurring tasks using Amazon EventBridge. For more information, see Tutorial: Schedule a recurring experiment.
You can monitor your experiment using any of the following features:
-
View your experiments in the AWS FIS console. For more information, see View your experiments.
-
View Amazon CloudWatch metrics for the target resources in your experiments or view AWS FIS usage metrics. For more information, see Monitor using CloudWatch.
-
Enable experiment logging to capture detailed information about your experiment as it runs. For more information see Experiment logging.
View your experiments
You can view the progress of a running experiment, and you can view experiments that have completed, stopped, or failed.
Stopped, completed, and failed experiments are automatically removed from your account after 120 days.
To view experiments using the console
Open the AWS FIS console at https://console.aws.amazon.com/fis/
. -
In the navigation pane, choose Experiments.
-
Choose the Experiment ID of the experiment to open its details page.
-
Do one or more of the following:
-
Check Details, State for the state of the experiment.
-
Choose the Actions tab for information about the experiment actions.
-
Choose the Targets tab for information about the experiment targets.
-
Choose the Timeline tab for a visual representation of the actions based on their start and end times.
-
To view experiments using the CLI
Use the list-experiments
Experiment states
An experiment can be in one of the following states:
-
pending – The experiment is pending.
-
initiating – The experiment is preparing to start.
-
running – The experiment is running.
-
completed – All actions in the experiment completed successfully.
-
stopping – The stop condition was triggered or the experiment was stopped manually.
-
stopped – All running or pending actions in the experiment are stopped.
-
failed – The experiment failed due to an error, such as insufficient permissions or incorrect syntax.
Action states
An action can be in one of the following states:
-
pending – The action is pending, either because the experiment hasn't started or the action is to start later in the experiment.
-
initiating – The action is preparing to start.
-
running – The action is running.
-
completed – The action completed successfully.
-
cancelled – The experiment stopped before the action started.
-
skipped – The action has been skipped.
-
stopping – The action is stopping.
-
stopped – All running or pending actions in the experiment are stopped.
-
failed – The action failed due to a client error, such as insufficient permissions or incorrect syntax.
Tag an experiment
You can apply tags to experiments to help you organize them. You can also implement tag-based IAM policies to control access to experiments.
To tag an experiment using the console
Open the AWS FIS console at https://console.aws.amazon.com/fis/
. -
In the navigation pane, choose Experiments.
-
Select the experiment and choose Actions, Manage tags.
-
To add a new tag, choose Add new tag, and specify a key and value.
To remove a tag, choose Remove for the tag.
-
Choose Save.
To tag an experiment using the CLI
Use the tag-resource
Stop an experiment
You can stop a running experiment at any time. When you stop an experiment, any post actions that have not been completed for an action are completed before the experiment stops. You cannot resume a stopped experiment.
To stop an experiment using the console
Open the AWS FIS console at https://console.aws.amazon.com/fis/
. -
In the navigation pane, choose Experiments.
-
Select the experiment, and choose Stop experiment.
-
In the confirmation dialog box, choose Stop experiment.
To stop an experiment using the CLI
Use the stop-experiment
List resolved targets
You can view information for resolved targets for an experiment after target resolution has ended.
To view resolved targets using the console
Open the AWS FIS console at https://console.aws.amazon.com/fis/
. -
In the navigation pane, choose Experiments.
-
Select the experiment, and choose Report.
-
View resolved target information under Resources.
To view resolved targets using the CLI
Use the list-experiment-resolved-targets