WorkSpace Pools Auto Scale Failed

The issue you're experiencing with your AWS WorkSpace Pool failing to scale out and in could be related to permissions or configuration. While an IAM role specifically for the directory is not typically required for scaling WorkSpaces, there are a few important points to consider:

1. Service-Linked Role: When integrating WorkSpaces with Application Auto Scaling, AWS automatically creates a service-linked role named AWSServiceRoleForApplicationAutoScaling_WorkSpacesPool in your account. This role uses the managed policy AWSApplicationAutoscalingWorkSpacesPoolPolicy, which grants Application Auto Scaling the necessary permissions to interact with WorkSpaces and CloudWatch.

2. Scalable Target Registration: Before scaling policies can work, you need to ensure that your WorkSpaces pool is properly registered as a scalable target with Application Auto Scaling. This registration process defines the resource ID, scalable dimension, and namespace for your WorkSpaces pool.

3. Scaling Policy Configuration: If you've removed all scale-out and scale-in policies, you'll need to recreate them with the appropriate conditions and actions. These policies define when and how your WorkSpaces pool should scale.

4. CloudWatch Alarms: Auto Scaling for WorkSpaces relies on CloudWatch alarms to trigger scaling actions. Ensure that these alarms are properly configured and functioning.

5. Capacity Limits: Verify that your account has sufficient capacity limits to support the desired scaling operations.

To troubleshoot the issue:

1. Check the Application Auto Scaling console or use the AWS CLI to verify that your WorkSpaces pool is correctly registered as a scalable target.

2. Review the CloudWatch logs for any error messages related to scaling operations.

3. Ensure that the AWSServiceRoleForApplicationAutoScaling_WorkSpacesPool role exists in your account and has the correct permissions.

4. Recreate your scaling policies, ensuring they are properly configured with appropriate conditions and actions.

5. Test the scaling by manually adjusting the desired capacity and monitoring the results.

If the issue persists after verifying these points, it may be worth contacting AWS support for further assistance, as there could be account-specific factors affecting your WorkSpaces pool's ability to scale.

Sources
Amazon WorkSpaces and Application Auto Scaling - Application Auto Scaling
Auto scaling for WorkSpaces Pools - Amazon WorkSpaces
AWS managed policies for Application Auto Scaling - Application Auto Scaling

AutoScaling is done by the Application AutoScaling service. Since you have scaling policies created, this means the Workspace Pool should be registered as a Scalable Target already. Assuming that hasn't been removed during your testing, can you run the below command, which should show if scaling was attempted and failed, or if the service chose not to scale for some reason (filling in your resourceID)? https://docs.aws.amazon.com/autoscaling/application/userguide/application-auto-scaling-scaling-activities.html#include-not-scaled-activities-with-the-aws-cli

aws application-autoscaling describe-scaling-activities --include-not-scaled-activities \
  --service-namespace workspaces --scalable-dimension workspaces:workspacespool:DesiredUserSessions \
  --resource-id workspacespool/wspool-123456

This is assuming the CloudWatch alarms went into the ALARM state and triggered the scaling policy. If that never happened, then either the alarms were edited out of band, or the metric never crossed the threshold