Select your cookie preferences

We use essential cookies and similar tools that are necessary to provide our site and services. We use performance cookies to collect anonymous statistics, so we can understand how customers use our site and make improvements. Essential cookies cannot be deactivated, but you can choose “Customize” or “Decline” to decline performance cookies.

If you agree, AWS and approved third parties will also use cookies to provide useful site features, remember your preferences, and display relevant content, including relevant advertising. To accept or decline all non-essential cookies, choose “Accept” or “Decline.” To make more detailed choices, choose “Customize.”

Zonal autoshift components

Focus mode
Zonal autoshift components - Amazon Application Recovery Controller (ARC)

The following diagram illustrates an example of an autoshift shifting traffic away from an Availability Zone. AWS starts an autoshift when internal telemetry indicates that there is an Availability Zone impairment that could potentially impact customers.

Diagram of an autoshift with three Availability Zones

The following are components of the zonal autoshift capabilities in ARC.

Zonal autoshift

Zonal autoshift shifts traffic away for a resource, without requiring you to take any action. Zonal autoshift is a capability in ARC where AWS starts an autoshift when internal telemetry indicates that there is an Availability Zone impairment that could potentially impact customers. Be aware that, in some cases, resources might be shifted away that are not experiencing impact.

Practice runs

When you enable zonal autoshift for a resource, you must also configure zonal autoshift practice runs for the resource. AWS performs a zonal shift for practice runs about weekly, for about 30 minutes. Practice runs make sure that your application can run normally with the loss of one Availability Zone. In a practice run, AWS shifts traffic for a resource away from one Availability Zone with a zonal shift, and then shifts traffic back when the practice run ends.

Practice run configuration

A practice run configuration defines the blocked dates and windows, if any, and the CloudWatch alarms that you specify for the practice run for a resource in zonal autoshift. You can edit a practice run at any time, to add or change blocked dates or windows, or to update the alarms for the practice run.

To enable zonal autoshift, you must have a practice run configuration in place for a resourceyou can also delete a practice run. To delete a practice run configuration for a resource, zonal autoshift must be disabled.

Practice run alarm

When you configure practice runs, you specify CloudWatch alarms that you create in CloudWatch, based on your resource and application requirements. The alarms that you specify can block a practice run from starting, or can stop a practice run in progress, if your application is adversely affected by the practice run.

If an alarm that you specify goes into an ALARM state, ARC ends the zonal shift for the practice run, so that traffic for the resource is no longer shifted away from the Availabilty Zone.

There are two types of alarms that you specify for practice runs: an outcome alarm, to monitor the health of your resource and application during the practice run, and a blocking alarm, which you can configure to prevent practice runs from starting, or to stop an in-progress practice run. The outcome alarm is required; the blocking alarm is optional.

Practice run outcome

ARC reports an outcome for each practice run. The following are the possible practice run outcomes:

  • PENDING: The zonal shift for the practice run is active (in progress). There's no outcome to return yet.

  • SUCCEEDED: The outcome alarm did not enter an ALARM state during the practice run, and the practice run completed the full 30 minute test period.

  • INTERRUPTED: The practice run ended for a reason that was not the outcome alarm entering an ALARM state. A practice run can be interrupted for a variety of reasons. For example, a practice run that ends because the blocking alarm specified for the practice run entered an ALARM state has an outcome of INTERRUPTED. For more information about reasons for an INTERRUPTED outcome, see Outcomes for practice runs.

  • FAILED: The outcome alarm entered an ALARM state during the practice run.

Built-in safety rules

Safety rules built into ARC prevent more than one traffic shift for a resource from being in effect at a time. That is, only one customer-initiated zonal shift, practice run zonal shift, or autoshift for the resource can be actively shifting traffic away from an Availability Zone. For example, if you start a zonal shift for a resource when it is currently shifted away with autoshift, your zonal shift takes precedence. For more information, see Outcomes for practice runs.

Resource identifier

The identifier for a resource to enable zonal autoshift for, which is the Amazon Resource Name (ARN) for the resource.

You can only enable zonal autoshift for resources in your account that are in an AWS service that is supported by ARC.

Note

You can only configure zonal autoshift for Application Load Balancers with cross-zone load balancing turned off.

Managed resource

Application Load Balancers register resources automatically with ARC for zonal autoshift. You must manually opt-in Network Load Balancer resources for zonal autoshift.

Resource name

The name of a managed resource in ARC.

Applied status

An applied status indicates whether a traffic shift is in effect for a resource. When you configure zonal autoshift, a resource can have more than one active traffic shift—that is, a practice run zonal shift, customer-initiated zonal shift, or autoshift. However, only one is applied, that is, is in effect for the resource at a time. The shift that has the status APPLIED determines the Availability Zone where application traffic has been shifted away for a resource, and when that traffic shift ends.

PrivacySite termsCookie preferences
© 2025, Amazon Web Services, Inc. or its affiliates. All rights reserved.