You cannot move an existing EC2 instance in place to another Availability Zone (AZ). The supported method is to create an Amazon Machine Image (AMI) from the source instance, launch a replacement in a subnet in the destination AZ, redirect networking, verify the replacement, and only then retire the original.
For a move within one Region, launch directly from the AMI in the target AZ. A move to another Region requires copying the AMI first. Expect a new private IPv4 address; a same-Region Elastic IP can be reassigned after the replacement is ready.
As an Amazon Associate I earn from qualifying purchases.
What “moving an EC2 instance” actually means
An EC2 instance is created in a specific subnet, and every subnet belongs to one Availability Zone. AWS does not provide an action that changes that placement after launch. AWS Support states: “You can’t move an existing EC2 instance to another subnet, Availability Zone, or VPC.”
The practical operation is therefore a controlled replacement:
#1 Best Overall
- Prepare the workload and its dependencies.
- Capture the operating system and EBS-backed configuration in an AMI.
- Launch a new instance from that AMI in a subnet in the destination AZ.
- Reconnect addresses, DNS, storage, monitoring, and integrations.
- Run acceptance checks while the source remains available for rollback.
- Stop or terminate the source only after the cutover is accepted.
This approach preserves the installed software and most instance configuration, but it does not preserve the original instance identity, subnet placement, private IP address, or any data that was never included in the image or a separate backup.
Plan the migration before creating an image
Make an inventory of everything that makes the instance reachable and useful. Record the source Region and AZ, destination AZ and subnet, instance type, key pair, security groups, IAM instance profile, attached EBS volumes, Elastic IP associations, DNS records, monitoring alarms, backup policies, scheduled jobs, and boot-time scripts.
Choose a cutover model
If the application can tolerate a brief interruption, stop the instance or quiesce the application before imaging it. This gives you a more consistent disk state. For a database or a write-heavy service, use the application’s own flush, pause, or replication procedure in addition to an EC2 stop when appropriate. If you create the AMI while writes continue, plan how to reconcile changes made after the image starts.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsConfirm destination capacity
Verify that the destination subnet has enough IPv4 addresses and that the requested instance type is offered in the target AZ. Confirm that the destination VPC has the required security groups, route tables, internet or NAT paths, and endpoints. If the replacement will be in a different VPC, security groups from the source VPC cannot be reused; create equivalent groups in the destination VPC.
Protect EBS and instance-store data
EBS volumes
EBS volumes are Availability-Zone scoped. A volume attached to an instance in one AZ cannot be attached directly to an instance in another AZ. To reproduce an EBS data volume in the target AZ, create a snapshot and create a new volume from that snapshot in the destination AZ, then attach it to the replacement using the required device mapping.
Creating a snapshot before AMI creation can reduce the time needed to create the image. Snapshot and image operations are incremental after the initial copy, but the first operation still depends on the amount of data. For a large file system without an existing snapshot, AWS notes that AMI creation can take several hours.
Instance-store volumes
Instance-store data is ephemeral and is not automatically captured by an AMI. Copy required files to durable storage or another backup target before the migration, and document how they will be restored on the replacement. Do not assume that an AMI contains data held on instance-store devices.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Create the AMI
Using the EC2 console
- Open the EC2 console in the Region that contains the source instance.
- Choose Instances, select the source, then choose Actions > Image and templates > Create image.
- Give the image a name and description that identify the source, date, and intended migration.
- Review the EBS volume mappings. Include the boot volume and any persistent data volumes that should be reproduced.
- Choose whether the instance should reboot during image creation. A reboot generally improves file-system consistency; disabling it avoids an interruption but increases the risk of an inconsistent image for active workloads.
- Create the image and monitor its status under AMIs. Do not launch the replacement until the AMI is available.
Keep the source instance and its snapshots intact while the image is being tested. If the image is intended only for this migration, label it clearly so it is not confused with a long-term golden image.
Using Systems Manager Automation
AWS provides the AWSSupport-CopyEC2Instance Systems Manager Automation runbook for copying an EC2 instance. Run it in the same AWS Region as the existing instance. The runbook can standardize image and copy steps, which is useful when the operation must be repeatable or auditable, but you still choose and verify the destination launch settings and traffic cutover.
Same-Region versus cross-Region moves
| Migration | Image handling | Address implications | Additional considerations |
|---|---|---|---|
| Different AZ in the same Region | Launch the available AMI directly into a subnet in the destination AZ. | The private IPv4 address changes. A same-Region Elastic IP can be detached from the source and attached to the replacement. | Recreate EBS data volumes from snapshots in the target AZ when they are not part of the boot image. |
| Different Region | Copy the AMI to the destination Region, wait for the copied AMI to become available, then launch there. | The original Elastic IP cannot be reused across Regions. Plan a new address and update DNS or another global traffic mechanism. | Check that the instance type, AMI architecture, key pair, IAM configuration, security services, and dependent regional resources exist in the destination Region. AMI copying and data transfer can incur charges. |
AMIs are Region-specific. A same-Region AZ change does not require an AMI copy; a cross-Region migration does. Copying an encrypted image also requires the appropriate key permissions and a usable key in the destination Region.
Rank #3
Launch the replacement in the destination AZ
- In the EC2 console, select the AMI and choose Launch instance from image.
- Choose an instance type that is available in the target AZ and provides adequate CPU, memory, network, and storage performance.
- Under network settings, select the destination VPC and a subnet explicitly located in the target AZ. Do not rely on an automatic subnet choice when AZ placement matters.
- Select the correct key pair, security groups, IAM instance profile, and monitoring options.
- Review storage mappings. Use the AMI’s boot mapping and add any new EBS volumes created from snapshots. Preserve required device names if the operating system or application expects them.
- Apply the same tags used for cost allocation, backups, ownership, and automation. Update any tag that identifies the old AZ or instance ID.
- Launch the instance and note its new instance ID and private IPv4 address.
Unless you explicitly select an available address in the destination subnet, EC2 assigns a new private IPv4 address from that subnet’s CIDR range. A private address from the old subnet cannot simply be carried into the new subnet.
Free tools Windows power users keep installed
One-click scans. No signup required.
Prepare operating-system and identity settings
Linux and general boot configuration
Check cloud-init or other first-boot scripts for assumptions about the old instance ID, hostname, subnet, mounted device names, local routes, or AZ-specific endpoints. Confirm that the network interface obtained the expected address and that all required file systems mounted with the intended ownership and permissions.
Windows and domain membership
A cloned Windows instance can create identity conflicts when it contacts a directory service. AWS recommends using Sysprep or removing a domain-joined instance from the domain before creating the replacement, according to your organization’s domain procedure. Rejoin the replacement and verify computer-account, policy, certificate, and service permissions before directing production traffic to it.
Move traffic and addresses
Elastic IP
For a same-Region move, disassociate the Elastic IP from the old instance, then associate it with the replacement’s network interface or instance. Perform this during the cutover window because clients using that address will follow whichever instance currently owns it. An Elastic IP cannot be reused across Regions.
DNS and application references
Update records that point to the old private address, public address, hostname, or instance ID. Check service-discovery entries, database allowlists, firewall rules, monitoring targets, deployment pipelines, license bindings, and downstream applications. Lowering DNS TTL ahead of the cutover can reduce caching time, but clients and resolvers may still retain the old value until their existing TTL expires.
If the application uses a load balancer, register the replacement target, wait for its health checks to pass, then drain or deregister the old target. If traffic is controlled by a static allowlist, add the replacement address before removing the old one so that validation access is not lost.
Validate the replacement before retiring the source
Use an explicit acceptance checklist rather than relying on the EC2 console showing a running state.
- Instance and system status: both status checks pass, the expected hostname is set, and the operating system has no boot or driver errors.
- Storage: every required EBS volume is present in the target AZ, mounted at the correct path, and readable and writable by the intended service account.
- Application health: the service starts automatically, responds to local and external health checks, and can read its configuration and dependent data.
- Network paths: security groups, network ACLs, routes, NAT or internet paths, private endpoints, and outbound allowlists behave as expected.
- Identity and permissions: IAM access, domain membership, certificates, secrets, and operating-system permissions work from the new instance.
- Operations: logs arrive, alarms evaluate, backups include the replacement, scheduled jobs run once, and management agents report the new instance.
- Data integrity: compare application records, transaction state, and recent files with the source or authoritative datastore.
Keep the source instance powered on or otherwise recoverable until these checks pass and your rollback window closes. If the source was stopped for imaging, do not terminate it merely because the replacement reached the running state.
Retire the old instance and clean up safely
- Confirm that production traffic no longer reaches the source and that the replacement has completed at least one successful backup or snapshot cycle.
- Follow the rollback plan’s waiting period. Stop the source first when you want a reversible hold; terminate it only when the business owner accepts the cutover.
- Keep the migration AMI and its snapshots for the required retention period. Deregister the AMI and delete associated snapshots only after rollback and retention requirements are satisfied.
- Remove unattached temporary EBS volumes, stale network interfaces, obsolete alarms, and old DNS or allowlist entries after confirming that no automation still references them.
Snapshot storage, AMI or snapshot copying, and data transfer may generate AWS charges. The duration and cost depend on data size, whether an existing snapshot can be reused, the copy mode, and the Regions involved.
Manual migration or Systems Manager runbook?
| Consideration | Manual EC2 workflow | AWSSupport-CopyEC2Instance runbook |
|---|---|---|
| Control over launch settings | Maximum control over subnet, instance type, storage, tags, security groups, and cutover sequence. | Automates the copy workflow, while destination launch choices and validation still require review. |
| Automation coverage | Each image, launch, networking, and cleanup action is performed and recorded by the operator or separate scripts. | Provides a repeatable Systems Manager Automation execution for the supported copy operation. |
| Auditability | Evidence is spread across console activity, change tickets, and operator notes unless you build additional logging. | Execution history is visible in Systems Manager, which can simplify operational review. |
| Rollback handling | Easy to keep the source untouched while you test, provided the operator follows a hold-and-verify plan. | Still requires a deliberate source-retention and traffic-rollback plan; automation does not decide when the migration is safe. |
| Best fit | One-off moves with unusual storage, networking, or application requirements. | Repeatable or standardized migrations where a documented automation runbook is valuable. |
Troubleshoot common migration failures
The replacement will not launch in the requested AZ
Verify that the subnet selected in the launch wizard belongs to the intended AZ, that the instance type is offered there, and that the subnet has available IPv4 addresses. A Region-wide instance type choice does not guarantee capacity in every AZ.
Best Value
A data volume is missing or will not attach
Check whether it was included in the AMI’s block-device mappings. If it was an independent EBS volume, create a snapshot-derived volume in the target AZ and attach it there. An EBS volume from the source AZ cannot be attached directly across AZ boundaries.
The application starts but cannot reach dependencies
Compare the replacement’s security groups, route tables, network ACLs, DNS resolver settings, proxy configuration, and allowlists with the source. Then inspect application logs for references to the old private IP, hostname, AZ, or instance ID.
Clients still reach the old instance
Check Elastic IP association, load-balancer target registration, DNS records, service discovery, and cached DNS TTLs. Do not terminate the source until you can prove that active traffic has drained or is intentionally being served by it as a rollback target.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Windows clients or services reject the replacement
Review the Sysprep or domain-removal step, computer-account status, certificates, group policies, and service permissions. A cloned domain identity can cause authentication or policy failures even when the operating system boots normally.
Quick Recap
Key decisions at a glance
| Requirement | Supported approach |
|---|---|
| Change AZ without changing the instance | Not supported; create and launch a replacement. |
| Move within one Region | Create an AMI, launch it in a subnet in the target AZ, and reassign a same-Region Elastic IP if used. |
| Move to another Region | Copy the AMI to the destination Region, then launch and configure a new instance there. |
| Keep the same private IPv4 address | Not generally possible across subnets; update DNS and application references. |
| Preserve EBS data | Include volumes in the image or recreate independent volumes from snapshots in the target AZ. |
| Preserve instance-store data | Back it up and restore it separately; AMI creation does not capture it automatically. |
| Reduce cutover risk | Quiesce or stop as appropriate, retain the source, validate comprehensively, and clean up only after rollback is no longer required. |
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
