Error modifying dms replication instance

The error you're encountering suggests that there might be a capacity issue in the Availability Zone where your DMS replication instance is currently located. This can happen when AWS doesn't have enough available resources to accommodate the requested instance type in that specific Availability Zone. Here are a few steps you can try to resolve this issue:

1. Try a different Availability Zone: If possible, attempt to modify the replication instance and select a different Availability Zone within the same region. This might help if the capacity issue is specific to the current AZ.

2. Wait and retry: Sometimes, capacity issues are temporary. You can wait for a short period (e.g., a few hours or a day) and try the modification again.

3. Consider Multi-AZ: Although you mentioned Single-AZ, switching to a Multi-AZ deployment might help, as it gives AWS more flexibility in resource allocation across multiple Availability Zones.

4. Use a different instance type: If the t3.small is not available, you could try a different instance type with similar or slightly higher specifications.

5. Create a new instance: As an alternative, you could create a new replication instance with the desired specifications (t3.small) in a different Availability Zone, then migrate your tasks to the new instance.

6. Contact AWS Support: If the issue persists, it might be worth reaching out to AWS Support for further assistance, as they can provide more specific guidance based on the current resource availability in your region.

Remember that when modifying a replication instance, you have the option to apply changes immediately or during the next scheduled maintenance window. Choosing to apply changes immediately might increase the chances of encountering capacity issues, so consider scheduling the change for the next maintenance window if your migration timeline allows for it.
Sources
Modifying a replication instance - AWS Database Migration Service
ModifyReplicationInstance - AWS Database Migration Service
Working with an AWS DMS replication instance - AWS Database Migration Service

This error occurs because of three known reasons:

1. If the IP addresses/free slots are not available on the subnet group.

2. If the availability zone does not support the instance class.

3. Unavailability of the instance class.

4. If the IP addresses/free slots are not available on the subnet group:

When you attach a subnet group on your replication instance, it contains the set of subnets that allow you to configure the instance in respective availability zones. The subnet contains a set of free IP addresses that you can use to configure the AWS resources in your account. There comes a point when there are no free slots available and that results in not allowing you to modify and use new AWS resources.

2. If the availability zone does not support the instance class:

Can execute the below AWS CLI command , the output shows supported availability zone result.

CLI command: aws dms describe-orderable-replication-instances
--query "OrderableReplicationInstances[?ReplicationInstanceClass=='dms.t3.small' && contains(EngineVersion, '3.5.2')]"
--region your_region_name

[+] describe-orderable-replication-instances: https://docs.aws.amazon.com/cli/latest/reference/dms/describe-orderable-replication-instances.html

3. Unavailability of the instance class:

This scenario is internal to AWS where at a given point in time the specific instance class, "dms.t3.small" would not be available to provide as there has been lot of demand during that time. It has nothing to do with free slots available at the customer end but the pool of requests that comes in for the class.

In such scenarios which is true in your case, trying out the modification of instance class upgrade again at a later point in time helps with the request. Alternate solution that AWS suggests you is to try and upgrade the instance to a different class first and then later you can perform modification to the intended class.

[+] Insufficient instance capacity: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/troubleshooting-launch.html#troubleshooting-launch-capacity

RECOMMENDATIONS

1. Launch a new replication instance with the desired capacity: The first option is to do as the error message suggests and launch a new replication instance with the required capacity. You can do so using the same subnet group as our automation will ensure that the replication instance is launched in the AZs where there is capacity available, or you can use a subnet group that does not include the AZ that does not currently have enough capacity.

2. Force an AZ switch for your current DMS replication instance: If you wish to keep the same replication instance, you can consider converting the same to a Multi-AZ deployment. This will launch a new secondary instance in a different AZ. After the second instance is launched, you can reboot the replication instance, making sure to do so with the option to fail over checked. This will make the new instance in the new AZ the primary instance. You can then remove the old primary instance from the deployment by converting the replication instance back to a Single-AZ deployment. You will then be able to change to the desired capacity in the new AZ where there is capacity available.

3. Scale your DMS instance to a different size/class and resize the instance at a later time, if needed.

Below I've also attached a few knowledge articles that discuss these capacity errors for your reference:

[+] https://repost.aws/knowledge-center/ec2-insufficient-capacity-errors