aws.eks.Cluster opting into EKS Auto Mode with built in node pools is currently requiring the cluster to be replaced

[flostadler]

Describe what happened

Users should be able to opt into EKS Auto Mode with existing clusters without requiring the cluster to be recreated. This is possible without using the built in nodepools, but currently not possible if users utilize one of the built in nodepools (system, general-purpose)

The set behavior is:

• Enabling computeConfig.enabled and setting a computeConfig.nodeRoleArn is a valid in-place change
• Disabling computeConfig.enabled even with a computeConfig.nodeRoleArn specified is a valid in-place change. The role is essentially a noop at thi point
• With computeConfig.enabled, changing computeConfig.nodeRoleArn from one ARN value to another is not a valid in-place change; this change can only be made by re-placing the cluster

Currently, opting into EKS Auto Mode with an existing cluster and specifying a computeConfig.nodeRoleArn is wanting to replace the cluster since the nodeRoleArn value is changing (from undefined to defined). This should be a valid transition that does not require replacements.

This is caused by the following upstream bug: hashicorp/terraform-provider-aws#40411

Sample program

n/a

Log output

n/a

Affected Resource(s)

aws.eks.Cluster

Output of pulumi about

pulumi-aws v6.63.0

Additional context

No response

Contributing

Vote on this issue by adding a 👍 reaction.
To contribute a fix for this issue, leave a comment (and link to your pull request, if you've opened one already).

[flostadler]

The upstream issue has been fixed, waiting for the next upstream release to pull the fix

[flostadler]

Fixed by #4925