If your Amazon EKS cluster has been stuck in the DELETING state for more than two weeks, don’t worry—you can take steps to resolve the issue. Here’s a simple guide to help you troubleshoot and complete the deletion process.
First, check if there are any remaining resources linked to your cluster. Make sure there are no managed node groups, Fargate profiles, EC2 worker nodes, VPC endpoints, or updates in progress. You can verify this using the AWS CLI commands:
-
To list node groups:
aws eks list-nodegroups –cluster-name your-cluster-name –region your-region
-
To list Fargate profiles:
aws eks list-fargate-profiles –cluster-name your-cluster-name –region your-region
-
To check for ongoing updates:
aws eks list-updates –name your-cluster-name –region your-region
If these commands return empty lists, then your cluster is not holding onto any resources via these common channels.
Next, examine the cluster’s description to identify any remaining resources. Use:
aws eks describe-cluster –name your-cluster-name –region your-region
In the output, look for leftover Network Interfaces (ENIs), especially those labeled as AWS-managed control plane interfaces. In some cases, these ENIs might still be marked with RequesterManaged = true and Status = in-use.
If you try to delete these ENIs manually, you might encounter errors like “Network interface is currently in use.” This indicates that the interfaces are still active and cannot be removed automatically. The only resource remaining could be the security group used by the control plane, which might only be associated with these ENIs.
Here’s what you can do:
-
Contact AWS Support: Since control plane ENIs are managed resources, the safest and most reliable route is to ask AWS Support for assistance. They have the tools to forcefully remove residual resources that can’t be deleted manually and can help complete the cluster deletion.
-
Verify no other resources are attached or lingering in your AWS account that could be preventing cleanup.
It’s a common issue that sometimes the control plane resources linger due to internal dependencies not visible to users. While it might seem that nothing else remains, AWS Support is the best resource for resolving these issues smoothly and safely.
Be cautious about attempting to delete network interfaces or security groups manually beyond what’s described, as it could impact other resources.
In summary, the key steps involve confirming no active resources are left, and if the cluster still won’t delete, reaching out to AWS Support for their intervention. They can assist in cleaning up those fine-grained resources and ensuring your environment is fully cleared.
Good luck, and don’t hesitate to ask for help from the AWS support team—they are there to assist with exactly this kind of situation.