eksctl
eksctl copied to clipboard
[Bug] Can't reuse instanceRoleARN in multiple nodegroups with AccessEntry
What were you trying to accomplish?
With the switch to access entries, creating multiple self managed nodes using the same instanceRoleARN no longer works.
What happened?
The first node successfully gets created however subsequent nodes fail to create due to the AccessEntry already existing.
How to reproduce it?
$ vi cluster.yaml
apiVersion: eksctl.io/v1alpha5
kind: ClusterConfig
metadata:
name: test
region: us-west-2
nodeGroups:
- name: ng-1
instanceType: m5.large
desiredCapacity: 1
volumeSize: 80
iam:
instanceRoleARN: "arn:aws:iam::999999999999:role/myRole"
- name: ng-2
instanceType: m5.large
desiredCapacity: 1
volumeSize: 80
iam:
instanceRoleARN: "arn:aws:iam::999999999999:role/myRole"
$ eksctl create cluster -f cluster.yaml
// fails
Logs
2024-01-24 20:11:48 [!] 1 error(s) occurred and cluster hasn't been created properly, you may wish to check CloudFormation console
From CFN stack:
Logical ID: AccessEntry
Status: CREATE_FAILED
Status reason: arn:aws:iam::999999999999:role/myRole|test already exists in stack <other_nodegroup_stack>
Anything else we need to know?
Versions
$ eksctl info
eksctl version: 0.169.0
kubectl version: v1.22.0
OS: linux
Hello DanielCKennedy :wave: Thank you for opening an issue in eksctl
project. The team will review the issue and aim to respond within 1-5 business days. Meanwhile, please read about the Contribution and Code of Conduct guidelines here. You can find out more information about eksctl
on our website
Hello Team .
We are encountering the same issue, which has led us to refrain from using EKS access entries. As a workaround, we have downgraded the eksctl version to a lower one from 0.167.0.
Looks like 0.166.0 does not have this problem. Thanks for calling it out @ajaydevtron !
Can we get real fix for this issue, using 0.166.0 is the workaround but it won't last longer? Thank you for support.