[Bug] Regression: eksctl 0.183.0 fails to create managed nodegroup in AWS China

[artem-nefedov]

What happened?

Using eksctl 0.183 and EKS 1.30, it will fail to create managed nodegroup in AWS China due to stack wait timeout.

How to reproduce it?

Create EKS 1.30 cluster in aws-cn with 1 managed nodegroup. Nodegroup uses amd64 spot instances with types ["t3.large","t3a.large"].

Logs

2024-06-24 10:10:20 [ℹ]  eksctl version 0.183.0
2024-06-24 10:10:20 [ℹ]  using region cn-north-1
2024-06-24 10:10:20 [ℹ]  setting availability zones to [cn-north-1a cn-north-1b]
2024-06-24 10:10:20 [ℹ]  subnets for cn-north-1a - public:192.168.0.0/19 private:192.168.64.0/19
2024-06-24 10:10:20 [ℹ]  subnets for cn-north-1b - public:192.168.32.0/19 private:192.168.96.0/19
2024-06-24 10:10:20 [ℹ]  nodegroup "managed-ng-1" will use "" [AmazonLinux2023/1.30]
2024-06-24 10:10:20 [ℹ]  using Kubernetes version 1.30
2024-06-24 10:10:20 [ℹ]  creating EKS cluster "redacted" in "cn-north-1" region with managed nodes
2024-06-24 10:10:20 [ℹ]  1 nodegroup (managed-ng-1) was included (based on the include/exclude rules)
2024-06-24 10:10:20 [ℹ]  will create a CloudFormation stack for cluster itself and 0 nodegroup stack(s)
2024-06-24 10:10:20 [ℹ]  will create a CloudFormation stack for cluster itself and 1 managed nodegroup stack(s)
... <abbreviated>
2024-06-24 10:45:55 [ℹ]  waiting for CloudFormation stack "eksctl-redacted-nodegroup-managed-ng-1"
2024-06-24 10:46:28 [ℹ]  waiting for CloudFormation stack "eksctl-redacted-nodegroup-managed-ng-1"
2024-06-24 10:46:28 [!]  1 error(s) occurred and cluster hasn't been created properly, you may wish to check CloudFormation console
2024-06-24 10:46:28 [ℹ]  to cleanup resources, run 'eksctl delete cluster --region=cn-north-1 --name=redacted'
2024-06-24 10:46:28 [✖]  exceeded max wait time for StackCreateComplete waiter

Anything else we need to know?

Problem is only reproduced in AWS China (global works normally). Downgrading to eksctl 0.180.0 resolves the problem.

Versions

eksctl 0.183.0
EKS 1.30