You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Tell us about your request
I'm trying to use "IAM Roles for Service Accounts" in private VPC without Internet, Nat Gateways. It works fine in VPC with internet access.
Which service(s) is this request for?
EKS
Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
When I deploy pod with a service account in spec everything looks fine, pod is mutating and the environment variables and necessary setting are automatically injected. Then I'm accessing pod via kubectl exec and run command
aws s3 ls --region eu-west-1 --debug
or any other and I see that it's trying to connect to global sts Starting new HTTPS connection (1): sts.amazonaws.com:443 and failed with a timeout. How can I force it to use my regional sts vpc endpoint like sts.eu-west-1.amazonaws.com?
Are you currently working around this issue?
I even tried to add cname record in coredns but recived
botocore.exceptions.SSLError: SSL validation failed for https://sts.amazonaws.com/ ("hostname 'sts.amazonaws.com' doesn't match either of 'sts.eu-west-1.amazonaws.com', '*.sts.eu-west-1.vpce.amazonaws.com'",)
The text was updated successfully, but these errors were encountered:
@alexey-boyko Can you retry? I've recreated the same, and was able to get IAM roles for service accounts working in a private VPC without outbound internet ("air-gapped").
I think this PR needs to be merged for cleaner integration.
I was able to use VPC endpoints for STS only after updating AWS_STS_REGIONAL_ENDPOINTS=regional on my workload
Tell us about your request
I'm trying to use "IAM Roles for Service Accounts" in private VPC without Internet, Nat Gateways. It works fine in VPC with internet access.
Which service(s) is this request for?
EKS
Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?
When I deploy pod with a service account in spec everything looks fine, pod is mutating and the environment variables and necessary setting are automatically injected. Then I'm accessing pod via kubectl exec and run command
or any other and I see that it's trying to connect to global sts
Starting new HTTPS connection (1): sts.amazonaws.com:443
and failed with a timeout. How can I force it to use my regional sts vpc endpoint like sts.eu-west-1.amazonaws.com?Are you currently working around this issue?
I even tried to add cname record in coredns but recived
The text was updated successfully, but these errors were encountered: