DNS Connectivity Issues for Pods in Amazon EKS Cluster

Question

Hello,

I'm currently facing an issue with Pods running inside an Amazon EKS cluster. These Pods need to connect to a few internal on-premises services from one of the microservices deployed on our AWS account, which is connected via DirectConnect. Unfortunately, I'm encountering some connectivity problems related to DNS.

When attempting to run a curl command to the internal URL from within the Pod, I encounter the following error:
Error: getaddrinfo ENOTFOUND

As an alternative, I tried using the IP address of the URL, and it worked. However, I had to use the -k and -v curl options to make it work.

I would greatly appreciate any advice or guidance on how to resolve this DNS issue. Your expertise would be invaluable in helping me understand and overcome this challenge. Thank you in advance for your assistance.

Gary Mclean · Accepted Answer

If you are using the AWS route53 vpc resolver then you will need to setup route53 outbound endpoints in the vpc of where your eks is.

Then you will need to create a rule for the dns domain you are trying to resolve along with any domains which are included in cname records.

The rules define the domains, endpoints and name servers to resolve on prem host names.

You can instead have these in a central network account as such and share them to each vpc using RAM.

https://docs.aws.amazon.com/Route53/latest/DeveloperGuide/resolver-forwarding-outbound-queries.html

DNS Connectivity Issues for Pods in Amazon EKS Cluster

Relevant content