Adjust CoreDNS Configuration in Kubernetes

FPT Managed Kubernetes Engine (M-FKE) uses CoreDNS to manage and resolve domains for the cluster. Since CoreDNS is managed by FPT, users cannot modify the CoreDNS Corefile directly. This guide shows how to use a ConfigMap to adjust the desired CoreDNS configuration in M-FKE.

Rewrite DNS

You can configure CoreDNS in M-FKE to perform DNS rewriting by configuring the coredns-custom ConfigMap in the kube-system namespace as follows:

apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns-custom
  namespace: kube-system
data:
  test.server: |
    .com:8053 {
    log
    errors
    rewrite stop {
      name regex (.*)\..com {1}.default.svc.cluster.local
      answer name (.*)\.default\.svc\.cluster\.local {1}..com
    }
    forward . /etc/resolv.conf # you can redirect this to a specific DNS server such as 10.0.0.10, but that server must be able to resolve the rewritten domain name
    }

After applying, delete the 2 CoreDNS pods in the kube-system namespace to reload the ConfigMap configuration for CoreDNS.

Custom forward server

If you need to specify a DNS server to resolve domains in the cluster, configure as follows:

apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns-custom
  namespace: kube-system
data:
  test.server: | # you may select any name here, but it must end with the .server file extension
    .com:8053 {
        forward foo.com 1.1.1.1
        forward . 2.2.2.2
    }

Where 1.1.1.1 and 2.2.2.2 are DNS servers you define. forward . 2.2.2.2 means CoreDNS will send resolution requests for *.com domains to DNS server 2.2.2.2.

After applying, delete the 2 CoreDNS pods in the kube-system namespace to reload the ConfigMap configuration.

Custom domains

If you want to configure a custom domain that can only be resolved internally within the cluster (for example, resolving the domain puglife.local, which is not a valid public domain), without a coredns-custom ConfigMap the M-FKE cluster will not be able to resolve it. Configure as follows:

apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns-custom
  namespace: kube-system
data:
  puglife.server: | # you may select any name here, but it must end with the .server file extension
    puglife.local:8053 {
        errors
        cache 30
        forward . 192.11.0.1  # this is my test/dev DNS server
    }

After applying, delete the 2 CoreDNS pods in the kube-system namespace to reload the ConfigMap configuration.

Stub domain

CoreDNS can be configured with a stub domain as follows. Make sure to update the custom domain and IP address to match the values in your infrastructure environment:

apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns-custom
  namespace: kube-system
data:
  test.server: | # you may select any name here, but it must end with the .server file extension
    abc.com:8053 {
     errors
     cache 30
     forward . 1.2.3.4
    }
    my.cluster.local:8053 {
        errors
        cache 30
        forward . 2.3.4.5
    }

After applying, delete the 2 CoreDNS pods in the kube-system namespace to reload the ConfigMap configuration.

Hosts plugin

You can use CoreDNS to add custom DNS records as follows:

apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns-custom # this is the name of the configmap you can overwrite with your changes
  namespace: kube-system
data:
    test.override: | # you may select any name here, but it must end with the .override file extension
          hosts { 
              10.0.0.1 example1.org
              10.0.0.2 example2.org
              10.0.0.3 example3.org
              fallthrough
          }

After applying, delete the 2 CoreDNS pods in the kube-system namespace to reload the ConfigMap configuration.

Enable DNS query logging

You can use DNS logging to help debug domain resolution in CoreDNS by adding the following configuration to the coredns-custom ConfigMap in the kube-system namespace:

apiVersion: v1
kind: ConfigMap
metadata:
  name: coredns-custom
  namespace: kube-system
data:
  log.override: | # you may select any name here, but it must end with the .override file extension
        log

After applying, delete the 2 CoreDNS pods in the kube-system namespace to reload the ConfigMap configuration.