Skip to content

Custom VPC DNS

Due to the isolation of the user-defined VPC and the default VPC network, the coredns deployed in the default VPC cannot be accessed from within the custom VPC. If you wish to use the intra-cluster domain name resolution capability provided by Kubernetes within your custom VPC, you can refer to this document and utilize the vpc-dns CRD to do so.

This CRD eventually deploys a coredns that has two NICs, one in the user-defined VPC and the other in the default VPC to enable network interoperability and provide an internal load balancing within the custom VPC through the custom VPC internal load balancing.

Deployment of vpc-dns dependent resources

apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  labels:
    kubernetes.io/bootstrapping: rbac-defaults
  name: system:vpc-dns
rules:
  - apiGroups:
    - ""
    resources:
    - endpoints
    - services
    - pods
    - namespaces
    verbs:
    - list
    - watch
  - apiGroups:
    - discovery.k8s.io
    resources:
    - endpointslices
    verbs:
    - list
    - watch
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  annotations:
    rbac.authorization.kubernetes.io/autoupdate: "true"
  labels:
    kubernetes.io/bootstrapping: rbac-defaults
  name: vpc-dns
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: system:vpc-dns
subjects:
- kind: ServiceAccount
  name: vpc-dns
  namespace: kube-system
---
apiVersion: v1
kind: ServiceAccount
metadata:
  name: vpc-dns
  namespace: kube-system
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: vpc-dns-corefile
  namespace: kube-system
data:
  Corefile: |
    .:53 {
        errors
        health {
          lameduck 5s
        }
        ready
        kubernetes cluster.local in-addr.arpa ip6.arpa {
          pods insecure
          fallthrough in-addr.arpa ip6.arpa
        }
        prometheus :9153
        forward . /etc/resolv.conf {
          prefer_udp
        }
        cache 30
        loop
        reload
        loadbalance
    }

In addition to the above resources, the feature relies on the nat-gw-pod image for routing configuration.

Configuring Additional Network

apiVersion: "k8s.cni.cncf.io/v1"
kind: NetworkAttachmentDefinition
metadata:
  name: ovn-nad
  namespace: default
spec:
  config: '{
      "cniVersion": "0.3.0",
      "type": "kube-ovn",
      "server_socket": "/run/openvswitch/kube-ovn-daemon.sock",
      "provider": "ovn-nad.default.ovn"
    }'

Configuring Configmap for vpc-dns

Create a configmap under the kube-system namespace to configure the vpc-dns usage parameters that will be used later to start the vpc-dns function:

apiVersion: v1
kind: ConfigMap
metadata:
  name: vpc-dns-config
  namespace: kube-system
data:
  coredns-vip: 10.96.0.3
  enable-vpc-dns: "true"
  nad-name: ovn-nad
  nad-provider: ovn-nad.default.ovn
  • enable-vpc-dns:enable vpc dns feature, true as default
  • coredns-image:dns deployment image. Defaults to the clustered coredns deployment version
  • coredns-vip:The vip that provides lb services for coredns.
  • coredns-template:The URL where the coredns deployment template is located. defaults to the current version of the ovn directory. coredns-template.yaml default is https://raw.githubusercontent.com/kubeovn/kube-ovn/<kube-ovn version>/yamls/coredns-template.yaml.
  • nad-name:Configured network-attachment-definitions Resource name.
  • nad-provider:The name of the provider to use.
  • k8s-service-host:The ip used for coredns to access the k8s apiserver service, defaults to the apiserver address within the cluster.
  • k8s-service-port:The port used for coredns to access the k8s apiserver service, defaults to the apiserver port within the cluster.

Deploying vpc-dns

configure vpc-dns yaml:

kind: VpcDns
apiVersion: kubeovn.io/v1
metadata:
  name: test-cjh1
spec:
  vpc: cjh-vpc-1
  subnet: cjh-subnet-1
  replicas: 2
  • vpc : The name of the vpc used to deploy the dns component.
  • subnet:Sub-name for deploying dns components.
  • replicas: vpc dns deployment replicas

View information about deployed resources:

# kubectl get vpc-dns
NAME        ACTIVE   VPC         SUBNET   
test-cjh1   false    cjh-vpc-1   cjh-subnet-1   
test-cjh2   true     cjh-vpc-1   cjh-subnet-2 

ACTIVE : true Customized dns component deployed, false No deployment.

Restrictions: only one custom dns component will be deployed under a VPC

  • When multiple vpc-dns resources are configured under a VPC (i.e., different subnets for the same VPC), only one vpc-dns resource is in the state true``, and the others arefasle`.
  • When the true vpc-dns is removed, the other false vpc-dns will be obtained for deployment.

Validate deployment results

To view vpc-dns Pod status, use label app=vpc-dns to view all vpc-dns pod status:

# kubectl -n kube-system get pods -l app=vpc-dns
NAME                                 READY   STATUS    RESTARTS   AGE
vpc-dns-test-cjh1-7b878d96b4-g5979   1/1     Running   0          28s
vpc-dns-test-cjh1-7b878d96b4-ltmf9   1/1     Running   0          28s

View switch lb rule status information:

# kubectl -n kube-system get slr
NAME                VIP         PORT(S)                  SERVICE                             AGE
vpc-dns-test-cjh1   10.96.0.3   53/UDP,53/TCP,9153/TCP   kube-system/slr-vpc-dns-test-cjh1   113s

Go to the Pod under this VPC and test the dns resolution:

nslookup kubernetes.default.svc.cluster.local 10.96.0.3

The subnet where the switch lb rule under this VPC is located and the pods under other subnets under the same VPC can be resolved.

微信群 Slack Twitter Support Meeting

Comments