How do you find the cluster/service CIDR for a Kubernetes cluster, once it is already running?
I know for Minikube, it is 10.0.0.1/24.
For GKE, you can find out via
gcloud container clusters describe XXXXXXX --zone=XXXXXX |
grep -e clusterIpv4Cidr -e servicesIpv4Cidr
But how do you find out on a generic Kubernetes cluster, particularly via kubectl?
Get Services IPs range
kubectl cluster-info dump | grep -m 1 service-cluster-ip-range
You will see something like e.g. --service-cluster-ip-range=xxx.yy.0.0/13
Get Pods IPs range
kubectl cluster-info dump | grep -m 1 cluster-cidr
You will see something like e.g. --cluster-cidr=xxx.yy.0.0/11
I spent hours searching for a generic way to do this. I gave up searching and wrote my own. As of Kubernetes 1.18, this method works across cloud providers, beyond just GKE.
SVCRANGE=$(echo '{"apiVersion":"v1","kind":"Service","metadata":{"name":"tst"},"spec":{"clusterIP":"1.1.1.1","ports":[{"port":443}]}}' | kubectl apply -f - 2>&1 | sed 's/.*valid IPs is //')
echo $SVCRANGE
172.21.0.0/16
This one liner works by feeding an invalid service cluster IP into kubectl apply and parsing the error output, which provides the service CIDR information.
With kubeadm
kubeadm config view | grep Subnet
Below output is self explanatory
podSubnet: 10.10.0.0/16
serviceSubnet: 10.96.0.0/12
If you did not specify --pod-network-cidr or --service-cidr, the defaults are used. Get defaults:
kubeadm config print init-defaults
Or get the configmap:
kubectl --namespace kube-system get configmap kubeadm-config -o yaml
Similarly as proposed before, but with only the CIDR returned:
POD_CIDR=$(kubectl cluster-info dump | grep -m 1 -Po '(?<=--cluster-cidr=)[0-9.\/]+')
SERVICE_RANGE=$(kubectl cluster-info dump | grep -m 1 -Po '(?<=--service-cluster-ip-range=)[0-9.\/]+')