Error when applying Calico network on Kubernetes cluster

11/22/2019

I did kubeinit using the following command (I am trying to setup Kubernetes on RHEL 7.6)

kubeadm init --apiserver-advertise-address=15.217.230.99 --pod-network-cidr=15.232.10.195/27

I want to use the calico network. Since I can't use 192.168.0.0/16 network, I had to wget the calico.yaml from https://docs.projectcalico.org/v3.9/manifests/calico.yaml and then modify CALICO_IPV4POOL_CIDR to have the value 15.232.10.195/27 (First thing is, I don't know if I am doing it correctly here. I am very new to Kubernetes and trying to setup my first ever cluster) When I try to apply the file using command (as a sudo user):

kubectl apply -f ./calico.yaml

I get the following error:

unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused
unable to recognize "./calico.yaml": Get http://localhost:8080/api?timeout=32s: dial tcp [::1]:8080: connect: connection refused

My api server runs on port 6443. That is what I see in my kubeadm join token generated by kubeadm init.

Can someone please correct me where I am making mistakes? Is it ok to use any other mask with calico network than 192.168.0.0/16? I can't use that since it is already being used in our network.

I also want to join Windows nodes in addition to linux nodes on my cluster. Is Calico network a correct approach OR recommendation is something else instead. I would like to know before I initialize the network on my cluster so that I can do the right thinhs

Thanks

-- Andy Johnson
kubeadm
kubectl
kubernetes

2 Answers

1/19/2020

Follow the below steps to overcome this issue:

mkdir -p $HOME/.kube

sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config

sudo chown $(id -u):$(id -g) $HOME/.kube/config
-- Anbazhagan
Source: StackOverflow

11/22/2019

After struggling for several hours, I found that kubectl apply command needs to run as regular user and not a root user. Make sure that you are executing the command for a user where we copied the .kube/config file OR it will not work After doing that, I see that my master is now ready.

-- Andy Johnson
Source: StackOverflow