I have my kubernetes cluster setup and I want to check for the nodes from worker/minion node, can we run kubectl form worker/minion node?
Yes, you just need to have the proper client credentials and you can run kubectl from anywhere that has network access to the apiserver. See Sharing Cluster Access with kubeconfig for the instructions to get a kubeconfig file onto your worker node.
Related
I have deployed 1 master and 3 nodes on VM's.
I can run successfully "kubectl" command on the server's SSH CLI. I can deploy pods, all fine.
But I couldn't find how can I run "kubectl" command from my local and manage the K8S cluster? How can I do that?
Thanks!
You might have a kubeconfig file somewhere on the VMs. You can copy this one to your local device under $HOME/.kube/config, so kubectl knows how to access the cluster.
For more information, see the kubernetes documentation.
From your local machine run:
kubectl config get-contexts
Then run the below (replace cluster-name with the cluster name you want to communicate with):
kubectl config use-context cluster-name
If the cluster name you want to communicate with is not listed, it means you haven't got to context file to the cluster.
I did an SSH on a worker node inside the cluster and I run kubectl in there. I created a PV, a PVC and a deployment. I read on the documentation that PV is a cluster-wide object. My question is what happens in this case? In other words, Does running kubectl inside a worker node has the same effect as running it from master node?
Short answer: yes. kubectl connects to the configured API server which controls the whole cluster.
I have created a Kubernetes cluster and one of instance in the cluster is inactive
I want to review the configured Kubernetes Engine cluster of an inactive configuration by which command should I check?
Should I use this "kubectl config get-contexts"?
or
kubectl config use-context and kubectl config view?
Am beginner to cloud please anyone explains?
The kubectl config get-context will not help you debug why the instance is failing. Basically it will just show you the list ot contexts. A context is a group of cluster access parameters. Each context contains a Kubernetes cluster, a user, and a namespace. The current context is the cluster that is currently the default for kubectl . On other hand the kubectl config view will just print you kubeconfig settings.
The best way to start is the Kubernestes official documentation. It provides a good basic steps for troubleshoouting your cluster. Some of the steps can be applied to GKE as well as the Kubeadm or Minikube clusters.
If you're using GKE, then you can read the nodes logs from Stackdriver. This document is excellent start when you want to check the logs directly in the log viewer.
If one of your instaces report NotReady after listing them with kubectl get nodes I suggest to ssh to that instances and check kubernetes components (kubelet and kube-proxy). You can view the GKE nodes from the instances page.
Kube Proxy logs:
/var/log/kube-proxy.log
If you want to check the kubelet logs, they're a unit in systemd in COS that can be accessed using jorunactl.
Kubelet logs:
sudo journalctl -u kubelet
For further debugging it is worth mentioning that that GKE master is a node inside a Google managed project and it is different from your cluster project.
For the detailed master logs you will have open a google support ticket. Here is more information about how GKE cluster architecture works, in case there's something related to the api-server.
Let me know if that was helpful.
You can run below command to check status of all the nodes of a kubernetes cluster. Pleases note if you are using GKE managed service you will not be able to see status of master nodes, you will only see status of worker nodes.
kubectl get nodes -o wide
kubectl describe node nodename
You can also run below command to check status of control plane components.
kubectl get componentstatus
You can use the below command to get list of all the nodes in GKE cluster:
kubectl get nodes -o wide
Once you have the list of nodes, you can describe the node to get the events"
kubectl describe node <Node-Name>
Based on the events you can debug the node.
I have two IP'S master node and worker node? I need to deploy some services using these. I don't know anything about kubernetes ,what is master node and worker node?
How do I start?
You should start from the very basic things..
Kubernetes concept page is your starting point.
The Kubernetes Master is a collection of three processes that run on a
single node in your cluster, which is designated as the master node.
Those processes are: kube-apiserver, kube-controller-manager and
kube-scheduler.
Each individual non-master node in your cluster runs
two processes: kubelet, which communicates with the Kubernetes Master.
kube-proxy, a network proxy which reflects Kubernetes networking
services on each node.
Regarding you question in comment: read Organizing Cluster Access Using kubeconfig Files. Make sure you have kubeconfig file in the right place..
I have deployment my kubernetes cluster using kubeadm.
Now I want to gather cluster based information like master node IP, port on which apiserver is listening and name of the cluster.
With kubectl cluster-info gives me some data but I am looking to fetch cluster level information with the help of K8s rest API.
One way which i have tried is look for apiserver pod and get the data. It's giving me cluster level data but I need some other cleaner way of doing it.
Thanks in advance!
If you have ran the apiserver, you can access the kubernetes REST API on port 8001.
One way to expose it is like this :
sudo kubectl proxy --address='0.0.0.0' --port=8001 --accept-hosts='^*$'&
then you can visit http://YOUR_VM_IP:8001/api
there you can see all the list of APIs and all the information you want.