I have a mongodb cluster that I connect to it by port-forwarding the clusterip service on my local.
kubectl port-forward svc/mongodb-svc 27017:27017
I specify readPreference in the db client app, but the service always connects me to random node regardless of my read preferences. As primary node may change in the future, I don't want to create anyother service per node either. So my current workaround is connecting to the pod directly
kubectl port-forward pod/mongodb-0 27017:27017
But this is not an ideal solution to me. Does anyone know anything about getting this work with a service?
Thank you!
Related
we are creating a vmware carvel package, I need do a sanity check for zookeeper, how can i check the output in gke?
Zookeeper output
Curl and localhost is failing to connect.
The problem is that your service type is ClusterIP which is not available from outside the Kubernetes cluster. There are two ways you could do this
If you need to really access this regularly from outside the cluster, you should deploy a service of type NodePort or LoadBalancer or an ingress. These would be reachable from outside
If you only want to check something quickly, you can temporarily make zookeeper visible with
kubectl port-forward zookeeper-0 2181
I know minikube should be used for local only, but i'd like to create a test environment for my applications.
In order to do that, I wish to expose my applications running inside the minikube cluster to external access (from any device on public internet - like a 4G smartphone).
note : I run minikube with --driver=docker
kubectl get services
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
web8080 NodePort 10.99.39.162 <none> 8080:31613/TCP 3d1h
minikube ip
192.168.49.2
One way to do it is as follows :
firewall-cmd --add-port=8081/tcp
kubectl port-forward --address 0.0.0.0 services/web8080 8081:8080
then I can access it using :
curl localhost:8081 (directly from the machine running the cluster inside a VM)
curl 192.168.x.xx:8081 (from my Mac in same network - this is the private ip of the machine running the cluster inside a VM)
curl 84.xxx.xxx.xxx:8081 (from a phone connected in 4G - this is the public ip exposed by my router)
I don't want to use this solution because kubectl port-forward is weak and need to be run every time the port-forwarding is no longer active.
How can I achieve this ?
(EDITED) - USING LOADBALANCER
when using LoadBalancer type and minikube tunnel, I can expose the service only inside the machine running the cluster.
kubectl get services
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
my-service LoadBalancer 10.111.61.218 10.111.61.218 8080:31831/TCP 3d3h
curl 10.111.61.218:8080 (inside the machine running the cluster) is working
but curl 192.168.x.xx:8080 (from my Mac on same LAN) is not working
Thanks
Minikube as a development tool for a single node Kubernetes cluster provides inherent isolation layer between Kubernetes and the external devices (being specific the inbound traffic to your cluster from LAN/WAN).
Different --drivers are allowing for flexibility when it comes to the place where your Kubernetes cluster will be spawned and how it will behave network wise.
A side note (workaround)!
As your minikube already resides in a VM and uses --driver=docker you could try to use --driver=none (you will be able to curl VM_IP:NodePort from the LAN). It will spawn your Kubernetes cluster directly on the VM.
Consider checking it's documentation as there are some certain limitations/disadvantages:
Minikube.sigs.k8s.io: Docs: Drivers: None
As this setup is already basing on the VM (with unknown hypervisor) and the cluster is intended to be exposed outside of your LAN, I suggest you going with the production-ready setup. This will inherently eliminate the connectivity issues you are facing. Kubernetes cluster will be provisioned directly on a VM and not in the Docker container.
Explaining the --driver=docker used: It will spawn a container on a host system with Kubernetes inside of it. Inside of this container, Docker will be used once again to spawn the necessary Pods to run the Kubernetes cluster.
As for the tools to provision your Kubernetes cluster you will need to chose the option that suits your needs the most. Some of them are the following:
Kubeadm
Kubespray
MicroK8S
After you created your Kubernetes cluster on a VM you could forward the traffic from your router directly to your VM.
Additional resources that you might find useful:
Stackoverflow.com: Questions Expose Kubernetes cluster to the Internet (Virtualbox with minikube)
curl $(minikube ip):$NODE_PORT : Now we can test that the app is exposed outside of the cluster using curl, the IP of the Node and the externally exposed port.
For you : curl 192.168.49.2:31613
Use nginx reverse-proxy
https://www.zepworks.com/posts/access-minikube-remotely-kvm/
install nginx, then in nginx.conf add this
stream {
server {
listen 8081;
proxy_pass 192.168.49.2:8080;
}
}
restart nginx
One way that I use to get around the fact that the process of kubectl port-forward stops after a while is to create a detach session using tmux following this. With that, I haven't had any problems with the exact same Minikube cluster configuration that you have.
I know minikube should be used for local only, but i'd like to create a test environment for my applications.
In order to do that, I wish to expose my applications running inside the minikube cluster to external access (from any device on public internet - like a 4G smartphone).
note : I run minikube with --driver=docker
kubectl get services
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
web8080 NodePort 10.99.39.162 <none> 8080:31613/TCP 3d1h
minikube ip
192.168.49.2
One way to do it is as follows :
firewall-cmd --add-port=8081/tcp
kubectl port-forward --address 0.0.0.0 services/web8080 8081:8080
then I can access it using :
curl localhost:8081 (directly from the machine running the cluster inside a VM)
curl 192.168.x.xx:8081 (from my Mac in same network - this is the private ip of the machine running the cluster inside a VM)
curl 84.xxx.xxx.xxx:8081 (from a phone connected in 4G - this is the public ip exposed by my router)
I don't want to use this solution because kubectl port-forward is weak and need to be run every time the port-forwarding is no longer active.
How can I achieve this ?
(EDITED) - USING LOADBALANCER
when using LoadBalancer type and minikube tunnel, I can expose the service only inside the machine running the cluster.
kubectl get services
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
my-service LoadBalancer 10.111.61.218 10.111.61.218 8080:31831/TCP 3d3h
curl 10.111.61.218:8080 (inside the machine running the cluster) is working
but curl 192.168.x.xx:8080 (from my Mac on same LAN) is not working
Thanks
Minikube as a development tool for a single node Kubernetes cluster provides inherent isolation layer between Kubernetes and the external devices (being specific the inbound traffic to your cluster from LAN/WAN).
Different --drivers are allowing for flexibility when it comes to the place where your Kubernetes cluster will be spawned and how it will behave network wise.
A side note (workaround)!
As your minikube already resides in a VM and uses --driver=docker you could try to use --driver=none (you will be able to curl VM_IP:NodePort from the LAN). It will spawn your Kubernetes cluster directly on the VM.
Consider checking it's documentation as there are some certain limitations/disadvantages:
Minikube.sigs.k8s.io: Docs: Drivers: None
As this setup is already basing on the VM (with unknown hypervisor) and the cluster is intended to be exposed outside of your LAN, I suggest you going with the production-ready setup. This will inherently eliminate the connectivity issues you are facing. Kubernetes cluster will be provisioned directly on a VM and not in the Docker container.
Explaining the --driver=docker used: It will spawn a container on a host system with Kubernetes inside of it. Inside of this container, Docker will be used once again to spawn the necessary Pods to run the Kubernetes cluster.
As for the tools to provision your Kubernetes cluster you will need to chose the option that suits your needs the most. Some of them are the following:
Kubeadm
Kubespray
MicroK8S
After you created your Kubernetes cluster on a VM you could forward the traffic from your router directly to your VM.
Additional resources that you might find useful:
Stackoverflow.com: Questions Expose Kubernetes cluster to the Internet (Virtualbox with minikube)
curl $(minikube ip):$NODE_PORT : Now we can test that the app is exposed outside of the cluster using curl, the IP of the Node and the externally exposed port.
For you : curl 192.168.49.2:31613
Use nginx reverse-proxy
https://www.zepworks.com/posts/access-minikube-remotely-kvm/
install nginx, then in nginx.conf add this
stream {
server {
listen 8081;
proxy_pass 192.168.49.2:8080;
}
}
restart nginx
One way that I use to get around the fact that the process of kubectl port-forward stops after a while is to create a detach session using tmux following this. With that, I haven't had any problems with the exact same Minikube cluster configuration that you have.
I am new to K8s and I am trying to migrate my service (which currently utilizes docker-compose.yml) to k8s. My service
deploys zipkin and elasticsearch
and these can be accessed at 'localhost:9411' and 'localhost:9200' respectively.
The most commonly used solution I found online was 'kompose' and I tried to run,
kompose up
2.
kompose convert
kubectl apply -f *****-deployment.yaml, ****-service.yaml
Once I finish this, I run kubectl get pods and I can see my deployments, but elasticsearch and zipkin are no more responsive on their respective localhost ports.
Ouput of 'kubectl get pods'
Output of 'docker ps'
Output of curl http://localhost:9200
Can someone tell me why this is happening and how to debug?
It is solved now; all I had to do was port forwarding.
kubectl port-forward zipkin-774cc77659-g929n 9411:9411
Thanks,
By default you service is exposed as ClusterIP, in this case your service will be accessible from within your cluster.
You can use port forwarding "With this connection in place you can use your local workstation to debug your application that is running in the pod" as described in the answer above.
Another approach is to use other "service types" like NodePort.
You can find more information here Publishing services (ServiceTypes)
I am new to Kubernetes. I have searched that you can use a LoadBalancer or NodePort to expose services. I am trying to follow tutorials on https://kubernetes.io/docs/tutorials/.
I am able to expose service with LoadBalancer but not NodePort. Here is the command I used.
kubectl expose deployment/kubernetes-bootcamp --type="NodePort" --port 8080
I guess you're specifically following 'Use a Service to Access an Application in a Cluster'
Just to note in an answer what you came to already in the comments, you don't set a --port parameter on the expose in that example and just on the run. You set the port that the Pod responds to in the run command but not the port exposed externally in expose - k8s will choose that for you automatically from the NodePort range and then tell you which one it chose. Then you find a public IP for a node in order to access.