C4W3:Ungraded Lab: Building ML Pipelines with Kubeflow

Course Q&A Machine Learning Engineering for Production(MLOps) Deploying Machine Learning Models in Production
1

Trying to deploy Kubeflow Pipelines, I have this error “The connection to the server localhost:8080 was refused”
can anyone help me to fix this pb?

2

Hi @Touhami_Wala ,

Check that the api server is actually running and hasn’t crashed by running this command:

docker ps | grep kube-apiserver

But the most likely problem is that you don’t have a .kube directory with the right config in it. So you can also try this:

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

Hope it helps.

Kader,
Mentor MELP

3

Have you created a local cluster using kind?

1 Like
4

Hello @saou_a and all, I need help installing/configuring kind.

After inputing this on my terminal and not getting any errors: curl -Lo ./kind https://kind.sigs.k8s.io/dl/v0.8.1/kind-linux-amd64 && chmod +x ./kind && mv ./kind /home/xxx/Downloads/kind

When I try to create a cluster using “kind create cluster” or “sudo kind create cluster” I keep getting an error of “kind: command not found”

Any help?