oh, a custom image could do the trick...

Hi All, Getting this error after creating cluster. Please help KUBECONFIG=~/.kube/config kubectl cluster-info E0524 205106.452182 29254 memcache.go:287] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0524 205106.455395 29254 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0524 205106.457168 29254 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0524 205106.458029 29254 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request Kubernetes control plane is running at https://127.0.0.1:6443 :~/charts/tinkerbell$ trusted_proxies=$(kubectl get nodes -o jsonpath='{.items[*].spec.podCIDR}' | tr ' ' ',') E0524 234102.295917 40788 memcache.go:287] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0524 234102.302173 40788 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0524 234102.303085 40788 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0524 234102.305462 40788 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request

Hi 👋 I’m creating the cluster with embedded etcd, I wonder what’s the best way to configure the default etcd storage? The etcd doc says it can be tuned via

--quota-backend-bytes

, is this exposed by k3d?

A separated question, let’s say I created the cluster via the following command, how many etcd instances do I have? In other words, how many storage do I get?

k3d cluster create multiserver --servers 3

I have tried following the FAQ for Authenticating containerd against DockerHub, but it does not profit as I expect. My pulls from deployments in k3d are still pulling from the unauthenticated account. I still have to create a

.dockerconfigjson

secret and use

imagePullSecrets

on each of the deployments. This is what FAQ says to do.

# saved as e.g. $HOME/registries.yaml
configs:
  "<http://docker.io|docker.io>":
    auth:
      username: "$USERNAME"
      password: "$PASSWORD"

I have a k3d config yaml

registries:
  create:
    name: registry.local
    hostPort: "9000"
  config: |
    mirrors:
      <http://docker.io|docker.io>:
        endpoint:
        - "<http://host.k3d.internal:6000>"
    configs:
      "<http://docker.io|docker.io>":
        auth:
          username: "myusername"
          password: "mypassword"

I have a registry mirror at

localhost:6000

in Docker Engine (Docker Desktop).

iptables issue? OS X docker 20.10.4

docker run --privileged --name=k3d --hostname=k3d -p 6443:6443  rancher/k3d 
...
subject=CN = docker:dind client
Getting CA Private Key
/certs/client/cert.pem: OK
ip: can't find device 'ip_tables'
modprobe: can't change directory to '/lib/modules': No such file or directory

...
level=warning msg="Running iptables --wait -t nat -L -n failed with message: `iptables v1.8.7 (legacy): can't initialize iptables table `nat': iptables who? (do you need to insmod?)\nPerhaps iptables or your kernel needs to be upgraded.`, error: exit status 3"

how easy is it to replicate k3d config to a self hosted production k3s cluster with 3 machines?

docker pull <http://ghcr.io/k3d-io/k3d-tools:sha-1afe360|ghcr.io/k3d-io/k3d-tools:sha-1afe360> 

sha-1afe360: Pulling from k3d-io/k3d-tools
Digest: sha256:8bead49614ce6313bfe8ed3c59cf6d662a60fbc02e7e083610a606c04907d1f5
Status: Downloaded newer image for <http://ghcr.io/k3d-io/k3d-tools:sha-1afe360|ghcr.io/k3d-io/k3d-tools:sha-1afe360>

if someone can let me know how to fix this i'll submit the PR ty

there are 8 gpus in my machine, i set the gpus parameter for k3d cluster, value of gpus is

--gpus='"device=1,2"'

, but, i entered the agent node container and got all gpus through running

nvidia-smi

. i inspected the agent node container, and found the DeviceIDs is 1,2. so, why i got 8 gpus in agent node by running command

nvidia-smi

?

Hello, I have a question about port mapping in k3d cluster

if I create a cluster with k3d cluster create --api-port 6550 -p “12345:80@loadbalancer” --agents 2

I can map the ingress port 80 to localhost:12345

but if I create a cluster with k3d cluster create --api-port 6550 -p “12345:7777@loadbalancer” --agents 2, the ingress port 7777 doesn’t map to localhost:12345

I don’t understand why ?

@worried-receptionist-67815 has left the channel

@worried-receptionist-67815 has left the channel

*

Hi, is there a way to edit the

env

part of a running K3D cluster? I created and populated a cluster (so I don't want to destroy it) but I need to edit env.

Hi, can you help me with https://github.com/k3d-io/k3d/issues/1308

I'm trying to wait for the metrics server to be ready using: ++ kubectl -n kube-system wait deployment.apps/metrics-server --for=condition=Ready But even after this I'm getting: E0712 075944.754990 387453 memcache.go:287] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0712 075944.761806 387453 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request E0712 075944.775681 387453 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request Is there a condition I can wait for so that I can be sure the metrics server is available before I continue?

I’m playing around with k3d on macOS running docker-for-mac. I have a deployment running - 1 pod, 3 containers. How come I don’t see those containers with

docker ps

? What magic is at play here? 🙃

How do I disable traefik in a k3d cluster config file? I tried: k3s-server-arg: - "--no-deploy" - "traefik" but when I run: k3d cluster create --config cluster-config.yaml --wait I get: FATA[0000] Schema Validation failed for config file cluster-config.yaml: - options: Additional property k3s-server-arg is not allowed

I'm seeing a problem a problem where delete namespace is hanging, anyone have any suggestions as to what the issue is or how to fix other than deleting the whole cluster and restarting kubectl delete namespace cattle-system namespace "cattle-system" deleted While it says the namespace has been deleted the command never completes and interrupting and rerunning the delete again the namespace still exists and the command hangs again

I have a yaml config file for k3d which has a volumes section that looks like:

volumes: # repeatable flags are represented as YAML lists
  - volume: ${PWD}/cluster_storage/:/var/lib/rancher/k3s/storage/ # Volume for database and other storage
    nodeFilters:
      - server:0
      - agent:*

I'm trying to get this to work properly for wsl2 and docker-for-win, but having some trouble. PVCs fail to start likely because the path is not mounting properly. Is there a guide or pattern somewhere that would help get k3d volume mounting working in wsl2 (ubuntu)?

how come k3d uses nginx as proxy into k3s? why not use traefic which k3s uses as default?

for local development, are you using k8s volumes for source code? is there a better solution?

If it's code that supports hot reloading then yes. There are also tools like

air

to make that work for compiled languages ( Go in this case). Otherwise there are e.g. tilt or skaffold for dev loops. In acorn (acorn.io) we have a built-in dev mode even.

I have restarted the host, so server and agents containers have been restarted, but, i got the error in agent containers:

Waiting to retrieve agent configuration; server is not ready: failed to retrieve configuration from server: <https://127.0.0.1:6444/v1-k3s/config>: 401 Unauthorized

, now, server is ready and agents is not ready in k3d cluster, so what should i do for resolving this problem?

When I run: kubectl get pods -n kube-system I don't see the api server, controller manager listed I do see coredns listed why is this?