Running kompose under k3s failed

[moluzhui]

I installed K3s with the following command and succeeded

# export INSTALL_K3S_EXEC="--write-kubeconfig ~/.kube/config --write-kubeconfig-mode 666"
# curl -sfL https://docs.rancher.cn/k3s/k3s-install.sh | INSTALL_K3S_MIRROR=cn sh -

Then I tried starting docker-compos.yaml with the official example

version: "2"

services:
  redis-master:
    image: gcr.io/google_containers/redis:e2e
    ports:
      - "6379"
  redis-slave:
    image: gcr.io/google_samples/gb-redisslave:v1
    ports:
      - "6379"
    environment:
      - GET_HOSTS_FROM=dns
  frontend:
    image: gcr.io/google-samples/gb-frontend:v4
    ports:
      - "80:80"
    environment:
      - GET_HOSTS_FROM=dns

But when I started, I found it only launched the service did not start deployments

# kompose -f docker-compose.yaml up
INFO We are going to create Kubernetes Deployments, Services and PersistentVolumeClaims for your Dockerized application. If you need different kind of resources, use the 'kompose convert' and 'kubectl create -f' commands instead. 
 
INFO Deploying application in "default" namespace 
INFO Successfully created Service: frontend       
INFO Successfully created Service: redis-master   
INFO Successfully created Service: redis-slave    

Your application has been deployed to Kubernetes. You can run 'kubectl get deployment,svc,pods,pvc' for details.

# kubectl get deployment,svc,pods,pvc
NAME                   TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
service/kubernetes     ClusterIP   10.43.0.1       <none>        443/TCP    14h
service/frontend       ClusterIP   10.43.62.105    <none>        80/TCP     8s
service/redis-master   ClusterIP   10.43.11.115    <none>        6379/TCP   8s
service/redis-slave    ClusterIP   10.43.180.220   <none>        6379/TCP   8s

Whether I have no configuration somewhere or kompose itself is not available on k3s?

OS is Ubuntu 18.04.2 LTS
k3s version is v1.17.4+k3s1 (3eee8ac3)
kompose version is 1.21.0 (992df58d8)

[hangyan]

Can you run the kompose convert command to see if there are any deployment yaml generated?

[moluzhui]

@hangyan As shown below, when I run the kompose convert command, it generated the corresponding deployements yaml

# kompose -f docker-compose.yaml convert
INFO Kubernetes file "frontend-service.yaml" created 
INFO Kubernetes file "redis-master-service.yaml" created 
INFO Kubernetes file "redis-slave-service.yaml" created 
INFO Kubernetes file "frontend-deployment.yaml" created 
INFO Kubernetes file "redis-master-deployment.yaml" created 
INFO Kubernetes file "redis-slave-deployment.yaml" created

Is this a bug or my configuration problem?

[hangyan]

@moluzhui Can you use kubectl to apply this yaml file and see how it goes?

[moluzhui]

@hangyan kubectl apply result is as follows:

# kubectl apply -f frontend-service.yaml -f redis-master-deployment.yaml -f redis-slave-deployment.yaml -f frontend-deployment.yaml -f redis-master-service.yaml -f redis-slave-service.yaml
service/frontend created
deployment.apps/redis-master created
deployment.apps/redis-slave created
deployment.apps/frontend created
service/redis-master created
service/redis-slave created

# kubectl get pods,svc,deploymentsNAME                                READY   STATUS    RESTARTS   AGE
pod/redis-slave-84c955db9d-wkxcz    1/1     Running   0          4s
pod/redis-master-584d6cd459-s22xt   1/1     Running   0          4s
pod/frontend-56f98d5478-r54vl       1/1     Running   0          4s

NAME                   TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)    AGE
service/kubernetes     ClusterIP   10.43.0.1       <none>        443/TCP    42s
service/frontend       ClusterIP   10.43.45.183    <none>        80/TCP     4s
service/redis-master   ClusterIP   10.43.10.0      <none>        6379/TCP   4s
service/redis-slave    ClusterIP   10.43.112.129   <none>        6379/TCP   4s

NAME                           READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/redis-slave    1/1     1            1           4s
deployment.apps/redis-master   1/1     1            1           4s
deployment.apps/frontend       1/1     1            1           4s

Everything seems normal.

[hangyan]

That's weird. I will use your compose file to reproduce this later. But you can use this as a workaround for now.

[moluzhui]

@hangyan Even scarier, I found the same problem with kompose up on the k83:v1.16.0 created on minikube.

[moluzhui]

That's weird. I will use your compose file to reproduce this later. But you can use this as a workaround for now.

[moluzhui]

That's weird. I will use your compose file to reproduce this later. But you can use this as a workaround for now.

Is it my yaml file error or a bug in kompose

[edmundmiller]

I can reproduce the issue on k3s, kompose up fails to create any pods, but kompose convert followed by kubectl apply works.

[hangyan]

close by #1297