google cloud shell support

[matti]

$ go get sigs.k8s.io/kind
...
$ kind create cluster
Creating cluster 'kind-1' ...
 ✓ Ensuring node image (kindest/node:v1.12.3) 🖼
 ✓ [kind-1-control-plane] Creating node container 📦
 ✓ [kind-1-control-plane] Fixing mounts 🗻
 ✓ [kind-1-control-plane] Starting systemd 🖥
 ✓ [kind-1-control-plane] Waiting for docker to be ready 🐋
 ✗ [kind-1-control-plane] Starting Kubernetes (this may take a minute) ☸
Error: failed to create cluster: failed to apply overlay network: exit status 1
Usage:
  kind create cluster [flags]
Flags:
      --config string   path to a kind config file
  -h, --help            help for cluster
      --image string    node docker image to use for booting the cluster
      --name string     cluster context name (default "1")
      --retain          retain nodes for debugging when cluster creation fails
      --wait duration   Wait for control plane node to be ready (default 0s)
Global Flags:
      --loglevel string   logrus log level [panic, fatal, error, warning, info, debug] (default "warning")
failed to create cluster: failed to apply overlay network: exit status 1

[BenTheElder]

Huh, that particular failure is a bit surprising, thanks for reporting it!

We can probably avoid it when we finish offline support by using fixed versions of the overlay network and pre-obtaining the overlay images.

Possibly related to #136

/assign
/kind bug
/priority important-soon

[BenTheElder]

#426 is one of the blockers, there also appears to be something strange with the networking.

This almost works, something is wrong with talking to the API server, will need to investigate further.

$ cat ./kind-config.yaml
kind: Cluster
apiVersion: kind.sigs.k8s.io/v1alpha3
networking:
  apiServerPort: 6443
nodes:
- role: control-plane
  extraMounts:
  - containerPath: /sys/class/dmi/id/product_name
    hostPath: /home/bentheelder/product_name

$  cat ./product_name
kind

$ kind create cluster --config=./kind-config.yaml --loglevel=debug

[BenTheElder]

I think the web preview feature interferes. https://cloud.google.com/shell/docs/features#web_preview

It should be possible to stick to lower ports for the API server etc. which might fix that, but this will require changes to kind.

[BenTheElder]

#426 was fixed, but there's still some funky networking blocking this.

[aojea]

@BenTheElder you can define the port range in the host, it's defined in the kernel in
/proc/sys/net/ipv4/ip_local_port_range

How is the setup? is kind running in a VM in GCE?

[BenTheElder]

In Google cloud shell? Pretty sure https://cloud.google.com/shell/docs/features#web_preview is the issue. *From:*Antonio Ojea <[email protected]> *Date:*Fri, May 3, 2019, 05:46 *To:*kubernetes-sigs/kind *Cc:*Benjamin Elder, Mention @BenTheElder <https://github.com/BenTheElder> you can define the port range

…

in the host, it's defined in the kernel in /proc/sys/net/ipv4/ip_local_port_range — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#182 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAHADKYKQ43S4WEWTGWGWNDPTQX2RANCNFSM4GKWBPHA> .

[aojea]

It's working for me running from https://console.cloud.google.com/cloudshell , but don't know if this is the same situation you are describing 😅

antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ go get sigs.k8s.io/kind
antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ kind create cluster
Creating cluster "kind" ...
 ✓ Ensuring node image (kindest/node:v1.14.1) 🖼
 ✓ Preparing nodes 📦
export KUBECONFIG="$(kind get kubeconfig-path --name="kind")"
kubectl cluster-info
antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ export KUBECONFIG="$(kind get kubeconfig-path --name="kind")"
antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ kubectl get nodes
NAME                 STATUS     ROLES    AGE   VERSION
kind-control-plane   NotReady   master   25s   v1.14.1
antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ kubectl get nodes -o wide
NAME                 STATUS     ROLES    AGE   VERSION   INTERNAL-IP   EXTERNAL-IP   OS-IMAGE                                  KERNEL-VERSION   CONTAINER-RUNTIME
kind-control-plane   NotReady   master   29s   v1.14.1   172.18.0.2    <none>        Ubuntu Disco Dingo (development branch)   4.14.104+        containerd://1.2.6-0ubuntu1

The port forwarded is within the allowed range (2000 to 65000)

antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ docker ps
CONTAINER ID        IMAGE                  COMMAND                  CREATED              STATUS              PORTS                                  NAMES
c6d2f3edda33        kindest/node:v1.14.1   "/usr/local/bin/entr…"   About a minute ago   Up About a minute   38797/tcp, 127.0.0.1:38797->6443/tcp   kind-control-plane

And the port range defined is within the webshell range (2000 to 65000) too

antonio_ojea_garcia@cloudshell:~ (agile-sprite-239512)$ cat /proc/sys/net/ipv4/ip_local_port_range
32768   60999

[BenTheElder]

Did your node ever go to ready though? Testing this...

…

On Fri, May 3, 2019 at 9:01 AM Antonio Ojea ***@***.***> wrote: It's working for me running from https://console.cloud.google.com/cloudshell , but don't know if this is the same situation you are describing 😅 ***@***.***:~ (agile-sprite-239512)$ go get sigs.k8s.io/kind ***@***.***:~ (agile-sprite-239512)$ kind create cluster Creating cluster "kind" ... ✓ Ensuring node image (kindest/node:v1.14.1) 🖼 ✓ Preparing nodes 📦 export KUBECONFIG="$(kind get kubeconfig-path --name="kind")" kubectl cluster-info ***@***.***:~ (agile-sprite-239512)$ export KUBECONFIG="$(kind get kubeconfig-path --name="kind")" ***@***.***:~ (agile-sprite-239512)$ kubectl get nodes NAME STATUS ROLES AGE VERSION kind-control-plane NotReady master 25s v1.14.1 ***@***.***:~ (agile-sprite-239512)$ kubectl get nodes -o wide NAME STATUS ROLES AGE VERSION INTERNAL-IP EXTERNAL-IP OS-IMAGE KERNEL-VERSION CONTAINER-RUNTIME kind-control-plane NotReady master 29s v1.14.1 172.18.0.2 <none> Ubuntu Disco Dingo (development branch) 4.14.104+ containerd://1.2.6-0ubuntu1 The port forwarded is within the allowed range (2000 to 65000) ***@***.***:~ (agile-sprite-239512)$ docker ps CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES c6d2f3edda33 kindest/node:v1.14.1 "/usr/local/bin/entr…" About a minute ago Up About a minute 38797/tcp, 127.0.0.1:38797->6443/tcp kind-control-plane And the port range defined is within the webshell range (2000 to 65000) too ***@***.***:~ (agile-sprite-239512)$ cat /proc/sys/net/ipv4/ip_local_port_range 32768 60999 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#182 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAHADK4CIXOMKBWBFKOI6QTPTROVFANCNFSM4GKWBPHA> .

[BenTheElder]

I observe an issue dialing the CRI socket, only some of the static pods come up. not sure why, but overall this not working quite right is not surprising. cat /proc/1/cgroup 😉

[aojea]

hehe, s/working/runs without errors/
definitively it needs more testing

[BenTheElder]

did a little more poking, something is killing containerd 🤔

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[BenTheElder]

it's not clear to me that this is clearly feasible or that any similar tool works there.

it would be neat, but I don't think we have the time.

if anyone is interested in the future feel free to try again and report back / file a new issue

[sirdarckcat]

fyi kubernetes/minikube#6404 got it working

[aojea]

I think that this may fix it #1995

[BenTheElder]

This is indeed fixed at HEAD, and will work in the next release (soon!), you can test it now by running the following from your cloudshell:

GO111MODULE=on go get sigs.k8s.io/kind@master
export PATH="${HOME}/gopath/bin:${PATH}"
kind create cluster

[BenTheElder]

thanks @aojea @sirdarckcat :-)

[BenTheElder]