Build and release ARM images

[Nwhouse]

Hello.
I am deploying for the first time to a Kube Raspberry Pi Cluster and seeing the below

NAMESPACE              NAME                                         READY   STATUS              RESTARTS   AGE
kube-system            coredns-f9fd979d6-45phj                      1/1     Running             1          33h
kube-system            coredns-f9fd979d6-tw9xm                      1/1     Running             1          33h
kube-system            etcd-pimaster                                1/1     Running             1          33h
kube-system            kube-apiserver-pimaster                      1/1     Running             1          33h
kube-system            kube-controller-manager-pimaster             1/1     Running             1          33h
kube-system            kube-flannel-ds-arm64-7x4v5                  1/1     Running             1          33h
kube-system            kube-flannel-ds-arm64-csr4j                  1/1     Running             0          33h
kube-system            kube-flannel-ds-arm64-m4fjz                  1/1     Running             0          33h
kube-system            kube-flannel-ds-arm64-sq7cz                  1/1     Running             0          33h
kube-system            kube-proxy-496tp                             1/1     Running             1          33h
kube-system            kube-proxy-m5dgq                             1/1     Running             0          33h
kube-system            kube-proxy-slv9x                             1/1     Running             0          33h
kube-system            kube-proxy-wrsbg                             1/1     Running             0          33h
kube-system            kube-scheduler-pimaster                      1/1     Running             1          33h
kubernetes-dashboard   dashboard-metrics-scraper-7b59f7d4df-7j9qm   1/1     Running             0          5h2m
kubernetes-dashboard   kubernetes-dashboard-5dbf55bd9d-clqtz        1/1     Running             0          5h2m
projectcontour         contour-56f877f95f-cht2n                     0/1     ContainerCreating   0          4h32m
projectcontour         contour-56f877f95f-xhd6q                     0/1     ContainerCreating   0          4h32m
projectcontour         contour-certgen-v1.8.0-6jjcf                 0/1     Error               0          4h32m
projectcontour         contour-certgen-v1.8.0-6vtp6                 0/1     Error               0          4h32m
projectcontour         envoy-l5xdw                                  0/2     Init:0/1            0          4h32m
projectcontour         envoy-smhqt                                  0/2     Init:0/1            0          4h32m
projectcontour         envoy-tt77c                                  0/2     Init:0/1            0          4h32m

Any assistance would be great.

Thanks

[youngnick]

Hi @Nwhouse, thanks for this issue. I edited your comment to fix the formatting.

The contour-certgen job needs to successfully run before Contour will be able to start.
I would suggest checking out the logs for contour-certgen-v1.8.0-6jjcf or contour-certgen-v1.8.0-6vtp6.

[Nwhouse]

Hi @youngnick
So I am new to this and building my first Kube cluster. As I had been messing around with it quite a bit I decided to just scrap it and start over. Once I got my cluster up I deployed Contour. As such I am still seeing the same results.
I am also trying to figure out where to acquire the logs from. If you could assist me with finding the logs I will post them so hopefully we can figure out what the issue is.

Thanks!!!

[skriss]

The outputs of kubectl -n projectcontour describe pod contour-certgen-v1.8.0-6jjcf and kubectl -n projectcontour logs pod/contour-certgen-v1.8.0-6jjcf should be a good start (substitute the correct pod name since they've changed).

[Nwhouse]

Thanks @skriss

Name:         contour-certgen-v1.8.1-42jf2
Namespace:    projectcontour
Priority:     0
Node:         pinode3/10.0.0.43
Start Time:   Thu, 03 Sep 2020 20:42:07 +0000
Labels:       app=contour-certgen
              controller-uid=31571ab5-08d9-48f8-a4a2-751189943596
              job-name=contour-certgen-v1.8.1
Annotations:  <none>
Status:       Failed
IP:           10.244.3.3
IPs:
  IP:           10.244.3.3
Controlled By:  Job/contour-certgen-v1.8.1
Containers:
  contour:
    Container ID:  docker://3760e19ca57bbcf38cf83954c9228f384c1a6472d2bb41c7a170c16189b4263c
    Image:         docker.io/projectcontour/contour:latest
    Image ID:      docker-pullable://projectcontour/contour@sha256:6f3becaadaca4702c5f9304154aa45b7f76386e7be9b996606c571bc40ea099f
    Port:          <none>
    Host Port:     <none>
    Command:
      contour
      certgen
      --kube
      --incluster
      --overwrite
      --secrets-format=compact
      --namespace=$(CONTOUR_NAMESPACE)
    State:          Terminated
      Reason:       Error
      Exit Code:    1
      Started:      Thu, 03 Sep 2020 20:42:11 +0000
      Finished:     Thu, 03 Sep 2020 20:42:11 +0000
    Ready:          False
    Restart Count:  0
    Environment:
      CONTOUR_NAMESPACE:  projectcontour (v1:metadata.namespace)
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from contour-certgen-token-vhd49 (ro)
Conditions:
  Type              Status
  Initialized       True 
  Ready             False 
  ContainersReady   False 
  PodScheduled      True 
Volumes:
  contour-certgen-token-vhd49:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  contour-certgen-token-vhd49
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
                 node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Events:          <none>

xxxxxxx@PiMaster:~$ kubectl -n projectcontour logs pod/contour-certgen-v1.8.1-42jf2
standard_init_linux.go:211: exec user process caused "exec format error"

[jpeach]

Raspberry Pi Cluster - we don't build or ship ARM images for Contour.

[Nwhouse]

Ahhh. well that answers the question. Thanks!

…

On Thu, Sep 3, 2020 at 10:53 PM James Peach ***@***.***> wrote: Raspberry Pi Cluster - we don't build or ship ARM images for Contour. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#2868 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKSC6DHRKVYKZUR5BXDVGPLSEBJCXANCNFSM4QSNNNVQ> .

-- Nathan Whitehouse nathan@whitehouse-family.net P. 770.831.7460 M. 770.378.9067

[jpeach]

Updated title. ARM is totally reasonable for homelab usage and becoming more popular with cloud providers too.

[mattmoor]

Shameless plug for the new multi-arch support in github.com/google/ko 😉 😇

[skriss]

projectcontour/contour:main now has multi-arch images, including arm64: https://hub.docker.com/layers/projectcontour/contour/main/images/sha256-bf480a3603d37de2246fcebce45dca2e467f966fae7470317b1f5da03a7caf76?context=explore

It'd be great if folks interested in arm could test these out! Assuming all's well, we will ship a multi-arch image for the next Contour release.

[stevesloka]

I just revalidated that amd64 still worked on a local kind cluster (just to be sure the registry stuff is working).

[advancedwebdeveloper]

@stevesloka , got some issues with #3031 (comment)

[sunjayBhatia]

@advancedwebdeveloper can you open a new issue to describe the issues you are having?