尝试在Raspberry Pi 3B+上运行OpenFaas函数时获取OpenFaas

Question

我正在尝试使用OpenFaas项目和运行在2个Raspberry 3B+上的kubernetes集群来部署一个函数。不幸的是，应该处理函数的吊舱将进入ImageInspectError状态.我试着用Docker直接运行这个函数，它包含在Docker映像中，一切都很好。

我在OpenFaas github上打开了一个OpenFaas，维护人员告诉我直接向Kubernetes社区询问一些线索。

我的第一个问题是: ImageInspectError是什么意思?它从何而来？

这是我掌握的所有信息：

预期行为

应该跑了。

当前行为

NAMESPACE     NAME                                 READY     STATUS              RESTARTS   AGE
kube-system   etcd-masternode                      1/1       Running             1          1d
kube-system   kube-apiserver-masternode            1/1       Running             1          1d
kube-system   kube-controller-manager-masternode   1/1       Running             1          1d
kube-system   kube-dns-7f9b64f644-x42sr            3/3       Running             3          1d
kube-system   kube-proxy-wrp6f                     1/1       Running             1          1d
kube-system   kube-proxy-x6pvq                     1/1       Running             1          1d
kube-system   kube-scheduler-masternode            1/1       Running             1          1d
kube-system   weave-net-4995q                      2/2       Running             3          1d
kube-system   weave-net-5g7pd                      2/2       Running             3          1d
openfaas-fn   figlet-7f556fcd87-wrtf4              1/1       Running             0          4h
openfaas-fn   testfaceraspi-7f6fcb5897-rs4cq       0/1       ImageInspectError   0          2h
openfaas      alertmanager-66b98dd4d4-kcsq4        1/1       Running             1          1d
openfaas      faas-netesd-5b5d6d5648-mqftl         1/1       Running             1          1d
openfaas      gateway-846f8b5686-724q8             1/1       Running             2          1d
openfaas      nats-86955fb749-7vsbm                1/1       Running             1          1d
openfaas      prometheus-6ffc57bb8f-fpk6r          1/1       Running             1          1d
openfaas      queue-worker-567bcf4d47-ngsgv        1/1       Running             2          1d

testfaceraspi不运行。

舱内的原木：

$ kubectl logs testfaceraspi-7f6fcb5897-rs4cq -n openfaas-fn
Error from server (BadRequest): container "testfaceraspi" in pod "testfaceraspi-7f6fcb5897-rs4cq" is waiting to start: ImageInspectError

豆荚描述：

$ kubectl describe pod -n openfaas-fn  testfaceraspi-7f6fcb5897-rs4cq 
Name:           testfaceraspi-7f6fcb5897-rs4cq
Namespace:      openfaas-fn
Node:           workernode/10.192.79.198
Start Time:     Thu, 12 Jul 2018 11:39:05 +0200
Labels:         faas_function=testfaceraspi
                pod-template-hash=3929761453
Annotations:    prometheus.io.scrape=false
Status:         Pending
IP:             10.40.0.16
Controlled By:  ReplicaSet/testfaceraspi-7f6fcb5897
Containers:
  testfaceraspi:
    Container ID:   
    Image:          gallouche/testfaceraspi
    Image ID:       
    Port:           8080/TCP
    Host Port:      0/TCP
    State:          Waiting
      Reason:       ImageInspectError
    Ready:          False
    Restart Count:  0
    Liveness:       exec [cat /tmp/.lock] delay=3s timeout=1s period=10s #success=1 #failure=3
    Readiness:      exec [cat /tmp/.lock] delay=3s timeout=1s period=10s #success=1 #failure=3
    Environment:
      fprocess:  python3 index.py
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-5qhnn (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          False 
  PodScheduled   True 
Volumes:
  default-token-5qhnn:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-5qhnn
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason            Age                 From                 Message
  ----     ------            ----                ----                 -------
  Warning  DNSConfigForming  2m (x1019 over 3h)  kubelet, workernode  Search Line limits were exceeded, some search paths have been omitted, the applied search line is: openfaas-fn.svc.cluster.local svc.cluster.local cluster.local heig-vd.ch einet.ad.eivd.ch web.ad.eivd.ch

事件日志：

$ kubectl get events --sort-by=.metadata.creationTimestamp -n openfaas-fn
LAST SEEN   FIRST SEEN   COUNT     NAME                                              KIND         SUBOBJECT                        TYPE      REASON                  SOURCE                  MESSAGE
14m         1h           347       testfaceraspi-7f6fcb5897-rs4cq.1540db41e89d4c52   Pod                                           Warning   DNSConfigForming        kubelet, workernode     Search Line limits were exceeded, some search paths have been omitted, the applied search line is: openfaas-fn.svc.cluster.local svc.cluster.local cluster.local heig-vd.ch einet.ad.eivd.ch web.ad.eivd.ch
4m          1h           75        figlet-7f556fcd87-wrtf4.1540db421002b49e          Pod                                           Warning   DNSConfigForming        kubelet, workernode     Search Line limits were exceeded, some search paths have been omitted, the applied search line is: openfaas-fn.svc.cluster.local svc.cluster.local cluster.local heig-vd.ch einet.ad.eivd.ch web.ad.eivd.ch
10m         10m          1         testfaceraspi-7f6fcb5897-d6z78.1540df9ed8b91865   Pod                                           Normal    Scheduled               default-scheduler       Successfully assigned testfaceraspi-7f6fcb5897-d6z78 to workernode
10m         10m          1         testfaceraspi-7f6fcb5897.1540df9ed6eee11f         ReplicaSet                                    Normal    SuccessfulCreate        replicaset-controller   Created pod: testfaceraspi-7f6fcb5897-d6z78
10m         10m          1         testfaceraspi-7f6fcb5897-d6z78.1540df9eef3ef504   Pod                                           Normal    SuccessfulMountVolume   kubelet, workernode     MountVolume.SetUp succeeded for volume "default-token-5qhnn" 
4m          10m          27        testfaceraspi-7f6fcb5897-d6z78.1540df9eef5445c0   Pod                                           Warning   DNSConfigForming        kubelet, workernode     Search Line limits were exceeded, some search paths have been omitted, the applied search line is: openfaas-fn.svc.cluster.local svc.cluster.local cluster.local heig-vd.ch einet.ad.eivd.ch web.ad.eivd.ch
8m          9m           8         testfaceraspi-7f6fcb5897-d6z78.1540df9f670d0dad   Pod          spec.containers{testfaceraspi}   Warning   InspectFailed           kubelet, workernode     Failed to inspect image "gallouche/testfaceraspi": rpc error: code = Unknown desc = Error response from daemon: readlink /var/lib/docker/overlay2/l: invalid argument
9m          9m           7         testfaceraspi-7f6fcb5897-d6z78.1540df9f670fcf3e   Pod          spec.containers{testfaceraspi}   Warning   Failed                  kubelet, workernode     Error: ImageInspectError

复制步骤(对于bug)

1. 在一个2节点的OpenFaas集群上部署k8s
2. 用faas new testfaceraspi --lang python3-armhf创建函数
3. 在handler.py中添加以下代码： 导入json句柄(Req)：jsonl = json.loads(req)返回("Found“+ str(jsonl"nbFaces") +”OpenFaas函数中的faces on raspi !")
4. 更改.yml提供程序中的网关和图像:名称: faas：http://127.0.0.1:31112 功能: testfaceraspi: lang: python3-armhf处理程序：./testfaceraspi映像: gallouche/testfaceraspi
5. 运行faas build -f testfacepi.yml
6. 在DockerHub中使用docker login登录
7. 运行faas push -f testfacepi.yml
8. 运行faas deploy -f testfacepi.yml

你的环境

• FaaS版本(来自：faas-cli version的完整输出)： 提交: 3995a8197f1df1ecdf524844477cffa04e4690ea版本: 0.6.11
• Docker版本(来自：docker version的完整输出)： Client: Version: 18.04.0-ce API版本: 1.37 Git版本: go1.9.4 Git提交: 3d479c0构建: Tue Apr 10 18:25:24 2018年OS/Arch: linux/arm实验: false Orchestrator:群服务器: version: 18.04.0-ce API版本: 1.37 (最低版本1.12) Go版本: go1.9.4 Git提交: 3d479c0构建：2018年4月10日18:21:25 OS/Arch: linux/arm实验: false
• 操作系统和版本(例如Linux、Windows、MacOS)： 分发服务器ID: Raspbian描述: Raspbian /Linux9.4(弹出式)发行版: 9.4代码

提前谢谢，如果你还需要更多的信息，请告诉我。

盖洛什

Answer 1

我看到了这个错误，因为Kubernetes不支持docker版本。在Kubernetes版本1.11中，支持的版本为1.11.2至1.13.1和17.03.x。

我无法用OpenFaaS测试解决方案。