使用Let's Encrypt和Ingress设置kubernetes-Gnutls_handshake()失败:接收到意外的TLS数据包

时间:2019-03-21 21:48:46

标签: kubernetes nginx-ingress cert-manager

我正在使用标准过程为在Kubernetes上运行的应用程序启用HTTPS终止,使用: -入口nginx -AWS ELB经典 -用于加密的证书管理器

我已使用此处描述的过程:https://www.digitalocean.com/community/tutorials/how-to-set-up-an-nginx-ingress-with-cert-manager-on-digitalocean-kubernetes

以前我一直能够使Ingress与HTTP一起使用,但是我在HTTPS上遇到问题,当我尝试对URL进行cURL时出现以下错误:

$ curl -iv https://<SERVER>
* Rebuilt URL to: <SERVER>
*   Trying <IP_ADDR>...
* Connected to <SERVER> (<IP_ADDR>) port 443 (#0)
* found 148 certificates in /etc/ssl/certs/ca-certificates.crt
* found 592 certificates in /etc/ssl/certs
* ALPN, offering http/1.1
* gnutls_handshake() failed: An unexpected TLS packet was received.
* Closing connection 0
curl: (35) gnutls_handshake() failed: An unexpected TLS packet was received.

这是我目前拥有的:

在kube系统名称空间中运行的证书管理器:

$ kubectl get pods -n kube-system
    NAME                                     READY   STATUS      RESTARTS   AGE
    cert-manager-5f8db6f6c4-c4t4k            1/1     Running     0          2d1h
    cert-manager-webhook-85dd96d87-rxc7p     1/1     Running     0          2d1h
    cert-manager-webhook-ca-sync-pgq6b       0/1     Completed   2          2d1h

在ingress-nginx名称空间中的入口设置:

$ kubectl get all -n ingress-nginx
NAME                                            READY   STATUS    RESTARTS   AGE
pod/default-http-backend-587b7d64b5-ftws2       1/1     Running   0          2d1h
pod/nginx-ingress-controller-68bb4bfd98-zsz8d   1/1     Running   0          12h

NAME                           TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)                      AGE
service/default-http-backend   ClusterIP   <IP_ADDR_1>   <none>        80/TCP                       2d1h
service/ingress-nginx          NodePort    <IP_ADDR_2>   <none>        80:32327/TCP,443:30313/TCP   2d1h

NAME                                       DESIRED   CURRENT   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/default-http-backend       1         1         1            1           2d1h
deployment.apps/nginx-ingress-controller   1         1         1            1           12h

应用程序名称空间中的应用程序和入口:

$ kubectl get all -n app
NAME              READY   STATUS    RESTARTS   AGE
pod/appserver-0   1/1     Running   0          2d1h

NAME                TYPE        CLUSTER-IP    EXTERNAL-IP   PORT(S)         AGE
service/appserver   ClusterIP   <IP_ADDR>   <none>        22/TCP,80/TCP   2d1h

NAME                         DESIRED   CURRENT   AGE
statefulset.apps/appserver   1         1         2d1h

$ kubectl describe ingress -n app
Name:             appserver
Namespace:        app
Address:
Default backend:  default-http-backend:80 (<none>)
TLS:
  letsencrypt-prod terminates <SERVER>
Rules:
  Host                      Path  Backends
  ----                      ----  --------
  <SERVER>
                            /   appserver:80 (<none>)
Annotations:
  certmanager.k8s.io/cluster-issuer:  letsencrypt-prod
  kubernetes.io/ingress.class:        nginx
Events:                               <none>

这是入口资源的样子:

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  annotations:
    kubernetes.io/ingress.class: nginx
    certmanager.k8s.io/cluster-issuer: letsencrypt-prod
  name: appserver
  namespace: app
spec:
  tls:
  - hosts:
    - <SERVER>
    secretName: letsencrypt-prod
  rules:
  - host: <SERVER>
    http:
      paths:
      - backend:
          serviceName: appserver
          servicePort: 80
        path: /

我已完成的其他检查:

检查证书是否正确生成:

$ kubectl describe cert -n app
Name:         letsencrypt-prod
...
Status:
  Conditions:
    Last Transition Time:  2019-03-20T14:23:07Z
    Message:               Certificate is up to date and has not expired
    Reason:                Ready
    Status:                True
    Type:                  Ready

从cert-manager检查的日志:

$ kubectl logs -f cert-manager-5f8db6f6c4-c4t4k -n kube-system
...
I0320 14:23:08.368872       1 sync.go:177] Certificate "letsencrypt-prod" for ingress "appserver" already exists
I0320 14:23:08.368889       1 sync.go:180] Certificate "letsencrypt-prod" for ingress "appserver" is up to date
I0320 14:23:08.368894       1 controller.go:179] ingress-shim controller: Finished processing work item "app/appserver"
I0320 14:23:12.548963       1 controller.go:183] orders controller: syncing item 'app/letsencrypt-prod-1237734172'
I0320 14:23:12.549608       1 controller.go:189] orders controller: Finished processing work item "app/letsencrypt-prod-1237734172"

目前还不确定,还有什么值得检查的?

2 个答案:

答案 0 :(得分:1)

您似乎正在使用带有cert-manager的旧版nginx-ingress,这可能是导致错误的罪魁祸首。

我建议您尝试使用最新版本的Helm部署nginx-ingress(确保nginx-ingress-controller的图像标签为v0.23.0)。另外,请确保在头盔部署期间以最新版本(--version v0.7.0)部署证书管理器。默认情况下,您应该看到三个Pod:cert-manager,cert-manager-cainjector和cert-manager-webhook。

我必须禁用Webhook,因为尽管提供了正确的参数,但它仍阻止了证书的颁发,因此您可能必须这样做。

希望这会有所帮助!

答案 1 :(得分:1)

毕竟,问题似乎与我如何在AWS上的ELB classic上完成侦听器的设置有关。

我已经完成以下操作:

HTTP 80 -> HTTP <INGRESS_SVC_NODE_PORT_1>
HTTP 443 -> HTTP <INGRESS_SVC_NODE_PORT_2>

第一个错误是我使用HTTP而不是HTTPS作为443端口。当我尝试使用HTTPS时,我不得不输入对我来说没有意义的SSL证书,因为我正在使用“让我们加密”在Ingress级别进行SSL终止。

因此,侦听器的以下配置有效:

TCP 80 -> TCP <INGRESS_SVC_NODE_PORT_1>
TCP 443 -> TCP <INGRESS_SVC_NODE_PORT_2>