Rancher Ingress not initializing properly

[christianhuening]

hint:
I think this relates to #7131 but with Kubernetes Environment instead of Cattle.

Rancher Version:
1.2.1
Docker Version:
1.12.3
OS and where are the hosts located? (cloud, bare metal, etc):
Ubuntu 16 LTS hosted on-prem, bare metal
Setup Details: (single node rancher vs. HA rancher, internal DB vs. external DB)
single node Rancher, internal DB
Environment Type: (Cattle/Kubernetes/Swarm/Mesos)
Kubernetes 1.5.1
Steps to Reproduce:
Deploy my Kubernetes App with a defined Ingress ressource.

Results:
Load Balancer gets started but stops with this message log:

21.12.2016 13:35:09time="2016-12-21T12:35:09Z" level=info msg="KUBERNETES_URL is not set, skipping init of kubernetes controller"
21.12.2016 13:35:09time="2016-12-21T12:35:09Z" level=info msg="Starting Rancher LB service"
21.12.2016 13:35:10time="2016-12-21T12:35:10Z" level=info msg="LB controller: rancher"
21.12.2016 13:35:10time="2016-12-21T12:35:10Z" level=info msg="LB provider: haproxy"
21.12.2016 13:35:10time="2016-12-21T12:35:10Z" level=info msg="starting rancher controller"
21.12.2016 13:35:10time="2016-12-21T12:35:10Z" level=info msg="Healthcheck handler is listening on :10241"
21.12.2016 13:35:11time="2016-12-21T12:35:11Z" level=info msg=" -- starting haproxy\n[ALERT] 355/123510 (37) : Starting frontend GLOBAL: cannot bind UNIX socket [/run/haproxy/admin.sock]\n"
21.12.2016 13:35:12time="2016-12-21T12:35:12Z" level=info msg=" -- reloading haproxy config with the new config changes\n[WARNING] 355/123512 (48) : config : 'option forwardfor' ignored for proxy 'default' as it requires HTTP mode.\n"

and as a result I can access my exposed web UI through the Ingress LB, but any further routing is not working.

Expected:
Ingress LB initializes properly and all ressources are accessable.

[pulberg]

I see the same error in an LB in a Cattle env -

12/21/2016 11:45:31 AMtime="2016-12-21T17:45:31Z" level=info msg="KUBERNETES_URL is not set, skipping init of kubernetes controller" 12/21/2016 11:45:32 AMtime="2016-12-21T17:45:32Z" level=info msg="Starting Rancher LB service" 12/21/2016 11:45:32 AMtime="2016-12-21T17:45:32Z" level=info msg="LB controller: rancher" 12/21/2016 11:45:32 AMtime="2016-12-21T17:45:32Z" level=info msg="LB provider: haproxy" 12/21/2016 11:45:32 AMtime="2016-12-21T17:45:32Z" level=info msg="starting rancher controller" 12/21/2016 11:45:32 AMtime="2016-12-21T17:45:32Z" level=info msg="Healthcheck handler is listening on :10241" 12/21/2016 11:45:33 AMtime="2016-12-21T17:45:33Z" level=info msg=" -- starting haproxy\n[ALERT] 355/174532 (27) : Starting frontend GLOBAL: cannot bind UNIX socket [/run/haproxy/admin.sock]\n"

[mheiniger]

Could be related to the startup scripts of HAproxy: rancher/lb-controller#56

[aemneina]

Closing due to inactivity. The ingress controller and lb has changed a lot. If you still can reproduce this behavior in a recent version please holler.

[Vespira]

Hi, I have typically the same issue with much newer version of Rancher and Kubernetes.

Rancher Version:
1.5.10
Docker Version:
1.12.6
OS and where are the hosts located? (cloud, bare metal, etc):
CoreOS 4.9.24 on private sub network
Setup Details: (single node rancher vs. HA rancher, internal DB vs. external DB)
HA Rancher, with external DB
Environment Type: (Cattle/Kubernetes/Swarm/Mesos)
Kubernetes 1.5.4
Steps to Reproduce:
Deploy my Kubernetes App with a defined Ingress ressource.

I have the same log as the initial post of this thread. I tried create the Ingress with this yaml file with:
>kubectl apply -f ingress.yml

file content:

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: integration-ingress
spec:
  rules:
  - host: vm-docker1-dev.mycompany.com
    http:
      paths:
      - path: /purchases
        backend:
          serviceName: purchases
          servicePort: 8085
      - path: /products
        backend:
          serviceName: products
          servicePort: 8086
  - host: vm-docker2-dev.mycompany.com
    http:
      paths:
      - path: /purchases
        backend:
          serviceName: purchases
          servicePort: 8085
      - path: /products
        backend:
          serviceName: products
          servicePort: 8086

Can someone explain me how can we set this URL ? shouln't it be already done ? Thanks a lot !

[nathan-vp]

Initially I thought we encountered the same issue, but @superseb discovered a small typo in our YAML file - thanks for that!

@Vespira We run a newer version of Rancher, but for us the following YAML produces a working ingress. Can you try this?

demo-ingress.yml

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: nginx
  labels:
    name: nginx
spec:
 replicas: 3
 template:
   metadata:
     labels:
       app: nginx
   spec:
     containers:
     - name: nginx
       image: nginx
       ports:
       - containerPort: 80
---
kind: Service
apiVersion: v1
metadata:
  name: nginx
spec:
  selector:
    app: nginx
  ports:
  - port: 80
    targetPort: 80
    protocol: TCP
    name: http
---
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: ingress
  annotations:
    io.rancher.scheduler.global: "true"
spec:
  rules:
  - host: foo.bar.com
    http:
      paths:
      - backend:
          serviceName: nginx
          servicePort: 80

Apply: kubectl apply -f demo-ingress.yml

Our environment:

Rancher Version:
1.6.2
Docker Version:
1.12.6
OS and where are the hosts located? (cloud, bare metal, etc):
Ubuntu 14.04.5 LTS
Setup Details: (single node rancher vs. HA rancher, internal DB vs. external DB)
HA Rancher, external DB
Environment Type: (Cattle/Kubernetes/Swarm/Mesos)
Kubernetes 1.5.4