Ingress¶

What is an Ingress Controller?¶

In Kubernetes, Ingress allows external users and client applications access to HTTP services. Ingress consists of two components.

The Ingress Resource is a collection of rules for the inbound traffic to reach Services. These are Layer 7 (L7) rules that allow hostnames and optionally paths, to be directed to specific Services in Kubernetes.

An Ingress Controller which acts upon the rules set by the Ingress Resource, typically via an HTTP or L7 load balancer. It is vital that both pieces are properly configured to route traffic from an outside client to a Kubernetes Service.

Prerequisites¶

As mentioned above, simply defining an ingress resource is not sufficient to define Ingress, it also requires an ingress controller to implement the rule set provided by the resource.

Catalyst Cloud provides the native Octavia ingress controller and it also supports other common ingress controllers such as the Nginx ingress controller.

While all ingress controllers should fit the reference specification they do in fact operate slightly differently and the relevant vendor documentation should be consulted for details on the correct setup.

Using the Octavia ingress controller¶

This guide explains how to deploy an Octavia ingress controller in to a Kubernetes cluster running on an Openstack cloud.

Some working examples¶

The code examples below will walk you through all of the steps required to be able to setup and Octavia ingress on your cluster.

The following scenarios will be covered.

simple HTTP ingress

Simple ingress with Octavia¶

The test application¶

For the test application we will use the Google Cloud echoserver

# deployment-echoserver.yml

---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: echoserver-deployment
  labels:
    app: echoserver
spec:
  replicas: 3
  selector:
    matchLabels:
      app: echoserver
  template:
    metadata:
      labels:
        app: echoserver
    spec:
      containers:
      - name: echoserver
        image: gcr.io/google-containers/echoserver:1.10
        ports:
        - containerPort: 8080

Deploy this to the cluster with the following command.

$ kubectl apply -f deployment-echoserver.yml

The create a service to expose the pods on port 80.

$ kubectl expose deployment echoserver --type=NodePort --target-port=8080 --port 80 --name=echoserver-svc

Deploying the ingress¶

In order to setup up the ingress controller we need to create a service account for this purpose and associate it with an appropriate cluster role that has the necessary cluster rights.

# octavia-ing-rbac.yml

---
kind: ServiceAccount
apiVersion: v1
metadata:
  name: octavia-ingress-controller
  namespace: kube-system

---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: octavia-ingress-controller
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:
  - kind: ServiceAccount
    name: octavia-ingress-controller
    namespace: kube-system

$ kubectl apply -f octavia-ing-rbac.yml
serviceaccount/octavia-ingress-controller created
clusterrolebinding.rbac.authorization.k8s.io/octavia-ingress-controller created

The file octavia-ing-configmap.yml below, defines the configuration for the ingress controller. It requires several user, cluster and project specific details to be entered. The following commands can assist with acquiring this information.

The password entry and cluster name will need to be supplied by the user.

# to get the subnet-id for the cluster1
$ clustername='<YOUR_CLUSTER_NAME>'
$ openstack subnet list | grep $clustername | awk -F'\| ' '{ print $2 }'

# to get the floating-network-id
$ openstack network list --external -c ID -f value

# to get the project-id auth-url and region
$ openstack configuration show -c auth.project_id -f value -c auth_url -f value -c region_name -f value -f yaml

# to get the user-id
$ openstack token issue -c user_id -f value -f yaml

# octavia-ing-configmap.yml

---
kind: ConfigMap
apiVersion: v1
metadata:
  name: octavia-ingress-controller-config
  namespace: kube-system
data:
  config: |
    cluster-name: fluentd-test
    openstack:
      auth-url: <cloud_regions_authentication_url>
      user-id: <cloud_user_id>
      password: <cloud_user_password>
      project-id: <cloud_project_id>
      region: nz-hlz-1
    octavia:
      subnet-id: <cluster_subnet_id>
      floating-network-id: <cloud_regions_public_net_id>

Once the file has been updated with the correct information create the configmap.

$ kubectl apply -f octavia-ing-configmap.yml
configmap/octavia-ingress-controller-config created

Next we define the ingress itself. This provides the rules and conditions for the ingress controller to route traffic to the applications. We have defined the following:

a host name for host-based routing
a default path for the entry URL
link to the backend service echoserver-svc on port 80

# octavia-ing-ingress.yml

---
apiVersion: networking.k8s.io/v1beta1
kind: Ingress
metadata:
  name: test-octavia-ingress
  annotations:
    kubernetes.io/ingress.class: "openstack"
    octavia.ingress.kubernetes.io/internal: "false"
spec:
  rules:
  - host: api.sample.com
    http:
      paths:
      - path: /
        backend:
          serviceName: echoserver-svc
          servicePort: 80

$ kubectl apply -f octavia-ing-ingress.yml
ingress.networking.k8s.io/test-octavia-ingress created

Finally we deploy the ingress controller as a statefulset. This deploys a single replica running the octavia-ing-configmap.yml container. It mounts the configuration as a volume in the pod.

# octavia-ing-statefulset.yml

---
kind: StatefulSet
apiVersion: apps/v1
metadata:
  name: octavia-ingress-controller
  namespace: kube-system
  labels:
    k8s-app: octavia-ingress-controller
spec:
  replicas: 1
  selector:
    matchLabels:
      k8s-app: octavia-ingress-controller
  serviceName: octavia-ingress-controller
  template:
    metadata:
      labels:
        k8s-app: octavia-ingress-controller
    spec:
      serviceAccountName: octavia-ingress-controller
      tolerations:
        - effect: NoSchedule # Make sure the pod can be scheduled on master kubelet.
          operator: Exists
        - key: CriticalAddonsOnly # Mark the pod as a critical add-on for rescheduling.
          operator: Exists
        - effect: NoExecute
          operator: Exists
      containers:
        - name: octavia-ingress-controller
          image: docker.io/catalystcloud/octavia-ingress-controller:v1.18.0-catalyst
          imagePullPolicy: IfNotPresent
          args:
            - /bin/octavia-ingress-controller
            - --config=/etc/config/octavia-ingress-controller-config.yaml
          volumeMounts:
            - mountPath: /etc/kubernetes
              name: kubernetes-config
              readOnly: true
            - name: ingress-config
              mountPath: /etc/config
      hostNetwork: true
      volumes:
        - name: kubernetes-config
          hostPath:
            path: /etc/kubernetes
            type: Directory
        - name: ingress-config
          configMap:
            name: octavia-ingress-controller-config
            items:
              - key: config
                path: octavia-ingress-controller-config.yaml

$ kubectl apply -f octavia-ing-statefulset.yml
statefulset.apps/octavia-ingress-controller unchanged

Once the pod is deployed we need to wait until the ingress is assigned a floating IP address as shown below.

$ kubectl get ing
NAME                   HOSTS              ADDRESS          PORTS   AGE
test-octavia-ingress   api.sample.com     103.197.61.251   80      3m43s

Now we can test our connectivity with the following.

$ ip=103.197.61.251
$ curl -H "Host:api.sample.com" http://$ip/ping/


Hostname: echoserver-66ff84f846-ljc9h

Pod Information:
  -no pod information available-

Server values:
  server_version=nginx: 1.13.3 - lua: 10008

Request Information:
  client_address=10.0.0.12
  method=GET
  real path=/ping/
  query=
  request_version=1.1
  request_scheme=http
  request_uri=http://api.sample.com:8080/ping/

Request Headers:
  accept=*/*
  host=api.sample.com
  user-agent=curl/7.64.1

Request Body:
  -no body in request-

Cleanup¶

To remove this setup from your cluster run the following from the directory that holds the files you created during the deployment.

$ for i in octavia-ing-*; do kubectl delete -f $i ; done
configmap "octavia-ingress-controller-config" deleted
ingress.networking.k8s.io "test-octavia-ingress" deleted
serviceaccount "octavia-ingress-controller" deleted
clusterrolebinding.rbac.authorization.k8s.io "octavia-ingress-controller" deleted
statefulset.apps "octavia-ingress-controller" deleted

Using the Nginx ingress controller¶

This guide explains how to deploy an Nginx ingress controller in to a Kubernetes cluster running on an Openstack cloud.

This guide makes the following assumptions.

You have installed the OpenStack command line tools and sourced an OpenStack RC file, as explained at Command line interface (CLI)
You have installed kubectl and and familiar with its use.

Some working examples¶

In order to use the Nginx ingress controller we first need to install it into our cluster. While this can be done by hand creating all of the required deployments, services and roles it is far simpler to use the Helm nginx-ingress chart to do this.

The following 3 scenarios are all implemented using the same Helm chart with with differing configuration in order to achieve the necessary Nginx ingress behaviour. They also use the same container as the backend service.

simple HTTP ingress
Ingress with TLS termination
Ingress with PROXY protocol support

Setting up Helm¶

If you do not already have helm installed you can see installing helm for more details.

Once Helm is installed we need to ensure it will work correctly. As the Catalyst Cloud does make use of RBAC access controls within it’s clusters we need to also ensure that there is a correctly configured service account in our cluster for tiller. The following YAML will create this account and ensure that it has the correct RBAC roles to perform the necessary actions on behalf of Helm.

cat <<EOF | kubectl apply -f -apiVersion: v1 kind: ServiceAccount
metadata:
  name: tiller
  namespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: tiller
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:
  - kind: ServiceAccount
    name: tiller
    namespace: kube-system
EOF

Now that we have the service account in place we can initialise Helm.

$ helm init --service-account tiller --history-max 200

$ kubectl -n kube-system get pods

Simple ingress with Nginx¶

For the first example we will create a straight forward HTTP ingress controller that will direct traffic to a backend pod that will simply echo back details of the pod, the request and the associated headers it received.

$ kubectl run echoserver --image=gcr.io/google-containers/echoserver:1.10 --port=8080 --expose
kubectl run --generator=deployment/apps.v1 is DEPRECATED and will be removed in a future version. Use kubectl run --generator=run-pod/v1 or kubectl create instead.
service/echoserver created
deployment.apps/echoserver created

$ kubectl get pod,service
kubectl get pod,service
NAME                              READY   STATUS    RESTARTS   AGE
pod/echoserver-7cc8b87c6f-h8ls5   1/1     Running   0          34m

NAME                 TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)    AGE
service/echoserver   ClusterIP   10.254.58.23   <none>        8080/TCP   34m
service/kubernetes   ClusterIP   10.254.0.1     <none>        443/TCP    41d

Now we need to define the basic configuration for the ingress controller.

cat <<EOF > nginx-ingress-controller-helm-values.yaml
controller:
    publishService:
        enabled: true
EOF

Now create the nginx ingress controller using the helm chart.

$ helm install stable/nginx-ingress --name nginx-ingress -f nginx-ingress-controller-helm-values.yaml

NAME:   nginx-ingress
LAST DEPLOYED: Wed Aug  7 13:55:09 2019
NAMESPACE: default
STATUS: DEPLOYED

RESOURCES:
==> v1/Pod(related)
NAME                                            READY  STATUS             RESTARTS  AGE
nginx-ingress-controller-9d9ccb6f8-c8jsl        0/1    ContainerCreating  0         1s
nginx-ingress-default-backend-7d5dd85c4c-wrzzq  0/1    ContainerCreating  0         1s

==> v1/Service
NAME                           TYPE          CLUSTER-IP     EXTERNAL-IP  PORT(S)                     AGE
nginx-ingress-controller       LoadBalancer  10.254.49.193  <pending>    80:31227/TCP,443:31316/TCP  1s
nginx-ingress-default-backend  ClusterIP     10.254.94.54   <none>       80/TCP                      1s

==> v1/ServiceAccount
NAME           SECRETS  AGE
nginx-ingress  1        1s

==> v1beta1/ClusterRole
NAME           AGE
nginx-ingress  1s

==> v1beta1/ClusterRoleBinding
NAME           AGE
nginx-ingress  1s

==> v1beta1/Deployment
NAME                           READY  UP-TO-DATE  AVAILABLE  AGE
nginx-ingress-controller       0/1    1           0          1s
nginx-ingress-default-backend  0/1    1           0          1s

==> v1beta1/Role
NAME           AGE
nginx-ingress  1s

==> v1beta1/RoleBinding
NAME           AGE
nginx-ingress  1s


NOTES:
The nginx-ingress controller has been installed.
It may take a few minutes for the LoadBalancer IP to be available.
You can watch the status by running 'kubectl --namespace default get services -o wide -w nginx-ingress-controller'

An example Ingress that makes use of the controller:

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  annotations:
    kubernetes.io/ingress.class: nginx
  name: example
  namespace: foo
spec:
  rules:
    - host: www.example.com
      http:
        paths:
          - backend:
              serviceName: exampleService
              servicePort: 80
            path: /
  # This section is only required if TLS is to be enabled for the Ingress
  tls:
      - hosts:
          - www.example.com
        secretName: example-tls

  If TLS is enabled for the Ingress, a Secret containing the certificate and key must also be provided:

  apiVersion: v1
  kind: Secret
  metadata:
    name: example-tls
    namespace: foo
  data:
    tls.crt: <base64 encoded cert>
    tls.key: <base64 encoded key>
  type: kubernetes.io/tls

Now we need to wait until the service gets an external IP address

$ kubectl get service
NAME                            TYPE           CLUSTER-IP       EXTERNAL-IP      PORT(S)                      AGE
echoserver                      ClusterIP      10.254.58.23     <none>           8080/TCP                     49m
kubernetes                      ClusterIP      10.254.0.1       <none>           443/TCP                      41d
nginx-ingress-controller        LoadBalancer   10.254.204.209   202.49.241.135   80:30722/TCP,443:30897/TCP   2m32s
nginx-ingress-default-backend   ClusterIP      10.254.68.138    <none>           80/TCP

$ openstack loadbalancer list | grep nginx
| 09d21949-528f-4afa-a1fb-9441b4555670 | kube_service_ea0613ef-4b48-4b22-b39a-cfb146c81c8a_default_nginx-ingress-controller | eac679e4896146e6827ce29d755fe289 | 10.0.0.16   | ACTIVE              | octavia  |

Now create an ingress that routes the incoming requests to the echo service based on the the URL path “/ping”. This requires the use of the annotation kubernetes.io/ingress.class: nginx

cat <<EOF | kubectl apply -f -
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
    name: test-http
    annotations:
        kubernetes.io/ingress.class: nginx
spec:
    rules:
    - host: test.example.com
      http:
        paths:
        - backend:
            serviceName: echoserver
            servicePort: 8080
          path: /ping
EOF

Wait for IP address to be allocated

$ kubectl get ingress -w

NAME        HOSTS              ADDRESS          PORTS   AGE
test-http   test.example.com   202.49.241.135   80      107s

Send a request to the /ping URL on the client IP address seen on the echo service

$ ip=202.49.241.135
$ curl -H "Host:test.example.com" http://$ip/ping

Hostname: echoserver-7cc8b87c6f-h8ls5

Pod Information:
  -no pod information available-

Server values:
  server_version=nginx: 1.13.3 - lua: 10008

Request Information:
  client_address=192.168.73.66
  method=GET
  real path=/ping
  query=
  request_version=1.1
  request_scheme=http
  request_uri=http://test.example.com:8080/ping

Request Headers:
  accept=*/*
  host=test.example.com
  user-agent=curl/7.54.0
  x-forwarded-for=10.0.0.14
  x-forwarded-host=test.example.com
  x-forwarded-port=80
  x-forwarded-proto=http
  x-original-uri=/ping
  x-real-ip=10.0.0.14
  x-request-id=157496f47a599ef1b2754eb910fa6b6c
  x-scheme=http

Request Body:
  -no body in request-

Cleanup¶

Before moving on to the next example let’s clean up the resources we created in this example.

$ kubectl delete ingress test-http
$ helm delete --purge nginx-ingress

Nginx ingress with TLS support¶

In this example we will add TLS support to our previous example.

For simplicity we will use a self signed certificate. The following code will create this for us.

$ if [ ! -f ./certs/tls.key ]; then
  mkdir certs
  openssl req -x509 -sha256 -nodes -days 365 -newkey rsa:2048 \
      -keyout certs/tls.key \
      -out certs/tls.crt \
      -subj "/CN=test.example.com/O=Integration"
fi

Next we will create a TLS secret based using the certificates created in the previous step.

$ kubectl create secret tls tls-secret-test-example-com --key certs/tls.key --cert certs/tls.crt

Label the secret so it’s easier to delete later

$ (kubectl get secret -l group=test-example-com 2>/dev/null | grep tls-secret-test-example-com) || kubectl label secret tls-secret-test-example-com group=test-example-com

As the helm config will remain the same as the previous example, we can go ahead and deploy the ingress controller.

$ helm install stable/nginx-ingress --name nginx-ingress -f nginx-ingress-controller-helm-values.yaml
NAME:   nginx-ingress
LAST DEPLOYED: Wed Aug 21 12:39:01 2019
NAMESPACE: default
STATUS: DEPLOYED

<-- output truncated for brevity -->

Once the loadbalancer is active and has an external IP we can create an ingress, the same as the previous example, that routes the incoming requests for test.example.com to the echo service based on the the URL path “/ping”. This time we will also add configuration for TLS support. This change adds the hosts that the ingress will accept traffic for and the cluster secret that will provide the certificate used for the encryption.

$ cat <<EOF | kubectl apply -f -
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: test-with-tls
  annotations:
      kubernetes.io/ingress.class: nginx
spec:
  rules:
      - host: test.example.com
        http:
          paths:
          - backend:
              serviceName: echoserver
              servicePort: 8080
            path: /ping
  tls:
      - hosts:
        - test.example.com
        secretName: tls-secret-test-example-com
EOF

Once the ingress is active and has been assigned an external IP address we can test the service

$ kubectl get ingress -w
NAME            HOSTS              ADDRESS          PORTS     AGE
test-with-tls   test.example.com   202.49.241.145   80, 443   41s

$ ip=202.49.241.145
$ curl -H "Host:test.example.com" https://$ip/ping --insecure
Hostname: echoserver-7cc8b87c6f-h8ls5

Pod Information:
  -no pod information available-

Server values:
  server_version=nginx: 1.13.3 - lua: 10008

Request Information:
  client_address=192.168.73.67
  method=GET
  real path=/ping
  query=
  request_version=1.1
  request_scheme=http
  request_uri=http://test.example.com:8080/ping

Request Headers:
  accept=*/*
  host=test.example.com
  user-agent=curl/7.54.0
  x-forwarded-for=10.0.0.14
  x-forwarded-host=test.example.com
  x-forwarded-port=443
  x-forwarded-proto=https
  x-original-uri=/ping
  x-real-ip=10.0.0.14
  x-request-id=2e1fa5e968414311d47076cbc3c6dcc7
  x-scheme=https

Request Body:
  -no body in request-

Cleanup¶

Before moving on to the next example let’s clean up the resources we created in this example.

$ kubectl delete ingress test-with-tls
$ helm delete --purge nginx-ingress

Nginx ingress with PROXY protocol support¶

For the final example we will enable support for the PROXY protocol. This provides visibility of the originating servers IP address to the backend services.

There are 2 sets of configuration we need to enable for this support to function.

The configuration for the Nginx ingress requires the following parameters to be enabled.

use-proxy-protocol Enables or disables the PROXY protocol to receive client connection (real IP address) information passed through proxy servers and load balancers.
use-forwarded-headers If true, NGINX passes the incoming X-Forwarded-* headers to upstreams. Use this option when NGINX is behind another L7 proxy / load balancer that is setting these headers.

If false, NGINX ignores incoming X-Forwarded-* headers, filling them with the request information it sees. Use this option if NGINX is exposed directly to the internet, or it’s behind a L3/packet-based load balancer that doesn’t alter the source IP in the packets.
compute-full-forwarded-for Will append the remote address to the X-Forwarded-For header instead of replacing it. When this option is enabled, the upstream application is responsible for extracting the client IP based on its own list of trusted proxies.

For the cluster itself the following annotation needs to be added to the configuration.

PROXY protocol This option provides support so that you can use a Service in LoadBalancer mode to configure a load balancer outside of Kubernetes itself, that will forward connections prefixed with PROXY protocol.

The load balancer will send an initial series of octets describing the incoming connection.

This is the actual configuration that will be used by the helm chart.

cat <<EOF > nginx-ingress-controller-helm-values.yaml
controller:
    publishService:
        enabled: true
    config:
        use-forward-headers: "true"
        compute-full-forward-for: "true"
        use-proxy-protocol: "true"
    service:
        annotations:
          loadbalancer.openstack.org/proxy-protocol: "true"
EOF

We install it as we have previously.

$ helm install stable/nginx-ingress --name nginx-ingress -f nginx-ingress-controller-helm-values.yaml
NAME:   nginx-ingress
LAST DEPLOYED: Wed Aug 21 14:10:01 2019
NAMESPACE: default
STATUS: DEPLOYED

<-- output truncated for brevity -->

Finally we can set up the ingress as we have for the previous examples.

cat <<EOF | kubectl apply -f -
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
    name: test-with-proxy
    annotations:
        kubernetes.io/ingress.class: nginx
spec:
    rules:
    - host: test.example.com
      http:
        paths:
        - backend:
            serviceName: echoserver
            servicePort: 8080
          path: /ping
EOF

Once the external IP is available we can test it with curl as we have previously. The important thing to note here is that now we can see the originating IP address included in the request headers.

$ kubectl get ingress
NAME              HOSTS              ADDRESS          PORTS   AGE
test-with-proxy   test.example.com   202.49.241.165   80      80s

$ ip=202.49.241.165
$ curl -H "Host:test.example.com" http://$ip/ping
Hostname: echoserver-7cc8b87c6f-h8ls5

Pod Information:
  -no pod information available-

Server values:
  server_version=nginx: 1.13.3 - lua: 10008

Request Information:
  client_address=192.168.73.68
  method=GET
  real path=/ping
  query=
  request_version=1.1
  request_scheme=http
  request_uri=http://test.example.com:8080/ping

Request Headers:
  accept=*/*
  host=test.example.com
  user-agent=curl/7.54.0
  x-forwarded-for=203.109.145.15
  x-forwarded-host=test.example.com
  x-forwarded-port=80
  x-forwarded-proto=http
  x-original-uri=/ping
  x-real-ip=203.109.145.15
  x-request-id=a244d459cce51cec15f5482fd4983709
  x-scheme=http

Request Body:
  -no body in request-

Cleanup¶

Before moving on to the next example let’s clean up the resources we created in this example.

$ kubectl delete ingress test-with-proxy
$ helm delete --purge nginx-ingress