Ingress configuration for k8s in different namespaces

Question

I need to configure Ingress Nginx on azure k8s, and my question is if is possible to have ingress configured in one namespace et. ingress-nginx and some serivces in other namespace eg. resources? My files looks like so:

# ingress-nginx.yaml
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: nginx-ingress-controller
  namespace: ingress-nginx
spec:
  replicas: 3
  selector:
    matchLabels:
      app: ingress-nginx
  template:
    metadata:
      labels:
        app: ingress-nginx
      annotations:
        prometheus.io/port: '10254'
        prometheus.io/scrape: 'true' 
    spec:
      containers:
        - name: nginx-ingress-controller
          image: quay.io/kubernetes-ingress-controller/nginx-ingress-controller:0.12.0
          args:
            - /nginx-ingress-controller
            - --default-backend-service=$(POD_NAMESPACE)/default-http-backend
            - --configmap=$(POD_NAMESPACE)/nginx-configuration
            - --tcp-services-configmap=$(POD_NAMESPACE)/tcp-services
            - --udp-services-configmap=$(POD_NAMESPACE)/udp-services
            - --annotations-prefix=nginx.ingress.kubernetes.io
            - --publish-service=$(POD_NAMESPACE)/ingress-nginx
          env:
            - name: POD_NAME
              valueFrom:
                fieldRef:
                  fieldPath: metadata.name
            - name: POD_NAMESPACE
              valueFrom:
                fieldRef:
                  fieldPath: metadata.namespace
          ports:
          - name: http
            containerPort: 80
          - name: https
            containerPort: 443
          livenessProbe:
            failureThreshold: 3
            httpGet:
              path: /healthz
              port: 10254
              scheme: HTTP
            initialDelaySeconds: 10
            periodSeconds: 10
            successThreshold: 1
            timeoutSeconds: 1
          readinessProbe:
            failureThreshold: 3
            httpGet:
              path: /healthz
              port: 10254
              scheme: HTTP
            periodSeconds: 10
            successThreshold: 1
            timeoutSeconds: 1

# configmap.yaml
kind: ConfigMap
apiVersion: v1
metadata:
  name: nginx-configuration
  namespace: ingress-nginx
  labels:
    app: ingress-nginx
---
kind: ConfigMap
apiVersion: v1
metadata:
  name: tcp-services
  namespace: ingress-nginx
---
kind: ConfigMap
apiVersion: v1
metadata:
  name: udp-services
  namespace: ingress-nginx
---
# default-backend.yaml
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: default-http-backend
  labels:
    app: default-http-backend
  namespace: ingress-nginx
spec:
  replicas: 1
  selector:
    matchLabels:
      app: default-http-backend
  template:
    metadata:
      labels:
        app: default-http-backend
    spec:
      terminationGracePeriodSeconds: 60
      containers:
      - name: default-http-backend
        # Any image is permissible as long as:
        # 1. It serves a 404 page at /
        # 2. It serves 200 on a /healthz endpoint
        image: gcr.io/google_containers/defaultbackend:1.4
        livenessProbe:
          httpGet:
            path: /healthz
            port: 8080
            scheme: HTTP
          initialDelaySeconds: 30
          timeoutSeconds: 5
        ports:
        - containerPort: 8080
        resources:
          limits:
            cpu: 10m
            memory: 20Mi
          requests:
            cpu: 10m
            memory: 20Mi
---
apiVersion: v1
kind: Service
metadata:
  name: default-http-backend
  namespace: ingress-nginx
  labels:
    app: default-http-backend
spec:
  ports:
  - port: 80
    targetPort: 8080
  selector:
    app: default-http-backend

kind: Service
apiVersion: v1
metadata:
  name: ingress-nginx
  namespace: ingress-nginx
  labels:
    app: ingress-nginx
spec:
  externalTrafficPolicy: Local
  type: LoadBalancer
  selector:
    app: ingress-nginx
  ports:
  - name: http
    port: 80
    targetPort: http
  - name: https
    port: 443
    targetPort: https

        # app-ingress.yaml
apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: app-ingress
  namespace: ingress-nginx
  annotations:
    kubernetes.io/ingress.class: nginx
    nginx.ingress.kubernetes.io/rewrite-target: /
spec:
  tls:
    - hosts:
      - api-sand.fake.com
  rules:
  - host: api-sand.fake.com
    http:
      paths:
      - backend:
          serviceName: api-sand
          servicePort: 80
        path: /

And then I have some app running in the resources namespace, and problem is that I am getting the following error

error obtaining service endpoints: error getting service resources/api-sand from the cache: service resources/api-sand was not found

If I deploy api-sand in the same namespace where ingress is then this service works fine.

Solution 1

Instead of creating the ingress app-ingress in ingress-nginx namespace you should create it in the namespace where you have the service api-sandand the pod.

Alternatively there is way to achieve ingress in one namespace and service in another namespace via externalName.Checkout Kubernetes Cross Namespace Ingress Network

Here is an example referred from here.

kind: Service
apiVersion: v1
metadata:
  name: my-service
spec:
  type: ExternalName
  externalName: test-service.namespacename.svc.cluster.local

---

apiVersion: extensions/v1beta1
kind: Ingress
metadata:
  name: example-ingress
  annotations:
    kubernetes.io/ingress.class: "nginx"
spec:
  rules:
  - host: example.com
    http:
      paths:
      - path: /
        backend:
          serviceName: my-service
          servicePort: 80

Solution 2

I would like to simplify the answer a bit further for those who are reletively new to Kubernetes and its ingress options in particular. There are 2 separate things that need to be present for ingress to work:

1. Ingress Controller(essentially a separate Pod/Deployment along with a Service that can be used to utilize routing and proxying. Based on nginx container for example);
2. Ingress rules(a separate Kubernetes resourse with kind: Ingress. Will only take effect if Ingress Controller is already deployed)

Now, Ingress Controller can be deployed in any namespace and is, in fact, usually deployed in a namespace separate from your app services. It can out-of-the-box see Ingress rules in all namespaces in the cluster and will pick them up.
The Ingress rules, however, must reside in the namespace where the app that they configure reside.

There are some workarounds for that, but this is the most common approach.

Solution 3

It's possible actually, you can define ingress and a service with ExternalName type in namespace A, while the ExternalName points to DNS of the service in namespace B. For further details, please refer to this answer: https://stackoverflow.com/a/51899301/2995449

Solution 4

Outside traffic comes through ingress controller service that is responsible for routing the traffic based on the defined routing rules or what we call ingress rules in k8s world.

In other words, ingress resources are just routing rules (think of it in away that's similar to DNS records) so when you define an ingress resource you just defined a rule for ingress controller to work on and route traffic based on such defined rules.

Solution:

1. Since Ingress are nothing but routing rules, you could define such rules anywhere in the cluster (in any namespace) and controller should pick them up as it monitors creation of such resources and react accordingly.

   Here's how to create ingress easily using kubectl

   kubectl create ingress <name> -n namespaceName --rule="host/prefix=serviceName:portNumber"

   Note: Add --dry-run=client -oyaml to generate yaml manifest file

2. Or you may create a service of type ExternalName in the same namespace where you have defined your ingress. such external service can point to any URL (a service that lives outside namespace or even k8s cluster)

   Here's an example that shows how to create an ExternalName service using kubectl:

   kubectl create service externalname ingress-ns -n namespaceName --external-name=serviceName.namespace.svc.cluster.local --tcp=80:80 --dry-run=client -oyaml

this should generate something similar to the following:

kind: Service
apiVersion: v1
metadata:
  name: nginx
  namespace: ingress-ns
spec:
  type: ExternalName
  externalName: serviceName.namespace.svc.cluster.local #or any external svc
  ports:
  - port: 80 #specify the port of service you want to expose 
    targetPort: 80 #port of external service 

As described above, create an ingress as below: kubectl create ingress <name> -n namespaceName --rule="host/prefix=serviceName:portNumber"

Note: Add --dry-run=client -oyaml to generate yaml manifest file

View more solutions

Comments

[Choco]

nice answer: Can I ask if when creating the Ingress rules, do they all need to be defined in a single Ingress rule, or can there be multiple Ingress rules within each namespace, each mapping to a service within that namespace? tx

[yuranos]

Well, it's not as simple as with Network Policies. Not that NPs are simple, but there's a relatively straightforward way to merge them. With Ingress it can't be so simple because it's harder to standardize underlying Ingress Controller implementation, especially since those are built on top of completely independent components that existed long before Kubernetes inception, like Nginx. Having said that you are not the first one, @Choco, who entertain that thought: github.com/kubernetes/ingress-nginx/issues/1539

[Lukas]

How did you get it working? I keep getting service "default/my-service" is type "ExternalName", expected "NodePort" or "LoadBalancer";

[metasim]

I couldn't get it to work either. I'm wondering if it's a provider and/or version issue? I'm on EKS 1.14...

[Damith Udayanga]

(<error: endpoints "scc-worker-service" not found>) when describing the ingress, show this error. any clue to resolve?

[Marc]

So if I have 100 Microservices which should be accessible via 1 hostname (and separate paths) I have to put all those services in the same namespace so that they can use the same ingress resource??

[yuranos]

It's such scenarios, you'd need to come up with a more advances networking architecture. Maybe it will be several levels of ingresses/proxies/gateway. Maybe ExternalName services. Is it a real or a hypothetical example?

[YoShade]

if you read in the example link they speak only on ngnix plus

[Andy Lorenz]

hey @yuranos your answer looks great, and I do appreciate how you've addressed it to "those who are reletively new to Kubernetes" - but could you please expand on what you mean by "app" when you say "must reside in the namespace where the app that they configure reside" ? There's no such thing called an "app", so does that mean the "service", the "deployment", or all of them? thanks (from a relative newbie!)

[yuranos]

Hey @AndyLorenz. Thanks. Yes, it means any workload resource(deployment, job, replicaset) that configures/controls your pods. They all reside in the same namespace.

[Suge]

Should I put the LoadBalancer service beside Ingress Controller, or beside ingress?

[yuranos]

If you mean k8s service, then it's a separate level of abstraction, below your ingress. Ingress usually contains rules to point to k8s services.