Now that you have a continuously running, replicated application you can expose it on a network. Before discussing the Kubernetes approach to networking, it is worthwhile to contrast it with the "normal" way networking works with Docker.
By default, Docker uses host-private networking, so containers can talk to other containers only if they are on the same machine. In order for Docker containers to communicate across nodes, there must be allocated ports on the machine's own IP address, which are then forwarded or proxied to the containers. This obviously means that containers must either coordinate which ports they use very carefully or ports must be allocated dynamically.
Coordinating port allocations across multiple developers or teams that provide containers is very difficult to do at scale, and exposes users to cluster-level issues outside of their control. Kubernetes assumes that pods can communicate with other pods, regardless of which host they land on. Kubernetes gives every pod its own cluster-private IP address, so you do not need to explicitly create links between pods or map container ports to host ports. This means that containers within a Pod can all reach each other's ports on localhost, and all pods in a cluster can see each other without NAT. The rest of this document elaborates on how you can run reliable services on such a networking model.
This guide uses a simple nginx server to demonstrate proof of concept.
We did this in a previous example, but let's do it once again and focus on the networking perspective. Create an nginx Pod, and note that it has a container port specification:
service/networking/run-my-nginx.yaml
This makes it accessible from any node in your cluster. Check the nodes the Pod is running on:
kubectl apply -f ./run-my-nginx.yaml kubectl get pods -l run=my-nginx -o wide
NAME READY STATUS RESTARTS AGE IP NODE
my-nginx-3800858182-jr4a2 1/1 Running 0 13s 10.244.3.4 kubernetes-minion-905m
my-nginx-3800858182-kna2y 1/1 Running 0 13s 10.244.2.5 kubernetes-minion-ljyd
Check your pods' IPs:
kubectl get pods -l run=my-nginx -o yaml | grep podIP
podIP: 10.244.3.4
podIP: 10.244.2.5
You should be able to ssh into any node in your cluster and curl both IPs. Note that the containers are not using port 80 on the node, nor are there any special NAT rules to route traffic to the pod. This means you can run multiple nginx pods on the same node all using the same containerPort and access them from any other pod or node in your cluster using IP. Like Docker, ports can still be published to the host node's interfaces, but the need for this is radically diminished because of the networking model.
You can read more about the Kubernetes Networking Model if you're curious.
So we have pods running nginx in a flat, cluster wide, address space. In theory, you could talk to these pods directly, but what happens when a node dies? The pods die with it, and the Deployment will create new ones, with different IPs. This is the problem a Service solves.
A Kubernetes Service is an abstraction which defines a logical set of Pods running somewhere in your cluster, that all provide the same functionality. When created, each Service is assigned a unique IP address (also called clusterIP). This address is tied to the lifespan of the Service, and will not change while the Service is alive. Pods can be configured to talk to the Service, and know that communication to the Service will be automatically load-balanced out to some pod that is a member of the Service.
You can create a Service for your 2 nginx replicas with kubectl expose:
kubectl expose deployment/my-nginx
service/my-nginx exposed
This is equivalent to kubectl apply -f the following yaml:
service/networking/nginx-svc.yaml
This specification will create a Service which targets TCP port 80 on any Pod with the run: my-nginx label, and expose it on an abstracted Service port (targetPort: is the port the container accepts traffic on, port: is the abstracted Service port, which can be any port other pods use to access the Service). View Service API object to see the list of supported fields in service definition. Check your Service:
kubectl get svc my-nginx
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
my-nginx ClusterIP 10.0.162.149 <none> 80/TCP 21s
As mentioned previously, a Service is backed by a group of Pods. These Pods are exposed through endpoints. The Service's selector will be evaluated continuously and the results will be POSTed to an Endpoints object also named my-nginx. When a Pod dies, it is automatically removed from the endpoints, and new Pods matching the Service's selector will automatically get added to the endpoints. Check the endpoints, and note that the IPs are the same as the Pods created in the first step:
kubectl describe svc my-nginx
Name: my-nginx
Namespace: default
Labels: run=my-nginx
Annotations: <none>
Selector: run=my-nginx
Type: ClusterIP
IP: 10.0.162.149
Port: <unset> 80/TCP
Endpoints: 10.244.2.5:80,10.244.3.4:80
Session Affinity: None
Events: <none>
kubectl get ep my-nginx
NAME ENDPOINTS AGE
my-nginx 10.244.2.5:80,10.244.3.4:80 1m
You should now be able to curl the nginx Service on <CLUSTER-IP>:<PORT> from any node in your cluster. Note that the Service IP is completely virtual, it never hits the wire. If you're curious about how this works you can read more about the service proxy.
Kubernetes supports 2 primary modes of finding a Service - environment variables and DNS. The former works out of the box while the latter requires the CoreDNS cluster addon.
enableServiceLinks flag to false on the pod spec. When a Pod runs on a Node, the kubelet adds a set of environment variables for each active Service. This introduces an ordering problem. To see why, inspect the environment of your running nginx Pods (your Pod name will be different):
kubectl exec my-nginx-3800858182-jr4a2 -- printenv | grep SERVICE
KUBERNETES_SERVICE_HOST=10.0.0.1
KUBERNETES_SERVICE_PORT=443
KUBERNETES_SERVICE_PORT_HTTPS=443
Note there's no mention of your Service. This is because you created the replicas before the Service. Another disadvantage of doing this is that the scheduler might put both Pods on the same machine, which will take your entire Service down if it dies. We can do this the right way by killing the 2 Pods and waiting for the Deployment to recreate them. This time around the Service exists before the replicas. This will give you scheduler-level Service spreading of your Pods (provided all your nodes have equal capacity), as well as the right environment variables:
kubectl scale deployment my-nginx --replicas=0; kubectl scale deployment my-nginx --replicas=2; kubectl get pods -l run=my-nginx -o wide
NAME READY STATUS RESTARTS AGE IP NODE
my-nginx-3800858182-e9ihh 1/1 Running 0 5s 10.244.2.7 kubernetes-minion-ljyd
my-nginx-3800858182-j4rm4 1/1 Running 0 5s 10.244.3.8 kubernetes-minion-905m
You may notice that the pods have different names, since they are killed and recreated.
kubectl exec my-nginx-3800858182-e9ihh -- printenv | grep SERVICE
KUBERNETES_SERVICE_PORT=443
MY_NGINX_SERVICE_HOST=10.0.162.149
KUBERNETES_SERVICE_HOST=10.0.0.1
MY_NGINX_SERVICE_PORT=80
KUBERNETES_SERVICE_PORT_HTTPS=443
Kubernetes offers a DNS cluster addon Service that automatically assigns dns names to other Services. You can check if it's running on your cluster:
kubectl get services kube-dns --namespace=kube-system
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kube-dns ClusterIP 10.0.0.10 <none> 53/UDP,53/TCP 8m
The rest of this section will assume you have a Service with a long lived IP (my-nginx), and a DNS server that has assigned a name to that IP. Here we use the CoreDNS cluster addon (application name kube-dns), so you can talk to the Service from any pod in your cluster using standard methods (e.g. gethostbyname()). If CoreDNS isn't running, you can enable it referring to the CoreDNS README or Installing CoreDNS. Let's run another curl application to test this:
kubectl run curl --image=radial/busyboxplus:curl -i --tty
Waiting for pod default/curl-131556218-9fnch to be running, status is Pending, pod ready: false
Hit enter for command prompt
Then, hit enter and run nslookup my-nginx:
[ root@curl-131556218-9fnch:/ ]$ nslookup my-nginx Server: 10.0.0.10 Address 1: 10.0.0.10 Name: my-nginx Address 1: 10.0.162.149
Till now we have only accessed the nginx server from within the cluster. Before exposing the Service to the internet, you want to make sure the communication channel is secure. For this, you will need:
You can acquire all these from the nginx https example. This requires having go and make tools installed. If you don't want to install those, then follow the manual steps later. In short:
make keys KEY=/tmp/nginx.key CERT=/tmp/nginx.crt kubectl create secret tls nginxsecret --key /tmp/nginx.key --cert /tmp/nginx.crt
secret/nginxsecret created
kubectl get secrets
NAME TYPE DATA AGE
default-token-il9rc kubernetes.io/service-account-token 1 1d
nginxsecret kubernetes.io/tls 2 1m
And also the configmap:
kubectl create configmap nginxconfigmap --from-file=default.conf
configmap/nginxconfigmap created
kubectl get configmaps
NAME DATA AGE
nginxconfigmap 1 114s
Following are the manual steps to follow in case you run into problems running make (on windows for example):
# Create a public private key pair openssl req -x509 -nodes -days 365 -newkey rsa:2048 -keyout /d/tmp/nginx.key -out /d/tmp/nginx.crt -subj "/CN=my-nginx/O=my-nginx" # Convert the keys to base64 encoding cat /d/tmp/nginx.crt | base64 cat /d/tmp/nginx.key | base64
Use the output from the previous commands to create a yaml file as follows. The base64 encoded value should all be on a single line.
apiVersion: "v1" kind: "Secret" metadata: name: "nginxsecret" namespace: "default" type: kubernetes.io/tls data: tls.crt: "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" tls.key: "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"
Now create the secrets using the file:
kubectl apply -f nginxsecrets.yaml kubectl get secrets
NAME TYPE DATA AGE
default-token-il9rc kubernetes.io/service-account-token 1 1d
nginxsecret kubernetes.io/tls 2 1m
Now modify your nginx replicas to start an https server using the certificate in the secret, and the Service, to expose both ports (80 and 443):
service/networking/nginx-secure-app.yaml
Noteworthy points about the nginx-secure-app manifest:
/etc/nginx/ssl. This is setup before the nginx server is started.kubectl delete deployments,svc my-nginx; kubectl create -f ./nginx-secure-app.yaml
At this point you can reach the nginx server from any node.
kubectl get pods -o yaml | grep -i podip
podIP: 10.244.3.5
node $ curl -k https://10.244.3.5
...
<h1>Welcome to nginx!</h1>
Note how we supplied the -k parameter to curl in the last step, this is because we don't know anything about the pods running nginx at certificate generation time, so we have to tell curl to ignore the CName mismatch. By creating a Service we linked the CName used in the certificate with the actual DNS name used by pods during Service lookup. Let's test this from a pod (the same secret is being reused for simplicity, the pod only needs nginx.crt to access the Service):
service/networking/curlpod.yaml
kubectl apply -f ./curlpod.yaml kubectl get pods -l app=curlpod
NAME READY STATUS RESTARTS AGE
curl-deployment-1515033274-1410r 1/1 Running 0 1m
kubectl exec curl-deployment-1515033274-1410r -- curl https://my-nginx --cacert /etc/nginx/ssl/tls.crt ... <title>Welcome to nginx!</title> ...
For some parts of your applications you may want to expose a Service onto an external IP address. Kubernetes supports two ways of doing this: NodePorts and LoadBalancers. The Service created in the last section already used NodePort, so your nginx HTTPS replica is ready to serve traffic on the internet if your node has a public IP.
kubectl get svc my-nginx -o yaml | grep nodePort -C 5
uid: 07191fb3-f61a-11e5-8ae5-42010af00002
spec:
clusterIP: 10.0.162.149
ports:
- name: http
nodePort: 31704
port: 8080
protocol: TCP
targetPort: 80
- name: https
nodePort: 32453
port: 443
protocol: TCP
targetPort: 443
selector:
run: my-nginx
kubectl get nodes -o yaml | grep ExternalIP -C 1
- address: 104.197.41.11
type: ExternalIP
allocatable:
--
- address: 23.251.152.56
type: ExternalIP
allocatable:
...
$ curl https://<EXTERNAL-IP>:<NODE-PORT> -k
...
<h1>Welcome to nginx!</h1>
Let's now recreate the Service to use a cloud load balancer. Change the Type of my-nginx Service from NodePort to LoadBalancer:
kubectl edit svc my-nginx kubectl get svc my-nginx
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
my-nginx LoadBalancer 10.0.162.149 xx.xxx.xxx.xxx 8080:30163/TCP 21s
curl https://<EXTERNAL-IP> -k
...
<title>Welcome to nginx!</title>
The IP address in the EXTERNAL-IP column is the one that is available on the public internet. The CLUSTER-IP is only available inside your cluster/private cloud network.
Note that on AWS, type LoadBalancer creates an ELB, which uses a (long) hostname, not an IP. It's too long to fit in the standard kubectl get svc output, in fact, so you'll need to do kubectl describe service my-nginx to see it. You'll see something like this:
kubectl describe service my-nginx ... LoadBalancer Ingress: a320587ffd19711e5a37606cf4a74574-1142138393.us-east-1.elb.amazonaws.com ...
© 2022 The Kubernetes Authors
Documentation Distributed under CC BY 4.0.
https://kubernetes.io/docs/concepts/services-networking/connect-applications-service