This example shows how to build a simple multi-tier web application using Kubernetes and Docker.
The example combines a web frontend, a redis master for storage and a replicated set of redis slaves.
This example assumes that you have forked the repository and turned up a Kubernetes cluster:
$ cd kubernetes
$ hack/dev-build-and-up.shUse the file examples/guestbook-go/redis-master-controller.json to create a replication controller which manages a single pod. The pod runs a redis key-value server in a container. Using a replication controller is the preferred way to launch long-running pods, even for 1 replica, so the pod will benefit from self-healing mechanism in kubernetes.
Create the redis master replication controller in your Kubernetes cluster using the kubectl CLI:
$ cluster/kubectl.sh create -f examples/guestbook-go/redis-master-controller.jsonOnce that's up you can list the replication controllers in the cluster:
$ cluster/kubectl.sh get rc
CONTROLLER CONTAINER(S) IMAGE(S) SELECTOR REPLICAS
redis-master-controller redis-master gurpartap/redis name=redis,role=master 1List pods in cluster to verify the master is running. You'll see a single redis master pod. It will also display the machine that the pod is running on once it gets placed (may take up to thirty seconds).
$ cluster/kubectl.sh get pods
POD IP CONTAINER(S) IMAGE(S) HOST LABELS STATUS
redis-master-pod-hh2gd 10.244.3.7 redis-master gurpartap/redis kubernetes-minion-4.c.lucid-walker-725.internal/104.154.49.134 name=redis,role=master RunningIf you ssh to that machine, you can run docker ps to see the actual pod:
me@workstation$ gcloud compute ssh --zone us-central1-b kubernetes-minion-4
me@kubernetes-minion-3:~$ sudo docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS
d5c458dabe50 gurpartap/redis:latest "/usr/local/bin/redi 5 minutes ago Up 5 minutes(Note that initial docker pull may take a few minutes, depending on network conditions.)
A Kubernetes 'service' is a named load balancer that proxies traffic to one or more containers. The services in a Kubernetes cluster are discoverable inside other containers via environment variables. Services find the containers to load balance based on pod labels.
The pod that you created in Step One has the label name=redis and role=master. The selector field of the service determines which pods will receive the traffic sent to the service. Use the file examples/guestbook-go/redis-master-service.json to create the service in the kubectl cli:
$ cluster/kubectl.sh create -f examples/guestbook-go/redis-master-service.json
$ cluster/kubectl.sh get services
NAME LABELS SELECTOR IP PORT
redis-master <none> name=redis,role=master 10.0.186.234 6379This will cause all new pods to see the redis master apparently running on $REDIS_MASTER_SERVICE_HOST at port 6379. Once created, the service proxy on each node is configured to set up a proxy on the specified port (in this case port 6379).
Although the redis master is a single pod, the redis read slaves are a 'replicated' pod. In Kubernetes, a replication controller is responsible for managing multiple instances of a replicated pod.
Use the file examples/guestbook-go/redis-slave-controller.json to create the replication controller:
$ cluster/kubectl.sh create -f examples/guestbook-go/redis-slave-controller.json
$ cluster/kubectl.sh get rc
CONTROLLER CONTAINER(S) IMAGE(S) SELECTOR REPLICAS
redis-master-controller redis-master gurpartap/redis name=redis,role=master 1
redis-slave-controller redis-slave gurpartap/redis name=redis,role=slave 2The redis slave configures itself by looking for the Kubernetes service environment variables in the container environment. In particular, the redis slave is started with the following command:
redis-server --slaveof redis-master 6379Once that's up you can list the pods in the cluster, to verify that the master and slaves are running:
$ cluster/kubectl.sh get pods
POD IP CONTAINER(S) IMAGE(S) HOST LABELS STATUS
redis-master-pod-hh2gd 10.244.3.7 redis-master gurpartap/redis kubernetes-minion-4.c.lucid-walker-725.internal/104.154.49.134 name=redis,role=master Running
redis-slave-controller-i7hvs 10.244.2.7 redis-slave gurpartap/redis kubernetes-minion-3.c.lucid-walker-725.internal/104.154.52.39 name=redis,role=slave Running
redis-slave-controller-nyxxv 10.244.1.6 redis-slave gurpartap/redis kubernetes-minion-2.c.lucid-walker-725.internal/130.211.144.5 name=redis,role=slave RunningYou will see a single redis master pod and two redis slave pods.
Just like the master, we want to have a service to proxy connections to the read slaves. In this case, in addition to discovery, the slave service provides transparent load balancing to clients. The service specification for the slaves is in examples/guestbook-go/redis-slave-service.json
This time the selector for the service is name=redis,role=slave, because that identifies the pods running redis slaves. It may also be helpful to set labels on your service itself--as we've done here--to make it easy to locate them later.
Now that you have created the service specification, create it in your cluster with the kubectl CLI:
$ cluster/kubectl.sh create -f examples/guestbook-go/redis-slave-service.json
$ cluster/kubectl.sh get services
NAME LABELS SELECTOR IP PORT
redis-master <none> name=redis,role=master 10.0.186.234 6379
redis-slave name=redis,role=slave name=redis,role=slave 10.0.22.180 6379This is a simple Go net/http (negroni based) server that is configured to talk to either the slave or master services depending on whether the request is a read or a write. It exposes a simple JSON interface, and serves a jQuery-Ajax based UX. Like the redis read slaves it is a replicated service instantiated by a replication controller.
The pod is described in the file examples/guestbook-go/guestbook-controller.json. Using this file, you can turn up your guestbook with:
$ cluster/kubectl.sh create -f examples/guestbook-go/guestbook-controller.json
$ cluster/kubectl.sh get replicationControllers
CONTROLLER CONTAINER(S) IMAGE(S) SELECTOR REPLICAS
guestbook-controller guestbook kubernetes/guestbook:v2 name=guestbook 3
redis-master-controller redis-master gurpartap/redis name=redis,role=master 1
redis-slave-controller redis-slave gurpartap/redis name=redis,role=slave 2Once that's up (it may take ten to thirty seconds to create the pods) you can list the pods in the cluster, to verify that the master, slaves and guestbook frontends are running:
$ cluster/kubectl.sh get pods
POD IP CONTAINER(S) IMAGE(S) HOST LABELS STATUS
guestbook-controller-182tv 10.244.2.8 guestbook kubernetes/guestbook:v2 kubernetes-minion-3.c.lucid-walker-725.internal/104.154.52.39 name=guestbook Running
guestbook-controller-jzjpe 10.244.0.7 guestbook kubernetes/guestbook:v2 kubernetes-minion-1.c.lucid-walker-725.internal/104.154.37.86 name=guestbook Running
guestbook-controller-zwk1b 10.244.3.8 guestbook kubernetes/guestbook:v2 kubernetes-minion-4.c.lucid-walker-725.internal/104.154.49.134 name=guestbook Running
redis-master-pod-hh2gd 10.244.3.7 redis-master gurpartap/redis kubernetes-minion-4.c.lucid-walker-725.internal/104.154.49.134 name=redis,role=master Running
redis-slave-controller-i7hvs 10.244.2.7 redis-slave gurpartap/redis kubernetes-minion-3.c.lucid-walker-725.internal/104.154.52.39 name=redis,role=slave Running
redis-slave-controller-nyxxv 10.244.1.6 redis-slave gurpartap/redis kubernetes-minion-2.c.lucid-walker-725.internal/130.211.144.5 name=redis,role=slave RunningYou will see a single redis master pod, two redis slaves, and three guestbook pods.
Just like the others, you want a service to group your guestbook pods. The service specification for the guestbook is in examples/guestbook-go/guestbook-service.json. There's a twist this time - because we want it to be externally visible, we set the createExternalLoadBalancer flag on the service.
$ cluster/kubectl.sh create -f examples/guestbook-go/guestbook-service.json
$ cluster/kubectl.sh get services
NAME LABELS SELECTOR IP PORT
guestbook <none> name=guestbook 10.0.12.110 3000
redis-master <none> name=redis,role=master 10.0.186.234 6379
redis-slave name=redis,role=slave name=redis,role=slave 10.0.22.180 6379To play with the service itself, find the external IP of the load balancer:
$ cluster/kubectl.sh get services guestbook -o template --template='{{index . "publicIPs"}}'
current-context: "kubernetes-satnam_kubernetes"
Running: cluster/../cluster/gce/../../_output/dockerized/bin/linux/amd64/kubectl get services guestbook -o template --template={{index . "publicIPs"}}
[104.154.87.59]$
and then visit port 3000 of that IP address e.g. http://104.154.87.59:3000.
You may need to open the firewall for port 3000 using the console or the gcloud tool. The following command will allow traffic from any source to instances tagged kubernetes-minion:
$ gcloud compute firewall-rules create --allow=tcp:3000 --target-tags=kubernetes-minion kubernetes-minion-3000If you are running Kubernetes locally, you can just visit http://localhost:3000 For details about limiting traffic to specific sources, see the GCE firewall documentation.
You should delete the service which will remove any associated resources that were created e.g. load balancers, forwarding rules and target pools. All the resources (pods, replication controllers and service) can be deleted with a single command:
$ cluster/kubectl.sh delete -f examples/guestbook-go
current-context: "kubernetes-satnam_kubernetes"
Running: cluster/../cluster/gce/../../_output/dockerized/bin/linux/amd64/kubectl delete -f examples/guestbook-go
guestbook-controller
guestbook
redis-master-controller
redis-master
redis-slave-controller
redis-slaveTo turn down a Kubernetes cluster:
$ cluster/kube-down.sh