{"id":6272,"date":"2019-07-26T18:22:34","date_gmt":"2019-07-26T18:22:34","guid":{"rendered":"https:\/\/www.devopsschool.com\/blog\/?p=6272"},"modified":"2021-11-16T11:13:52","modified_gmt":"2021-11-16T11:13:52","slug":"working-with-kubernetes-cluster-using-kubectl-part-2","status":"publish","type":"post","link":"https:\/\/www.devopsschool.com\/blog\/working-with-kubernetes-cluster-using-kubectl-part-2\/","title":{"rendered":"Kubernetes Commands: kubectl run – Tutorials and Examples"},"content":{"rendered":"
#Imperatively working with your cluster. Run will \"generate\" a Deployment by default.<\/span>\n#This is pulling a specified image from Google's container registry.<\/span>\n#kubectl run, will convert a pod creation into a \"Deployment generation\"<\/span>\n#http:\/\/kubernetes.io\/docs\/user-guide\/kubectl-conventions\/#generators <\/span>\nkubectl run hello-world --image=gcr.io\/google-samples\/hello-app:1.0<\/span>\n\n#But let's deploy a single pod too...<\/span>\nkubectl run hello-world-pod --image=gcr.io\/google-samples\/hello-app:1.0<\/span> --generator=run-pod\/v1\n\n#Let's follow our pod and deployment status<\/span>\nkubectl get pods\nkubectl get deployment\nkubectl get pods -o wide\n\n#Remember, k8s is a container orchestrator and it's starting up containers on Nodes.<\/span>\n#Open a second terminal and ssh into the node that hello-world pod is running on.<\/span>\nssh aen@c1-node1\nsudo docker ps\nexit<\/span>\n\n#Back on the Master, we can pull the logs from the container. Which is going to be anything written to stdout. <\/span>\n#Maybe something went wrong inside our app, and our pod won't start. This is useful for troubleshooting.<\/span>\nkubectl logs hello-world-pod\n\n#Starting a process inside a container inside a pod.<\/span>\n#We can use this to launch any process as long as the executable\/binary is in the container.<\/span>\n#Launch a shell into the container. Callout that this is on the *pod* network.<\/span>\nkubectl exec -it hello-world-pod  -- \/bin\/sh\nhostname\nip addr\nexit<\/span>\n\n#Remember that first kubectl run we executed, it created a Deployment for us.<\/span>\n#Let's look more closely at the deployment<\/span>\n#Deployments are made of ReplicaSets!<\/span>\nkubectl get deployment hello-world\nkubectl get replicaset\nkubectl get pods\n\n#Let's take a closer look at our pod.<\/span>\n#Walk through the pods Events...<\/span>\n#Name, Containers, Ports, Conditions, and Events. <\/span>\n#Deployments are made of ReplicaSets!<\/span>\nkubectl describe deployment hello-world | more\n\n#Let's see what describe can tell us about a deployed Pod.<\/span>\n#Check out the Name, Node, Status, Containers, and events.<\/span>\nkubectl get pods\nkubectl describe pods $PASTEPODNAMEHERE | more\n\n#Expose the Deployment as a Serivce.<\/span>\n#This will create a Service for the ReplicaSet behind the Deployment<\/span>\n#We are exposing our serivce on port 80, connecting to an application running on 8080 in our pod.<\/span>\n#Port: Interal Cluster Port, the Service's port. You will point cluster resources here.<\/span>\n#TargetPort: The Pod's Serivce Port, your application. That one we defined when we started the pods.<\/span>\nkubectl expose deployment hello-world --port=80<\/span> --target-port=8080<\/span>\n\n#Check out the IP: and Port:, that's where we'll access this service.<\/span>\nkubectl get service hello-world\n\n#We can also get that information from using get<\/span>\nkubectl describe service hello-world\n\n#Access the service inside the cluster<\/span>\ncurl http:\/\/$SERVCIEIP:$PORT<\/span>\n\n#Access the pod's application directly, useful for troubleshooting.<\/span>\nkubectl get endpoints hello-world\ncurl http:\/\/$ENDPOINT:$TARGETORT<\/span>\n\n#Using kubectl to generate yaml or json files of our imperitive configuration.<\/span>\nkubectl get service hello-world -o yaml\nkubectl get service hello-world -o json\n\n#Exported resources are stripped of cluster-specific information.<\/span>\nkubectl get service hello-world -o yaml --export > service-hello-world.yaml\nkubectl get deployment hello-world -o yaml --export > deployment-hello-world.yaml\nls *.yaml\n\nmore service-hello-world.yaml\n\n#We can ask the API for more information about an object<\/span>\nkubectl explain service | more\n\n#And drill down further if needed, includes very good explanation of what's available for that resource<\/span>\nkubectl explain service.spec | more\nkubectl explain service.spec.ports\nkubectl explain service.spec.ports.targetPort\n\n#Let's remove everything we created imperitively and start over using a declarative model<\/span>\nkubectl delete service hello-world\nkubectl delete deployment hello-world\nkubectl delete pod hello-world-pod\nkubectl get all\n\n#Deploying applications declaratively<\/span>\n#we can use apply to create our resources from yaml.<\/span>\nkubectl apply -f deployment-hello-world.yaml\nkubectl apply -f service-hello-world.yaml \n\n#This re-creates everything we created, but in yaml<\/span>\nkubectl get all\n\n#scale up our deployment<\/span>\nvi deployment-hello-world.yaml\nChange replicas from 1<\/span> to 2<\/span>\n     replicas: 2<\/span>\n\n#update our configuration with apply<\/span>\nkubectl apply -f deployment-hello-world.yaml\n\n#And check the current configuration of our deployment<\/span>\nkubectl get deployment hello-world\n\n#Repeat the curl access to see the load balancing of the HTTP request<\/span>\nkubectl get service hello-world\ncurl http:\/\/$SERVICEIP:PORT<\/span>\n\n#We can edit the resources \"on the fly\" with kubectl edit. But this isn't reflected in our yaml. But is<\/span>\n#persisted in the etcd database...cluster store. Change 2 to 3.<\/span>\nkubectl edit deployment hello-world\n\n#Let's clean up our deployment and remove everything<\/span>\nkubectl delete deployment hello-world\nkubectl delete service hello-world\n\nkubectl get all\n<\/code><\/span>Code language:<\/span> PHP<\/span> (<\/span>php<\/span>)<\/span><\/small><\/pre>\n\n