Use kubectl to
deployment called lab2
kubectl create deployment lab2 --image=sirfragalot/docker-demo:dcus \
--dry-run=client -o yaml | tee lab2.yaml
kubectl apply -f lab2.yaml
kubectl scale deployment lab2 --replicas=5
kubectl scale deployment lab2 --replicas=1
kubectl set image deployment/lab2 docker-demo=nginx:latest --record
kubectl rollout history deployment/lab2
kubectl rollout undo deployment/lab2
kubectl get deployments \
--output=custom-columns=NAME:.metadata.name,IMAGE:.spec.template.spec.containers.*.image
Use kubectl to
deployment called deployme using image busybox with 3 replicas and command sleep 3600.deployment called deployme using image ubuntu
# create deployment called deployme declarative
kubectl create deployment deployme --image=busybox --replicas=3 --dry-run=client -o yaml -- sleep 3600 > deployme.yaml
kubectl apply -f deployme-1.yaml
# OR (imperative)
kubectl create deployment deployme --image=busybox --replicas=3 -- sleep 3600
# select one pod to look at using exec
kubectl exec deployme-XXXXXXXX-XXXX -- ps
# alter the image of deployment deployme
kubectl set image deployment deployme busybox=ubuntu
# check that pods have been deployed
kubectl get pods
It is bad practice not to specify the image tag (as it will default to ’latest’). It is best practice is to specify the image tag