Can I know why kubectl run sometimes create deployment and sometimes pod.
You can see 1st one creates pod and 2nd one creates deployment. only diff is --restart=Never
// 1
chams#master:~/yml$ kubectl run ng --image=ngnix --command --restart=Never --dry-run -o yaml
apiVersion: v1
kind: Pod
..
status: {}
//2
chams#master:~/yml$ kubectl run ng --image=ngnix --command --dry-run -o yaml
kubectl run --generator=deployment/apps.v1 is DEPRECATED and will be removed in a future version. Use kubectl run --generator=run-pod/v1 or kubectl create instead.
apiVersion: apps/v1
kind: Deployment
metadata:
creationTimestamp: null
labels:
run: ng
name: ng
..
status: {}
The flags are intentionally meant to create different kind of objects. I am copying from the help of kubectl run:
--restart='Always': The restart policy for this Pod. Legal values [Always,
OnFailure, Never]. If set to 'Always' a deployment is created, if set to
'OnFailure' a job is created, if set to 'Never', a regular pod is created. For
the latter two --replicas must be 1. Default 'Always', for CronJobs `Never`.
Never acts like a cronjob which is scheduled immediately.
Always creates a deployment and the deployment monitors the pod and restarts in case of failure.
By default kubectl run command creates a Deployment.
Using kubectl run command you can create and run a particular image, possibly replicated.
Creates a deployment or job to manage the created containers.
The difference in your case is seen in command (1st one) including restart policy argument.
If value of restart policy is set to 'Never', a regular pod is created. For the latter two --replicas must be 1. Default 'Always', for CronJobs Never.
Try to use command:
$ kubectl run --generator=run-pod/v1 ng --image=ngnix --command --dry-run -o yaml
instead of
$ kubectl run ng --image=ngnix --command --dry-run -o yaml
to avoid statement
"kubectl run --generator=deployment/apps.v1 is DEPRECATED and will be removed in a future version. Use kubectl run --generator=run-pod/v1 or kubectl create instead."
More information you can find here:docker-kubectl, kubectl-run.
Related
I'm running the example in chapter "Service Discovery" of the book "Kubernetes up and running". The original command to run a deployment is kubectl run alpaca-prod --image=gcr.io/kuar-demo/kuard-amd64:blue --replicas=3 --port=8080 --labels="ver=1,app=alpaca,env=prod", however in K8s version 1.25, the --replicate parameter in run command is not supported any more. I planned to run without replica and then use "kubectl scale" to scale the deployment later. Problem is the run command only creates a pod, not a deployment (the scale command expects a deployment). So how do i promote my pod to a deployment, my kubernetes verion is 1.25?
There is no way to promote it however you can change label and all those stuff but instead of that you can create the new deployment delete the existing POD.
So easy step you can take input of existing running POD to YAML file first
kubectl get pod <POD name> -o yaml > pod-spec.yaml
Create deployment spec YAML file now
kubectl create deployment deploymentname --image=imagename --dry-run=client -o yaml > deployment-spec.yaml
Edit the deployment-spec.yaml file
and in other tab pod-spec.yaml you can copy the Spec part from POD file to new deployment file.
Once deployment-spec.yaml is ready you can apply it. Make sure if you are running service labels get matched properly
kubectl apply -f deployment-spec.yaml
Delete the single running POD
kubectl delete pod <POD name>
I attempted but there is an error..i also see See 'kubectl run --help' for usage.
but i can't fix it..
kubectl run pod pod4 --image=aamirpinger/helloworld:latest --port=80 --annotaions=createdBy="Muhammad Shahbaz" --restart=Never
Error: unknown flag: --annotaions
kubectl run supports specifying annotations via the --annotations flag that can be specified multiple times to apply multiple annotations.
For example:
$ kubectl run --image myimage --annotations="foo=bar" --annotations="another=one" mypod
results in the following:
$ kubectl get pod mypod -o yaml
apiVersion: v1
kind: Pod
metadata:
annotations:
foo: bar
another: one
[...]
kubectl run doesn't have an option to set annotations.
Unless you're running a one-off debugging pod, it's usually better practice to write out the full (Deployment) YAML file, commit to source control, and install it using kubectl apply -f. That will let you specify any Kubernetes object property you need to.
As David Maze mentioned ,there is no --annotations flag for kubectl run command.It is better to write deployment yaml file compared to running using kubectl run command.
However you can add annotations to kubernetes resources using Kubectl annotate command.All Kubernetes objects support the ability to store additional data with the object as annotations.
Hope this helps.
When I use the kubectl run command instead of creating a deployment it creates a pod/selenium-node-chrome and as a result, I am unable to scale the selenium-node-chrome using the replicas command.
PS C:\Users\Test> kubectl run selenium-node-chrome --image selenium/node-chrome:latest --env="HUB_PORT_4444_TCP_ADDR=selenium-hub" --env="HUB_PORT_4444_TCP_PORT=4444"
pod/selenium-node-chrome created
PS C:\Users\Test> kubectl scale deployment selenium-node-chrome --replicas=5
Error from server (NotFound): deployments.extensions "selenium-node-chrome" not found
The video tutorial that I followed successfully created deployment "selenium-node-chrome" after running the same command. Please I need help and I am new to Kubernetes. Thanks.
You should use a generator
kubectl run selenium-node-chrome \
--image selenium/node-chrome:latest \
--env="HUB_PORT_4444_TCP_ADDR=selenium-hub" \
--env="HUB_PORT_4444_TCP_PORT=4444" \
--generator=deployment/apps.v1beta1
https://v1-17.docs.kubernetes.io/docs/reference/kubectl/conventions/#generators
All generators are deprecated in Kubernetes version 1.18. From the docs here
Note: All kubectl generators are deprecated. See the Kubernetes v1.17
documentation for a list of generators and how they were used.
You can use kubectl create deployment my-dep --image=busybox to create a deployment.
Also to create a yaml file use kubectl create deployment my-dep --image=busybox --dry-run=client -o yaml > deployment.yaml and then edit the yaml file to add env or any other details and apply the yaml via kubectl apply -f deployment.yaml
Using kubectl command line, is it possible to define the exact pod name?
I have tried with
kubectl run $pod-name --image imageX
However, the resulted pod name is something like $pod-name-xx-yyy-nnn.
So without using a yaml file, can I define the pod name using kubectl CLI?
kubectl run creates a Deployment by default. A Deployment starts a ReplicaSet that manages the pods/replicas... and therefore has a generated pod name.
Run pod
To run a single pod you can add --restart=Never to the kubectl run command.
kubectl run mypod --restart=Never --image=imageX
The $pod-name is a variable with some value let's say "hello". So when you are running the kubectl run command it will create a deployment. The deployment will create a replicaset with name "hello-xxxx" and the replicaset will create pod with name "replicasetname-xxx".
If you want to create a pod using kubectl run use the below command "kubectl run times --generator=run-pod/v1 hello --image=busybox".
It will create a pod with name hello. You are suppose to replace hello and image name.
Otherwise you could use "kubectl create pod hello --image=busybox".
I want to debug the pod in a simple way, therefore I want to start the pod without deployment.
But it will automatically create a deployment
$ kubectl run nginx --image=nginx --port=80
deployment "nginx" created
So I have to create the nginx.yaml file
---
apiVersion: v1
kind: Pod
metadata:
name: nginx
spec:
containers:
- name: nginx
image: nginx
ports:
- containerPort: 80
And create the pod like below, then it creates pod only
kubectl create -f nginx.yaml
pod "nginx" created
How can I specify in the command line the kind:Pod to avoid deployment?
// I run under minikue 0.20.0 and kubernetes 1.7.0 under Windows 7
kubectl run nginx --image=nginx --port=80 --restart=Never
--restart=Always: The restart policy for this Pod. Legal values [Always, OnFailure, Never]. If set to Always
a deployment is created, if set to OnFailure a job is created, if set to Never, a regular pod is created. For the latter two --replicas must be 1. Default Always [...]
see official document https://kubernetes.io/docs/user-guide/kubectl-conventions/#generators
Now there are two ways one can create pod through command line.
kubectl run nginx --image=nginx --restart=Never
OR
kubectl run --generator=run-pod/v1 nginx1 --image=nginx
See official documentation.
https://kubernetes.io/docs/reference/kubectl/conventions/#generators
Use generators for this, default kubectl run will create a deployment object. If you want to override this behavior use "run-pod/v1" generator.
kubectl run --generator=run-pod/v1 nginx1 --image=nginx
You may refer the link below for better understanding.
https://kubernetes.io/docs/reference/kubectl/conventions/#generators
I'm relatively new to kubernetes, but it seems it has evolved quite a bit since this question was asked. As of its latest versions(I'm running v1.16) generators are deprecated and they are completely removed in v1.18.
See the corresponding ticket and the release notes.
Release notes explicitly say:
Remove all the generators from kubectl run. It will now only create
pods.
I've tested kubectl run with various --restart flags and never got any deployments created. What we do have now is called "naked" Pod. And while you might be tempted to use it, it goes against k8s best practices:
Don’t use naked Pods (that is, Pods not bound to a ReplicaSet or
Deployment) if you can avoid it. Naked Pods will not be rescheduled in
the event of a node failure.
When you are using "kubectl run nginx --image=nginx --port=80" it creates a deployment by default.
To create a pod you have two options.
kubectl run --generator=run-pod/v1 nginx --image=nginx --port=80
kubectl create pod nginx --image=nginx --port=80