In OpenShift CLI, how can I define the memory size within the oc new-app command?
I do not think that oc new-app allows you to specify this on the command line. So your best bet would be to use oc new-app and then use oc patch to edit the created Deployment.
Alternatively, you can use oc run to start a single Pod, with that command you can actually specify --limits and --requests, but that is probably not what you want:
oc run ruby-hello-world --image=ruby-hello-world --limits=cpu=200m,memory=400Mi --requests=cpu=100m,memory=200Mi
Related
Do anyone know the oc command which do the same job when I am clicking the "Up" arrow in deployment?
I just try and find the answer:
oc scale deployments/cakephp-dw-git --replicas=1
I'm trying to run a containerized app which is stored in Nexus docker hosted on url 12.23.34.55:8086
I'm trying to run it on my Openshift Cluster, but I'm getting error. Commands I'm using to run
oc create secret docker-registry mysecret --docker-server=http://12.23.34.55/ --docker-username=aditya --docker-password=aditya --docker-email=aditya#example.org
oc secrets link default mysecret --for=pull
My nexus is running on http://12.23.34.55:8081
Now I'm using command to launch in OpenShift using below command.
oc new-app 12.23.34.55:8085/mytestapp:11 --insecure-registry=true
as per $ oc new-app myregistry:5000/example/myimage
https://docs.openshift.com/container-platform/4.1/applications/application_life_cycle_management/creating-new-applications.html
But it does not work, it asks for password and not able to deploy from console too, can anyone help me with exact commmand.
Creating the secret is not enough for OpenShift to be able to pull from the registry. You still need to link that secret as well.
Take a look at the official documentation here:
https://docs.openshift.com/container-platform/4.1/openshift_images/managing_images/using-image-pull-secrets.html#images-allow-pods-to-reference-images-from-secure-registries_using-image-pull-secrets
Okay! I found an answer, so using private registry first we should import image using
oc import name url/imagename:tag
then we can create new app with the same
oc new app name
I would like to add some labels (commit hash, branch name,...) to images I create using Openshift source-to-image binary build. These labels will have naturally different values for every build.
Currently oc start-build does not even support -e flags to add environment variables. (At least is seems to, it works for Git source, its a bug?)
And for binary build does not supports --build-arg to pass argument for docker file.
The only way I was able to accomplish this to call oc set env bc [build-name] then start the build. And use Label in Dockerfile with values from environment variables.
My question is isn't there a better way to do this? (Ideally in a way that Dockerfile is not necessarily changed) Doesn't s2i supports passing --label to docker build behind?
Thank you.
Do you want to add environment variable when you start oc start-build ? Because you mentioned oc set env bc [build-name].
Then you can use --env=<key>=<value> option, refer Starting Build for more details.
$ oc start-build <buildconfig_name> --env=<key>=<value>
I hope it help you.
We are running OSE 3.2 and I'm trying to rsh with oc to various pods. I'm using Cygwin. As long as I pass it a command, it works, so I assume it's unable to give me a shell. I've tried setting my TERM environment variable to vt100, xterm, and ansi with no luck. I am able to rsh into pods with oc using the Windows cmd prompt with TERM not set at all, but I really don't like that thing and would prefer to use Cygwin for all functions. I've searched quite a bit for a solution to this, but have come up empty. Thanks much.
On my ubuntu laptop I was issuing some kubectl commands including running kubernetes from a local Docker container all was well ... at some point I then issued this command
kubectl config set-cluster test-doc --server=https://104.196.108.118
now my local kubectl fails to execute ... looks like the Server side needs to get reset back to default
kubectl version
Client Version: version.Info{Major:"1", Minor:"2", GitVersion:"v1.2.2", GitCommit:"528f879e7d3790ea4287687ef0ab3f2a01cc2718", GitTreeState:"clean"}
error: couldn't get version/kind; json parse error: json: cannot unmarshal string into Go value of type struct { APIVersion string "json:\"apiVersion,omitempty\""; Kind string "json:\"kind,omitempty\"" }
I deleted and reinstalled the gcloud SDK binaries and ran
mv ~/.config/gcloud ~/.config/gcloud~ignore
gcloud init
gcloud components update kubectl
How do I delete my local kubectl settings (on ubuntu 16.04) and start fresh ?
It's important to note that you've set a kubeconfig setting for your client. When you run kubectl version, you're getting the version for client and the server which in your case seems to be the issue with the version command.
Updating your config
You need to update the setting to the appropriate information. You can use the same command you used to set the server to change it to the correct server.
If you want to wipe the slate clean in terms of client config, you should remove the kubeconfig file(s). In my experience with the gcloud setup, this is just ~/.kube/config.
If you are running the cluster through google cloud engine, you can use gcloud to get the kubeconfig set for you as per the container engine quick start guide. The following assumes that you have defaults for the project, zone, and cluster set.
gcloud container clusters get-credentials CLUSTER_NAME
Removing kubectl - this isn't necessary
If your goal is to wholesale get rid of kubectl, you should remove the component rather than reseting gcloud.
gcloud components remove kubectl
But that won't solve your problem as it doesn't remove or reset ~/.kube/config when I run it on Mac and if you want to keep working with it, you'll need to reinstall kubectl.