Follow the instructions on Red Hat Developer Sandbox web page to create an account. This will allow you to start a cloud development environment running this repository.
Click the badge 
| CDE Startup | CDE Running VS Code |
|---|---|
 |
 |
The cloud development environment that you started runs in a container with all the pre-requisties to run this demo.
The following commands create the DevWorkspacesTemplates that contain the IDEs dfinition: vim, VisualStudio Code and IntelliJ.
kubectl apply -f ./editors-contributions/Create a cloud development environment using dw1.yaml that specify a git repo and a container with the devtools but no IDE:
kind: DevWorkspace
apiVersion: workspace.devfile.io/v1alpha2
metadata:
name: dw
spec:
started: true
template:
projects:
- name: quarkus-api
git:
remotes:
origin: https://github.com/l0rd/quarkus-api
components:
- name: dev-tooling
container:
image: quay.io/mloriedo/devtools:latest
EOFTo start the cloud development environment, open a Terminal from VS Code (Terminal -> New Terminal) and apply the DevWorkspace custom resource using kubectl:
# Create the CDE
kubectl apply -f dw1.yaml
# Wait for the workspace to be ready
kubectl wait --for=condition=Ready dw/dw --timeout=300s# Get the cloud development environment Pod name
export POD=`kubectl get pod -l "controller.devfile.io/devworkspace_name=dw" -o jsonpath={.items[].metadata.name}`
# Run the project build (`mvn clen install`) using kubectl exec
kubectl exec -ti ${POD} -- bash -c 'cd ${PROJECT_SOURCE} && mvn clean install'exit code 137: the build process has been killed with an OOM error (Out Of Memory). We are going to fix it in next step
We are going to use a slightly different DevWorkspace object:
container:
image: quay.io/mloriedo/devtools:latest
> memoryRequest: 2G
> memoryLimit: 10G
> cpuRequest: '1'
> cpuLimit: '4'
> contributions:
> - name: editor
> kubernetes:
> name: vscodeIn the same terminal apply the new DevWorkspace custom resource using kubectl:
# Update the CDE
kubectl apply -f dw2.yaml
# Wait for the workspace to be ready
kubectl wait --for=condition=Ready dw/dw --timeout=300s
# Retrieve the IDE URL
export IDE=`kubectl get dw dw -o jsonpath={.status.mainUrl}` && \
printf "\nOpen VS Code in your browser with the following link:\n\n\t${IDE}\n\n"Patch the DevWorkspace object to change IDE. Let's use vim:
kubectl patch dw dw \
--type merge \
-p '{ "spec": { "contributions": [ {"name": "editor", "kubernetes": { "name": "vim" } } ] } }' && \
sleep 2 && \
kubectl wait --for=condition=Ready dw/dw --timeout=300s && \
export IDE=`kubectl get dw dw -o jsonpath={.status.mainUrl}` && \
printf "\nOpen VIM in your browser with the following link:\n\n\t${IDE}\n\n"Or IntelliJ...
kubectl patch dw dw \
--type merge \
-p '{ "spec": { "contributions": [ {"name": "editor", "kubernetes": { "name": "intellij" } } ] } }' && \
sleep 2 && \
kubectl wait --for=condition=Ready dw/dw --timeout=300s && \
export IDE=`kubectl get dw dw -o jsonpath={.status.mainUrl}` && \
printf "\nOpen IntelliJ in your browser with the following link:\n\n\t${IDE}\n\n" Next DevWorkspace object specifies some commands 1-package and 2-runtests:
cpuRequest: '1'
cpuLimit: '4'
> commands:
> - id: 1-package
> exec:
> component: dev-tooling
> commandLine: './mvnw clean package -DskipTests=true -Dquarkus.profile=staging'
> - id: 2-runtests
> exec:
> component: dev-tooling
> commandLine: './mvnw test'
contributions:
- name: editorUse kubectl to apply those changes
# Update the CDE
kubectl apply -f dw3.yaml
# Restart the CDE to make sure the
# new commands are passed to VS Code
scripts/dw-restart.sh
# Wait for the workspace to be ready
kubectl wait --for=condition=Ready dw/dw --timeout=300sNow it's possible to run the predefined commands as VS Code tasks (Terminal -> Run Tasks -> devfile)
Next DevWorkspace object specifies a new component of type container named postgresql:
> - name: postgresql
> container:
> image: 'quay.io/centos7/postgresql-13-centos7@sha256:994f5c622e2913bda1c4a7fa3b0c7e7f75e7caa3ac66ff1ed70ccfe65c40dd75'
> env:
> - name: POSTGRESQL_USER
> value: user
> - name: POSTGRESQL_PASSWORD
> value: password
> - name: POSTGRESQL_DATABASE
> value: food_db
> - name: PGDATA
> value: /tmp/pgdataA PosgreSQL container will be run as a sidecar of the dev tools container in the cloud development environment Pod.
As for other DevWorkspace objects, let's use kubectl to apply it:
# Update the CDE
kubectl apply -f dw4.yaml
# Wait for the workspace to be ready
kubectl wait --for=condition=Ready dw/dw --timeout=300sNow that a postgres container is available, the task 2-runtests is executed successfully.
The following extra configurations are optional and are done automatically when using Dev Spaces or Eclipse Che but need to be done manually when applying DevWorkspaces objects manually.
The following command will create a secret named kube-config that will be automatically mounted in /home/user/.kube/config.
$ ./auto-mount/kubeconfig.shThe following command creates the dotfiles-config ConfigMap and the data of the ConfigMap will be automatically mounted as the file /home/user/.bashrc.d/ps1.sh. This will change the prompt (PS1) of CDE terminal.
$ kubectl apply -f ./auto-mount/bashrc.yaml The following command creates the git-credentials Secret that will be mounted as a git credentials file in the CDE. The user won't need to manually login to GitHub to git push.
export GITHUB_LOGIN="<gh-login>"
export GITHUB_TOKEN="<gh-token>"
export GIT_USER_EMAIL="<email>"
export GIT_USER_NAME="<name>"
$ ./auto-mount/gitconfig.shTODO
Build the container image with the devtools defined in the Dockerfile.dev file
IMAGE="quay.io/mloriedo/devtools"
docker build -t "${IMAGE}" -f Dockerfile.dev .
docker push "${IMAGE}"kubectl delete dw dw
kubectl delete cm git-config dotfiles-config
kubectl delete secret git-credentials kube-config
kubectl delete dwt vscode vscode-ds intellij vim