OpenShift Console

Codename: "Bridge"

The console is a more friendly kubectl in the form of a single page webapp. It also integrates with other services like monitoring, chargeback, and OLM. Some things that go on behind the scenes include:

Proxying the Kubernetes API under /api/kubernetes
Providing additional non-Kubernetes APIs for interacting with the cluster
Serving all frontend static assets
User Authentication

Quickstart

Dependencies:

node.js >= 18 & yarn >= 1.20
go >= 1.18+
oc or kubectl and an OpenShift or Kubernetes cluster
jq (for contrib/environment.sh)
Google Chrome/Chromium or Firefox for integration tests

Build everything:

This project uses Go modules, so you should clone the project outside of your GOPATH. To build both the frontend and backend, run:

./build.sh

Backend binaries are output to ./bin.

Configure the application

The following instructions assume you have an existing cluster you can connect to. OpenShift 4.x clusters can be installed using the OpenShift Installer. More information about installing OpenShift can be found at https://try.openshift.com/. You can also use CodeReady Containers for local installs, or native Kubernetes clusters.

OpenShift (no authentication)

For local development, you can disable OAuth and run bridge with an OpenShift user's access token. If you've installed OpenShift 4.0, run the following commands to login as the kubeadmin user and start a local console for development. Make sure to replace /path/to/install-dir with the directory you used to install OpenShift.

oc login -u kubeadmin -p $(cat /path/to/install-dir/auth/kubeadmin-password)
source ./contrib/oc-environment.sh
./bin/bridge

The console will be running at localhost:9000.

If you don't have kubeadmin access, you can use any user's API token, although you will be limited to that user's access and might not be able to run the full integration test suite.

OpenShift (with authentication)

If you need to work on the backend code for authentication or you need to test different users, you can set up authentication in your development environment. Registering an OpenShift OAuth client requires administrative privileges for the entire cluster, not just a local project. You must be logged in as a cluster admin such as system:admin or kubeadmin.

To run bridge locally connected to an OpenShift cluster, create an OAuthClient resource with a generated secret and read that secret:

oc process -f examples/console-oauth-client.yaml | oc apply -f -
oc get oauthclient console-oauth-client -o jsonpath='{.secret}' > examples/console-client-secret

If the CA bundle of the OpenShift API server is unavailable, fetch the CA certificates from a service account secret. Otherwise copy the CA bundle to examples/ca.crt:

oc get secrets -n default --field-selector type=kubernetes.io/service-account-token -o json | \
    jq '.items[0].data."ca.crt"' -r | python -m base64 -d > examples/ca.crt
# Note: use "openssl base64" because the "base64" tool is different between mac and linux

Finally run the console and visit localhost:9000:

./examples/run-bridge.sh

Enabling Monitoring Locally

In order to enable the monitoring UI and see the "Observe" navigation item while running locally, you'll need to run the OpenShift Monitoring dynamic plugin alongside Bridge. To do so, follow these steps:

Clone the monitoring-plugin repo: https://github.com/openshift/monitoring-plugin
cd to the monitoring-plugin root dir
Run

yarn && yarn start

Run Bridge in another terminal following the steps above, but set the following environment variable before starting Bridge:

export BRIDGE_PLUGINS="monitoring-plugin=http://localhost:9001"

Updating `tectonic-console-builder` image

Updating tectonic-console-builder image is needed whenever there is a change in the build-time dependencies and/or go versions.

In order to update the tectonic-console-builder to a new version i.e. v27, follow these steps:

Update the tectonic-console-builder image tag in files listed below:
- .ci-operator.yaml
- Dockerfile.dev
- Dockerfile.plugins.demo For example, tectonic-console-builder:27
Update the dependencies in Dockerfile.builder file i.e. v18.0.0.
Run ./push-builder.sh script build and push the updated builder image to quay.io. Note: You can test the image using ./builder-run.sh ./build-backend.sh. To update the image on quay.io, you need edit permission to the quay.io/coreos/ tectonic-console-builder repo.
Lastly, update the mapping of tectonic-console-builder image tag in [openshift/release](https:// github.com/openshift/release/blob/master/core-services/image-mirroring/supplemental-ci-images/mapping_supplemental_ci_images_ci) repository. Note: There could be scenario were you would have to add the new image reference in the "mapping_supplemental_ci_images_ci" file, i.e. to avoid CI downtime for upcoming release cycle. Optional: Request for the rhel-8-base-nodejs-openshift-4.15 nodebuilder update if it doesn't match the node version in tectonic-console-builder.

CodeReady Containers

If you want to use CodeReady for local development, first make sure it is set up, and the OpenShift cluster is started.

To login to the cluster's API server, you can use the following command:

oc login -u kubeadmin -p $(cat ~/.crc/machines/crc/kubeadmin-password) https://api.crc.testing:6443

… or, alternatively, use the CRC daemon UI (Copy OC Login Command --> kubeadmin) to get the cluster-specific command.

Finally, prepare the environment, and run the console:

source ./contrib/environment.sh
./bin/bridge

Native Kubernetes

If you have a working kubectl on your path, you can run the application with:

export KUBECONFIG=/path/to/kubeconfig
source ./contrib/environment.sh
./bin/bridge

The script in contrib/environment.sh sets sensible defaults in the environment, and uses kubectl to query your cluster for endpoint and authentication information.

To configure the application to run by hand, (or if environment.sh doesn't work for some reason) you can manually provide a Kubernetes bearer token with the following steps.

First get the secret ID that has a type of kubernetes.io/service-account-token by running:

kubectl get secrets

then get the secret contents:

kubectl describe secrets/<secret-id-obtained-previously>

Use this token value to set the BRIDGE_K8S_AUTH_BEARER_TOKEN environment variable when running Bridge.

Operator

In OpenShift 4.x, the console is installed and managed by the console operator.

Hacking

See CONTRIBUTING for workflow & convention details.

See STYLEGUIDE for file format and coding style guide.

Dev Dependencies

go 1.18+, nodejs/yarn, kubectl

Frontend Development

All frontend code lives in the frontend/ directory. The frontend uses node, yarn, and webpack to compile dependencies into self contained bundles which are loaded dynamically at run time in the browser. These bundles are not committed to git. Tasks are defined in package.json in the scripts section and are aliased to yarn run <cmd> (in the frontend directory).

Install Dependencies

To install the build tools and dependencies:

cd frontend
yarn install

You must run this command once, and every time the dependencies change. node_modules are not committed to git.

Interactive Development

The following build task will watch the source code for changes and compile automatically. If you would like to disable hot reloading, set the environment variable HOT_RELOAD to false.

yarn run dev

If changes aren't detected, you might need to increase fs.inotify.max_user_watches. See https://webpack.js.org/configuration/watch/#not-enough-watchers. If you need to increase your watchers, it's common to see multiple errors beginning with Error from chokidar.

Unit Tests

Run all unit tests:

./test.sh

Run backend tests:

./test-backend.sh

Run frontend tests:

./test-frontend.sh

Debugging Unit Tests

cd frontend; yarn run build
Add debugger; statements to any unit test
yarn debug-test route-pages
Chrome browser URL: 'chrome://inspect/#devices', click on the 'inspect' link in Target (v10...) section.
Launches chrome-dev tools, click Resume button to continue
Will break on any debugger; statements

Integration Tests

Cypress integration tests are implemented in Cypress.io.

To install Cypress:

cd frontend
yarn run cypress install

Launch Cypress test runner:

cd frontend
oc login ...
yarn run test-cypress-console

This will launch the Cypress Test Runner UI in the console package, where you can run one or all Cypress tests.

Important: when testing with authentication, set BRIDGE_KUBEADMIN_PASSWORD environment variable in your shell.

Execute Cypress in different packages

An alternate way to execute cypress tests is via test-cypress.sh which takes a -p <package> parameter to allow execution in different packages. It also can run Cypress tests in the Test Runner UI or in -- headless mode:

console>./test-cypress.sh
Runs Cypress tests in Test Runner or headless mode
Usage: test-cypress [-p] <package> [-s] <filemask> [-h true]
  '-p <package>' may be 'console, 'olm' or 'devconsole'
  '-s <specmask>' is a file mask for spec test files, such as 'tests/monitoring/*'. Used only in headless mode when '-p' is specified.
  '-h true' runs Cypress in headless mode. When omitted, launches Cypress Test Runner
Examples:
  test-cypress.sh                                       // displays this help text
  test-cypress.sh -p console                            // opens Cypress Test Runner for console tests
  test-cypress.sh -p olm                                // opens Cypress Test Runner for OLM tests
  test-cypress.sh -h true                               // runs all packages in headless mode
  test-cypress.sh -p olm -h true                        // runs OLM tests in headless mode
  test-cypress.sh -p console -s 'tests/crud/*' -h true  // runs console CRUD tests in headless mode

When running in headless mode, Cypress will test using its integrated Electron browser, but if you want to use Chrome or Firefox instead, set BRIDGE_E2E_BROWSER_NAME environment variable in your shell with the value chrome or firefox.

More information on Console's Cypress usage

More information on DevConsole's Cypress usage

How the Integration Tests Run in CI

The end-to-end tests run against pull requests using ci-operator. The tests are defined in this manifest in the openshift/release repo and were generated with ci-operator-prowgen.

CI runs the test-prow-e2e.sh script, which runs test-cypress.sh.

test-cypress.sh runs all Cypress tests, in all 'packages' (console, olm, and devconsole), in -- headless mode via:

test-cypress.sh -h true

For more information on test-cypress.sh usage please see Execute Cypress in different packages

Internationalization

See INTERNATIONALIZATION for information on our internationalization tools and guidelines.

Deploying a Custom Image to an OpenShift Cluster

Once you have made changes locally, these instructions will allow you to push changes to an OpenShift cluster for others to review. This involves building a local image, pushing the image to an image registry, then updating the OpenShift cluster to pull the new image.

Prerequisites

Docker v17.05 or higher for multi-stage builds
An image registry like quay.io or Docker Hub

Steps

Create a repository in the image registry of your choice to hold the image.
Build Image docker build -t <your-image-name> <path-to-repository | url>. For example:

docker build -t quay.io/myaccount/console:latest .

Push image to image registry docker push <your-image-name>. Make sure docker is logged into your image registry! For example:

docker push quay.io/myaccount/console:latest

Put the console operator in unmanaged state:

oc patch consoles.operator.openshift.io cluster --patch '{ "spec": { "managementState": "Unmanaged" } }' --type=merge

Update the console Deployment with the new image:

oc set image deploy console console=quay.io/myaccount/console:latest -n openshift-console

Wait for the changes to rollout:

oc rollout status -w deploy/console -n openshift-console

You should now be able to see your development changes on the remote OpenShift cluster!

When done, you can put the console operator back in a managed state to remove the custom image:

oc patch consoles.operator.openshift.io cluster --patch '{ "spec": { "managementState": "Managed" } }' --type=merge

Dependency Management

Dependencies should be pinned to an exact semver, sha, or git tag (eg, no ^).

Backend

Whenever making vendor changes:

Finish updating dependencies & writing changes
Commit everything except vendor/ (eg, server: add x feature)
Make a second commit with only vendor/ (eg, vendor: revendor)

Adding new or updating existing backend dependencies:

Edit the go.mod file to the desired version (most likely a git hash)
Run go mod tidy && go mod vendor
Verify update was successful. go.sum will have been updated to reflect the changes to go.mod and the package will have been updated in vendor.

Frontend

Add new frontend dependencies:

yarn add <package@version>

Update existing frontend dependencies:

yarn upgrade <package@version>

To upgrade yarn itself, download a new yarn release from https://github.com/yarnpkg/yarn/releases, replace the release in frontend/.yarn/releases with the new version, and update yarn-path in frontend/.yarnrc.

@patternfly

Note that when upgrading @patternfly packages, we've seen in the past that it can cause the JavaScript heap to run out of memory, or the bundle being too large if multiple versions of the same @patternfly package is pulled in. To increase efficiency, run the following after updating packages:

npx yarn-deduplicate --scopes @patternfly

Supported Browsers

We support the latest versions of the following browsers:

Edge
Chrome
Safari
Firefox

IE 11 and earlier is not supported.

Frontend Packages

console-dynamic-plugin-sdk [API] [Console Extensions]
console-plugin-shared
dev-console
eslint-plugin-console
integration-tests-cypress
knative-plugin
operator-lifecycle-manager [Descriptors README] [Descriptors API Reference]
rhoas-plugin

AObuchow/console