Skip to content

Latest commit

 

History

History
108 lines (66 loc) · 2.42 KB

running-locally.md

File metadata and controls

108 lines (66 loc) · 2.42 KB
Raw

Running Locally

Pre-requisites:

  • Golang
  • Yarn. brew install yarn
  • Docker
  • Kustomize
  • protoc brew install protobuf
  • jq
  • Kubernetes Cluster (we recommend Docker for Desktop + K3D, as this will allow you to test RBAC set-up, and is also fast)

Useful:

  • For a PS1 prompt showing your current kube context: kube-ps1 to help. brew install kube-ps1

K3D tip: You can set-up K3D to be part of your default kube config as follows

cp ~/.kube/config ~/.kube/config.bak
cat $(k3d get-kubeconfig --name='k3s-default') >> ~/.kube/config

Add to /etc/hosts:

127.0.0.1 minio
127.0.0.1 postgres
127.0.0.1 mysql

To install into the “argo” namespace of your cluster: Argo, MinIO (for saving artifacts and logs) and Postgres (for offloading or archiving):

make start

If you prefer MySQL:

make start DB=mysql

You’ll now have

Either:

You need the token to access the CLI or UI:

eval $(make env)

./dist/argo auth token

At this point you’ll have everything you need to use the CLI and UI.

User Interface

Tip: If you want to make UI changes without a time-consuming build:

cd ui
yarn install
yarn start

The UI will start up on http://localhost:8080.

Debugging

If you want to run controller or argo-server in your IDE (e.g. so you can debug it):

Start with only components you don't want to debug;

make start COMPONENTS=controller

Or

make start COMPONENTS=argo-server

To find the command arguments you need to use, you’ll have to look at the start target in the Makefile.`

Running Sonar Locally

This can only be done if you have already created a pull request.

Install the scanner:

brew install sonar-scanner

Run the tests:

make test CI=true
make test-reports/test-report.out

Perform a scan:

# the key is PR number (e.g. "2666"), the branch is the CI branch, e.g. "pull/2666"
SONAR_TOKEN=... sonar-scanner -Dsonar.pullrequest.key=... -Dsonar.pullrequest.branch=...

Clean

To clean-up everything:

make clean
kubectl delete ns argo
docker system prune -af