I've been using this project to test FLAME on minikube, to vet the k8s capabilities before trying anything in production. Here's my general workflow:
- First, clone this repo and install minikube
minikube startto fire up the clustereval $(minikube docker-env)to use minikube's docker repocdinto flame_test, if you're not already theredocker build -t flame_test:latest .to build the imagekubectl apply -f deployment.yamlto start up the k8s resources- Get a shell into one of the pods in the cluster (I included a deployment of testing pods in
deployment.yaml), e.g.kubectl exec -n flame-namespace --stdin --tty iex-debug-deployment-7bfbfbc76f-5jkpm -- /bin/bash - POST a request to one of the
flame-testpods, e.g.curl -X POST http://10.244.0.81:4000/call -d '{"data": 42}' -H "Content-Type: application/json" - You'll get a response of
{message: ${data * 2}}in your console! If you check the logs of whichever pod you sent the request to, you'll see a log line that tells you the node that actually processed your request! e.g.Processed task on node :"[email protected]"