Kubernetes¶

Deploy your dotflow pipeline to a Kubernetes cluster.

Create project¶

dotflow init
# Select cloud: kubernetes

Or generate files for an existing project:

dotflow cloud generate --platform kubernetes

Generated files¶

File	Description
`Dockerfile`	Container image
`deployment.yaml`	Kubernetes Deployment manifest
`service.yaml`	Kubernetes Service manifest

Prerequisites¶

Docker
kubectl
A Kubernetes cluster (minikube for local testing)

Deploy (local with minikube)¶

brew install minikube
minikube start
eval $(minikube docker-env)

docker build -t my_pipeline:latest .

kubectl apply -f deployment.yaml
kubectl apply -f service.yaml

Deploy (production)¶

Replace image in deployment.yaml with your container registry URL and change imagePullPolicy to Always:

kubectl apply -f deployment.yaml
kubectl apply -f service.yaml

View logs¶

kubectl logs -l app=my_pipeline

Important¶

Do not rename workflow.py or the main() function — the Dockerfile CMD depends on it
imagePullPolicy: Never is for local images only — change to Always for production
Project names with _ are automatically converted to - in Kubernetes resource names