scanner-research / kubequeue

Copy/paste framework for a simple job distributed job queue with Kubernetes

kubequeue

0. Install kubectl
1. Add your function to tasks.py, see add_one for example
2. Configure cluster-start.sh parameters (including CLUSTER_ID)
3. Configure IMAGE_NAME in image-build.sh and worker.yaml
4. Add dependencies to Dockerfile
5. Run ./image-build.sh
6. Run ./cluster-start.sh
7. Run ./kube-start.sh
8. Change the TASK and ARGS parameters to your input task/arguments in submit.py.
9. Replace print(outputs) with whatever you want to do with your outputs.
10. Run python3 submit.py and wait!