Layer-Scheduler

Introduction

Layer-aware and Resource-adaptive container Scheduler (LRScheduler) utilizes container image layer information to design and implement a node scoring and container scheduling mechanism, which can effectively reduce download cost when deploying containers and its weight dynamically adapts to enhance load balancing in edge clusters, increasing layer sharing scores when resource load is low to use idle resources effectively.

Quick Start

1. Make sure the runtime of Kubernetes is Docker

2. Please upload all the local images and images that will be deployed in containers to Docker registry

3. Log in all nodes to the image repository you are using

4. Set the site, account and password of your docker registry in the file named 'main.go'

5. Create a scheduler-config.yaml file and specify the use of the layer-scheduler (LRScheduler) for scheduling

6. Everytime you edit the files in the folder called 'layer', you are supposed to compile ‘main.go’ again by: go build -o kube-scheduler ./main.go

7. Start layer-scheduler (LRScheduler) which helps you get the logs of images, layers and score by: ./kube-scheduler --authentication-kubeconfig=/etc/kubernetes/scheduler.conf --authorization-kubeconfig=/etc/kubernetes/scheduler.conf --config=/etc/kubernetes/scheduler-config.yaml

8. Deploy a Pod and specify in the yaml file that it should use the layer-scheduler (LRScheduler) for scheduling

Prerequisites

1. OS-Image：CentOS Linux 7 (Core)

2. k8s version：v1.23.8

3. Docker version 20.10.8

4. go version go1.18 linux/amd64

5. Kubelet Version：1.23.8

6. Client Version: v1.23.8

7. Server Version: v1.23.8

8. Container Runtime Version: docker://20.10.8