https://www.chenshaowen.com/en/tags/scheduling/Recent content in Scheduling on Shaowen Chen's WebsiteHugo -- gohugo.ioen©2016 - {year}, All Rights Reserved.Sun, 24 Mar 2024 00:00:00 +0000weeklyhttps://www.chenshaowen.com/en/blog/the-basic-of-volcano.htmlSun, 24 Mar 2024 00:00:00 +0000Sun, 24 Mar 2024 00:00:00 +0000https://www.chenshaowen.com/en/blog/the-basic-of-volcano.html1. Introduction to Volcano Volcano is a Kubernetes-based resource scheduling system open-sourced by Huawei. Compared with the native scheduler, its notable features are: Supports gang scheduling Scheduling batch jobs easily runs into deadlock: for example, two jobs each need 10 Pods running simultaneously to start. When both jobs are submitted at the same time, it is possible that only part of the Pods of each are scheduled, so neither job can run properly and they wait on each other.微信公众号VolcanoAIKubernetesSchedulingOperationshttps://www.chenshaowen.com/en/blog/custom-cluster-scheduler-to-optimize-tekton.htmlWed, 26 Apr 2023 10:00:00 +0000Wed, 26 Apr 2023 10:00:00 +0000https://www.chenshaowen.com/en/blog/custom-cluster-scheduler-to-optimize-tekton.html1. A Constrained Build Environment Cannot Meet Build Requirements Tekton is a CICD engine based on a Kubernetes cluster, and it is more cloud-native than Jenkins. In plain terms, that means it is easier to develop plugins for, easier to scale, easier to observe, and more fun. Because code can only be stored on the company intranet, the build cluster can only be deployed on the office intranet.微信公众号OptimizationTektonCICDDevOpsKubernetesOperationsLearningSchedulingBuildMulti-Clusterhttps://www.chenshaowen.com/en/blog/how-to-set-up-pod-to-run-to-a-specified-node.htmlMon, 27 Jun 2022 00:00:00 +0000Mon, 27 Jun 2022 00:00:00 +0000https://www.chenshaowen.com/en/blog/how-to-set-up-pod-to-run-to-a-specified-node.html1. Specify a Node via nodeSelector When Creating a Workload Add a label to the node 1 kubectl label node node2 project=A Create the workload with the specified nodeSelector 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 cat <<EOF | kubectl apply -f - apiVersion: apps/v1 kind: Deployment metadata: name: nginx-nodeselector spec: replicas: 1 selector: matchLabels: app: nginx-nodeselector template: metadata: labels: app: nginx-nodeselector spec: nodeSelector: project: A containers: - name: nginx image: nginx EOF Check the workload 1 2 3 4 kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES nginx-nodeselector-7bb75b7687-7r5xk 1/1 Running 0 19s 10.微信公众号PodKubernetesSchedulingOperationsnodeSelectorLearninghttps://www.chenshaowen.com/en/blog/descheduler-makes-kubernetes-load-more-balanced.htmlSat, 11 Jun 2022 00:00:00 +0000Sat, 11 Jun 2022 00:00:00 +0000https://www.chenshaowen.com/en/blog/descheduler-makes-kubernetes-load-more-balanced.html1. Why Secondary Scheduling Is Needed The job of the Kubernetes scheduler is to bind a Pod to a single best node. To do this, the scheduler runs through a series of filtering and scoring steps. Kubernetes scheduling is based on Requests, but the actual usage of each Pod changes dynamically.微信公众号KubernetesdeschedulerSchedulingOperationshttps://www.chenshaowen.com/en/blog/affinity-of-kubernetes-scheduler.htmlSat, 30 May 2020 00:00:00 +0000Sat, 30 May 2020 00:00:00 +0000https://www.chenshaowen.com/en/blog/affinity-of-kubernetes-scheduler.html1. The Scheduler in Kubernetes kube-scheduler is the component in Kubernetes that decides which Node a Pending Pod runs on; it is called the scheduler. Kubernetes ships with a large number of built-in scheduling policies and also provides some advanced scheduling policies (nodeAffinity, podAffinity, and so on) for users, which微信公众号KubernetesSchedulingAffinityLearningOperations