0 views collected around this technical thread.
This article provides a step‑by‑step guide to building a custom Kubernetes controller in Go that uses Prometheus metrics to intelligently schedule Flink TaskManager Pods, covering the underlying scheduler concepts, code implementation, Docker image creation, RBAC setup, deployment, testing, and advanced considerations.
This article explains the role of kube-scheduler in Kubernetes, details its scheduling workflow and plugin framework, and demonstrates how to define a custom scheduler by configuring KubeSchedulerConfiguration, creating a ConfigMap, deploying it via a Deployment, and verifying its operation alongside the default scheduler.
The Kubernetes scheduler continuously watches for unscheduled Pods, places them in a priority queue, filters feasible Nodes, scores and selects the best Node using built‑in and custom plugins—adjusting the node‑sampling rate for large clusters, and allowing extensibility through extenders, multiple schedulers, and the scheduler framework configuration.
The article describes how 58 Cloud Search (WCS) was built as a fully self‑service search platform on Kubernetes, detailing its background, architecture, high‑availability master design with HAProxy and Keepalived, multi‑cluster deployment, automated indexing, fault recovery, resource‑aware scheduling, and operational optimizations that now serve over 200 business search services with billions of daily queries.