2 views collected around this technical thread.
This article explains the Linux kernel's CPU load‑balancing mechanism, covering concepts such as CPU load, balancing strategies, scheduling domains, groups, the PELT algorithm, CFS scheduling, trigger points, and practical code examples for both regular and real‑time tasks.
This article explains the Linux Completely Fair Scheduler (CFS), covering its design goals, core concepts such as virtual runtime, weight, red‑black tree management, load‑balancing mechanisms, scheduling policies, and answers common questions about its operation and kernel code.
This article explains Linux's CFS scheduler mechanics, including the impact of priority, virtual runtime, event‑driven scheduling, the differences between PELT and WALT load‑tracking algorithms, and the complexities introduced by big‑LITTLE architectures on task placement, throughput, and latency.
This article provides a comprehensive overview of the Linux process scheduler, explaining why scheduling is needed, how it works, the various scheduling policies such as real‑time and CFS, the mechanisms for priority and load balancing, and the historical evolution of Linux schedulers with code examples.
This article explains how Kubernetes resource requests and limits map to Linux cgroup settings via the CFS scheduler, illustrates the underlying calculations for cpu.shares and cpu.cfs_quota_us, and discusses the impact on programming languages such as Go and Java within containers.
This article explains the concepts, reasons, and implementation details of Linux process scheduling, covering cooperative and preemptive scheduling, run‑queue structures, wake‑up paths, scheduling ticks, context switching, priority handling, and the various scheduling classes used in the kernel.
An in‑depth overview of Linux’s Completely Fair Scheduler (CFS) explains its design goals, data structures such as red‑black trees and virtual runtime, the evolution from O(n) and O(1) schedulers, weight‑based fairness calculations, scheduling classes, and key kernel functions that implement task selection and timing.
The talk from DataFun Summit 2023 explains how Baidu's CFS storage builds a trillion‑file‑scale distributed file system by revisiting file system fundamentals, POSIX limitations, historical storage architectures, and introducing a lock‑free metadata service with single‑shard primitives, data‑layout optimizations, and a simplified client‑centric architecture that achieves high scalability and performance.
This article analyzes the limitations of the Linux Completely Fair Scheduler (CFS) for high‑priority workloads, introduces Tencent's custom offline BT scheduler that provides absolute preemption, and presents experimental results showing significant improvements in latency, CPU utilization, and carbon‑reduction for cloud‑native services.
The article explains how the Linux CFS scheduler balances load using three mechanisms—periodic balancer on busy CPUs, NOHZ idle balancer that wakes idle CPUs in tickless mode, and the new idle balancer that checks overload and cache state—detailing their triggers, IPI interactions, timing intervals, and key data structures.
PELT (Per‑Entity Load Tracking) replaces the coarse per‑run‑queue load model in Linux’s CFS scheduler with fine‑grained, decay‑based load and utility measurements for each scheduling entity, enabling more accurate load‑balancing, CPU‑frequency decisions, and hierarchical group scheduling across heterogeneous cores.
During Kuaishou's 2020 Spring Festival red‑envelope campaign, Tencent Cloud's Cloud File Storage (CFS) provided a high‑throughput, read‑heavy NFS solution that handled up to 5 GB/s across 1 700 pods, delivering 100% availability, low latency, and massive interaction volumes for billions of users.
This article surveys the evolution, core concepts, and concrete implementation details of Linux task scheduling, illustrating kernel data structures, scheduling entities, run‑queue management, and the CFS algorithm with code excerpts and explaining how the scheduler balances interactive and background workloads.
This article explains the concepts of nice and priority values in Linux, how they differ, their impact on process scheduling, and compares historic O(1) and modern CFS schedulers, including real‑time scheduling policies, providing practical command examples for administrators.