This article explains the fundamentals of Linux CPU affinity, compares soft and hard binding, shows how to use taskset, sched_setaffinity, numactl and cgroup cpuset, and provides real‑world case studies for servers, game frameworks, Nginx and Hadoop clusters.
CPU caches bridge the speed gap between processors and memory, but cache misses can dramatically degrade performance, especially under high concurrency or big‑data workloads; this article explains cache architecture, common miss causes, diagnostic tools like perf and cachestat, and practical optimization techniques for Linux systems.
When a production server hit 100% CPU at 3 AM, the author walks through a three‑minute, step‑by‑step method—quickly identifying the offending process, drilling into threads, and pinpointing problematic code—while sharing useful shell commands, common pitfalls, advanced safeguards like cgroup limits and eBPF tracing.
The thundering herd problem in Linux causes multiple processes or threads to wake up simultaneously for a single event, leading to wasted CPU cycles, excessive context switches, lock contention, and severe performance degradation, but it can be mitigated with kernel tweaks, epoll flags, SO_REUSEPORT, thread pools, and other strategies.
This guide explains how to use the iostat command to monitor Linux disk I/O, interprets its key metrics, compares I/O behavior to everyday queueing, and offers practical solutions for performance problems based on real output samples.
This article introduces bpftrace, an eBPF‑based dynamic tracing tool for Linux, explains its core concepts, technical architecture, installation methods, basic syntax, and demonstrates real‑world performance analysis, fault diagnosis, and security monitoring scenarios while comparing it with DTrace, SystemTap, and BCC.
This guide explains why disk I/O matters for server performance and walks through using the Linux dd command to create a test file, run read/write benchmarks, interpret the results from /proc/diskstats, and automate regular checks with a cron script.
This comprehensive guide explains how to diagnose Linux performance problems—from CPU, memory, disk I/O, and network bottlenecks to system load—using tools like top, vmstat, perf, iostat, netstat, and flame graphs, and demonstrates the methodology with a real‑world Nginx case study.
This guide explains how high concurrency and fast response depend on throughput and latency, defines key performance metrics, shows how to interpret average load, CPU context switches, and memory usage, and provides practical Linux tools and command‑line examples for diagnosing and tuning system performance.
This article provides an in‑depth technical walkthrough of Android’s traced_perf component, covering its code layout, integration with Perfetto, IPC mechanisms, perf_event handling, sample acquisition, stack unwinding, and trace writing, while illustrating each step with diagrams and reference links.
This comprehensive guide explains Linux performance optimization by covering core metrics such as throughput and latency, the meaning of average load, CPU context switching, memory management, common bottlenecks, and a suite of diagnostic tools like vmstat, pidstat, perf, and strace, plus practical tuning techniques for both applications and the operating system.
This article guides developers through classifying system‑level and business‑level bugs, using Linux utilities like perf, ps, and vmstat for quick root‑cause analysis, and outlines effective code‑design patterns and architectural strategies—caching, rate‑limiting, and high‑availability—to prevent and resolve production incidents.
This article walks through Linux process states, fork/exec creation, CPU vs load metrics, network I/O models, and a real‑world Nginx high‑load case, showing how strace, accept_mutex, and SO_REUSEPORT can resolve the issue.
This article introduces practical CPU and memory subsystem performance tuning for Kunpeng processors, covering optimization concepts, key parameters, common monitoring tools such as top, perf and numactl, and detailed methods like NUMA binding, prefetch control, timer tuning, TLB page size adjustment, and thread concurrency optimization.
This article explains how to monitor CPU usage on Linux using tools like top, vmstat, and gstat, analyzes common causes of high user and system CPU time, and provides practical tuning methods such as adjusting process priority, using ulimit, and configuring program‑specific CPU settings.
This article explains the meaning of iostat fields, clarifies common misconceptions about svctm and await, details each /proc/diskstats column, shows how to calculate key performance metrics, and illustrates practical analysis with real disk examples.
This article outlines how the Linux CPU scheduler manages threads and interrupts, explains context switches, run queue length, CPU utilization metrics, typical performance thresholds, and introduces monitoring tools like vmstat and mpstat, followed by real‑world troubleshooting examples.
Switching from traditional 512‑byte sectors to 4096‑byte (4K) sectors enhances error‑correction, increases usable capacity, and can affect I/O performance; proper partition alignment is crucial, as misaligned partitions cause significant read/write slowdowns across various Linux file systems, as shown by benchmark tests.
This article explains how to use the sar (System Activity Reporter) tool on Linux to collect and analyze performance data, covering past and real‑time statistics, CPU, memory, I/O and swap metrics, installation steps, and a full list of command‑line options.
