Performance

1. 首页
2. Performance

Redis Performance Optimization In-Depth: From Connection Pool to Memory Management 2026

Deep dive into Redis performance optimization across multiple dimensions: connection pool configuration, Pipeline batching, bigkey detection and governance, eviction policy selection (LRU vs LFU), systematic solutions for cache penetration/avalanche/breakdown, high availability architecture selection, and production-grade monitoring system. With extensive practical code examples and configuration templates.

数据库技术 2026年09月19日 0 点赞 0 评论 22 浏览

PyTorch 2.0 Compiler 深度剖析：从 TorchDynamo FX Graph 到 TorchInductor Triton CodeGen 的完全工程指南

PyTorch 2.0 Compiler 深度剖析：从 TorchDynamo Python 字节码拦截、FX Graph 中间表示捕获、AOTAutograd 自动微分分解、TorchInductor Triton CodeGen GPU 内核代码生成、算子融合与内存优化、CUDA Graph 捕获、动态形状支持、分布式并行编译、到生产级部署与性能基准测试的完整工程指南。

后端开发 2026年09月20日 0 点赞 0 评论 15 浏览

Linux io_uring: Async IO Revolution from Kernel to Production

Deep analysis of Linux io_uring: core architecture, three working modes, liburing dev practice, performance benchmarks, production deployment.

编程技术 2026年09月26日 0 点赞 0 评论 3 浏览

eBPF: Linux Kernel Programmability Revolution

Deep-dive technical guide covering eBPF architecture, XDP packet processing, observability tracing, runtime security, and practical development with libbpf and CO-RE. Complete benchmark data and production deployment patterns.

编程技术 2026年09月26日 0 点赞 0 评论 1 浏览

io_uring: Linux Kernel's Revolutionary Asynchronous I/O Framework

io_uring represents a fundamental rethinking of asynchronous I/O in the Linux kernel. Comprehensive deep dive into io_uring architecture, submission/completion queues, SQPOLL mode, fixed buffers, linked operations, multishot, and real-world performance benchmarks.

编程技术 2026年09月26日 0 点赞 0 评论 2 浏览

Linux io_uring: A Deep Engineering Analysis of the Next-Generation Asynchronous I/O Framework

An in-depth engineering analysis of Linux io_uring: dual-ring buffer architecture, system call interface, liburing programming, SQPOLL/IOPOLL modes, fixed buffers/files, performance benchmarks, and production deployment practices.

编程技术 2026年09月26日 0 点赞 0 评论 4 浏览

Linux Kernel Scheduler: A Deep Engineering Analysis of CPU Scheduling from CFS to EEVDF

Deep engineering analysis of Linux CPU scheduling: from CFS virtual runtime to EEVDF deadline-based scheduling, NUMA awareness, EAS, and performance benchmarks.

编程技术 2026年09月26日 0 点赞 0 评论 6 浏览

Linux Kernel Scheduler: A Deep Engineering Analysis of CPU Scheduling from CFS to EEVDF

Deep engineering analysis of Linux CPU scheduling: from CFS virtual runtime to EEVDF deadline-based scheduling, NUMA awareness, EAS, and performance benchmarks.

编程技术 2026年09月26日 0 点赞 0 评论 4 浏览