Nginx Slab算法研究  

Nginx Slab Algorithm Research

作　　者：宋雅琴 郭志川[1] SONG Yaqin1,2, GUO Zhichuan1(1National Network New Media Engineering Research Center, Institute of Acoustics, Chinese Academy of Sciences, Beijing, 100190,China,＂ ZUnlverslty. ＇ of Chinese Academy of Sciences, Beijing, 100190, Chin)

机构地区：[1]中国科学院声学研究所国家网络新媒体工程技术研究中心,北京100190 [2]中国科学院大学,北京100190

出　　处：《网络新媒体技术》2018年第2期54-61,共8页Network New Media Technology

摘　　要：Nginx设计了简单的内存池进行内存管理来降低开发中对内存资源管理的复杂度。Nginx各进程间使用共享内存的方式共享数据,而对共享内存的内存池进行管理的方法是基于经典的Slab算法,其通过构造小的内存块来避免内存碎片、使用链表方式连接有限的页面来提高分配速率。本文详细介绍Nginx的Slab算法,对其进行总结,并与经典的Linux内核的Slab算法对比。Nginx designed a simple memory pool for memory management to reduce the complexity of memory management. Nginx uses shared memory to share data among processes. The way to manage memory pool with shared memory is based on the classic Slab algo- rithm, which avoids memory fragmentation by constructing small blocks of memory and increases memory allocation rate by using linked lists to connect limited pages. This article details Nginx＇ s Slab algorithm, summarizes it, and compares it with the classic Slab algo- rithm in Linux kernel.

关 键 词：内存管理 Slab算法 Nginx服务器 内存池.内存碎片 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]