ab test nginx Nginx performance optimization

Transfer: https: //www.cnblogs.com/nulige/p/9369700.html

1. Performance Optimization Overview

Before doing performance optimization, we need to consider the following

1. System Structure current bottleneck
- OUTCOME MEASURES
- pressure test
2. Understand the business model
- Interface type of business
- The system hierarchical structure
3. Performance and Safety
- Good safety performance weak
- Good low safety performance

2. stress testing tool

1. Install the pressure testing toolab

[root@nginx-lua ~]# yum install httpd-tools -y

2. use a decompression tool measuring

[root@nginx-lua ~]# ab -n 200 -c 2 http://127.0.0.1/ //-n总的请求次数 //-c并发请求数 //-k是否开启长连接

3. Configure Nginxstatic website and tomcatdynamic web environment

[root@nginx-lua conf.d]# cat jsp.conf 
server {
        server_name localhost;
        listen 80;
        location / {
                root /soft/code;
                try_files $uri @java_page;
                index index.jsp index.html;
        }
        location @java_page{
                proxy_pass http://192.168.56.20:8080; } } //分别给Nginx准备静态网站 [root@nginx-lua ~]# cat /soft/code/bgx.html <h1> Ab Load </h1> //给Tomcat准备静态网站文件 [root@tomcat-node1-20 ROOT]# cat /soft/tomcat-8080/webapps/ROOT/bgx.html <h1> Ab Load </h1>

4. Use abtool pressure test

//进行压力测试
[root@Nginx conf.d]# ab -n2000 -c2 http://127.0.0.1/bgx.html ... Server Software: nginx/1.12.2 Server Hostname: 127.0.0.1 Server Port: 80 Document Path: /bgx.html Document Length: 19 bytes Concurrency Level: 200 # 总花费总时长 Time taken for tests: 1.013 seconds # 总请求数 Complete requests: 2000 # 请求失败数 Failed requests: 0 Write errors: 0 Total transferred: 510000 bytes HTML transferred: 38000 bytes # 每秒多少请求/s(总请求出/总共完成的时间) Requests per second: 9333.23 [#/sec] (mean) # 客户端访问服务端, 单个请求所需花费的时间 Time per request: 101.315 [ms] (mean) # 服务端处理请求的时间 Time per request: 0.507 [ms] (mean, across all concurrent requests) # 判断网络传输速率, 观察网络是否存在瓶颈 Transfer rate: 491.58 [Kbytes/sec] received

5. nginxUnder the bgxdocument is removed, the pressure will be measured again by the tomcatprocessing

Concurrency Level:      200
Time taken for tests: 1.028 seconds Complete requests: 2000 Failed requests: 0 Write errors: 0 Total transferred: 510000 bytes HTML transferred: 38000 bytes Requests per second: 1945.09 [#/sec] (mean) Time per request: 102.823 [ms] (mean) Time per request: 0.514 [ms] (mean, across all concurrent requests) Transfer rate: 484.37 [Kbytes/sec] received

3. Impact Performance

Easy to affect the performance of the overall concern

1. Network
- Network traffic
- Whether the network packet loss
- These requests affect the calling of http
2. System
- There is no hardware disk corruption, disk rate
- System load, memory, system stability
3. Service
- Connection optimization, optimization request
- Do the corresponding service settings as business forms
4. Program
- Interface Performance
- Processing speed
- Program execution efficiency
5. Database

More or less have some association between each of the services and infrastructure services, we need the entire layered architecture, find the corresponding short-board systems or services, then optimize

4. The system performance optimization

File handles, Linux everything is a file, the file handle is an index can be understood as
- With the file handle will call our process frequently increase
- There are restrictions on the default file handles, so that a process can not be unlimited calling
- Each process and the need to limit how much each service using the file handle
- File handle is required to be adjusted to optimize parameters
Setting mode
- Modify global system
- Users modify locality
- Locality process modifications

vim /etc/security/limits.conf // root soft nofile 65535 root hard nofile 65535 // all users globally * soft nofile 25535 * hard nofile 25535 // for Nginx process worker_rlimit_nofile 65535 for the root user; // // root user soft remind // hard // nofile limit the number of files to configure the maximum size of items // 65535

Note: You can use the ulimit -a command to view open files 65535 system has a maximum number of open files value.

5.Nginx Performance Optimization

CPUBetween pro and reduce the frequent migration process will continue, reducing the performance loss

1. View the current CPUphysical status

 
           [root@nginx ~] 
           # lscpu |grep "CPU(s)" 
          
           CPU(s):                24 
          
           On-line CPU(s) list:   0-23 
          
           NUMA node0 CPU(s):     0,2,4,6,8,10,12,14,16,18,20,22 
          
           NUMA node1 CPU(s):     1,3,5,7,9,11,13,15,17,19,21,23

//2颗物理cpu,没颗cpu12核心, 总共24核心

2. The Nginx workerprocess tied to a different core

 
           // 
           启动多少worker进程, 官方建议和cpu核心一致, 第一种绑定组合方式 
          
           #worker_processes 24; 
          
           #worker_cpu_affinity 000000000001 000000000010 000000000100 000000001000 000000010000 000000100000 000001000000 000010000000 000100000000 001000000000 010000000000 10000000000; 
          
           // 
           第二种方式 
          
           #worker_processes 2; 
          
           #worker_cpu_affinity 101010101010 010101010101; 
          
           // 
           最佳方式绑定方式 
          
           worker_processes auto; 
          
           worker_cpu_affinity auto;

3. Review the nginx workerprocess is bound to the correspondingcpu

ps -eo pid,args,psr|grep [n]ginx

4. NginxGeneral Optimization profile

 
           [root@nginx ~] 
           # cat nginx.conf 
          
           user nginx; 
          
           worker_processes auto; 
          
           worker_cpu_affinity auto; 
          
           error_log  
           /var/log/nginx/error 
           .log warn; 
          
           pid  
           /run/nginx 
           .pid; 
          
           #调整至1w以上,负荷较高建议2-3w以上 
          
           worker_rlimit_nofile 65535; 
          
           events { 
          
           use epoll; 
          
           #限制每个进程能处理多少个连接请求,10240x16 
          
           worker_connections 10240; 
          
           } 
          
           http { 
          
           include              
           /etc/nginx/mime 
           .types; 
          
           default_type        application 
           /octet-stream 
           ; 
          
           # 统一使用utf-8字符集 
          
           charset utf-8; 
          
           log_format  main   
           '$remote_addr - $remote_user [$time_local] "$request" ' 
          
           '$status $body_bytes_sent "$http_referer" ' 
          
           '"$http_user_agent" "$http_x_forwarded_for"' 
           ; 
          
           access_log   
           /var/log/nginx/access 
           .log  main; 
          
           # Core module 
          
           sendfile            on; 
          
           # 静态资源服务器建议打开 
          
           tcp_nopush          on; 
          
           # 动态资源服务建议打开,需要打开keepalived 
          
           tcp_nodelay         on; 
          
           keepalive_timeout   65; 
          
           # Gzip module 
          
           gzip  
           on; 
          
           gzip_disable  
           "MSIE [1-6]\." 
           ; 
          
           gzip_http_version 1.1; 
          
           # Virtal Server 
          
           include  
           /etc/nginx/conf 
           .d/*.conf; 
          
           }

1. Performance Optimization Overview

Before doing performance optimization, we need to consider the following

1. System Structure current bottleneck
- OUTCOME MEASURES
- pressure test
2. Understand the business model
- Interface type of business
- The system hierarchical structure
3. Performance and Safety
- Good safety performance weak
- Good low safety performance

2. stress testing tool

1. Install the pressure testing toolab

[root@nginx-lua ~]# yum install httpd-tools -y

2. use a decompression tool measuring

[root@nginx-lua ~]# ab -n 200 -c 2 http://127.0.0.1/ //-n总的请求次数 //-c并发请求数 //-k是否开启长连接

3. Configure Nginxstatic website and tomcatdynamic web environment

[root@nginx-lua conf.d]# cat jsp.conf 
server {
        server_name localhost;
        listen 80;
        location / {
                root /soft/code;
                try_files $uri @java_page;
                index index.jsp index.html;
        }
        location @java_page{
                proxy_pass http://192.168.56.20:8080; } } //分别给Nginx准备静态网站 [root@nginx-lua ~]# cat /soft/code/bgx.html <h1> Ab Load </h1> //给Tomcat准备静态网站文件 [root@tomcat-node1-20 ROOT]# cat /soft/tomcat-8080/webapps/ROOT/bgx.html <h1> Ab Load </h1>

4. Use abtool pressure test

//进行压力测试
[root@Nginx conf.d]# ab -n2000 -c2 http://127.0.0.1/bgx.html ... Server Software: nginx/1.12.2 Server Hostname: 127.0.0.1 Server Port: 80 Document Path: /bgx.html Document Length: 19 bytes Concurrency Level: 200 # 总花费总时长 Time taken for tests: 1.013 seconds # 总请求数 Complete requests: 2000 # 请求失败数 Failed requests: 0 Write errors: 0 Total transferred: 510000 bytes HTML transferred: 38000 bytes # 每秒多少请求/s(总请求出/总共完成的时间) Requests per second: 9333.23 [#/sec] (mean) # 客户端访问服务端, 单个请求所需花费的时间 Time per request: 101.315 [ms] (mean) # 服务端处理请求的时间 Time per request: 0.507 [ms] (mean, across all concurrent requests) # 判断网络传输速率, 观察网络是否存在瓶颈 Transfer rate: 491.58 [Kbytes/sec] received

5. nginxUnder the bgxdocument is removed, the pressure will be measured again by the tomcatprocessing

Concurrency Level:      200
Time taken for tests: 1.028 seconds Complete requests: 2000 Failed requests: 0 Write errors: 0 Total transferred: 510000 bytes HTML transferred: 38000 bytes Requests per second: 1945.09 [#/sec] (mean) Time per request: 102.823 [ms] (mean) Time per request: 0.514 [ms] (mean, across all concurrent requests) Transfer rate: 484.37 [Kbytes/sec] received

3. Impact Performance

Easy to affect the performance of the overall concern

1. Network
- Network traffic
- Whether the network packet loss
- These requests affect the calling of http
2. System
- There is no hardware disk corruption, disk rate
- System load, memory, system stability
3. Service
- Connection optimization, optimization request
- Do the corresponding service settings as business forms
4. Program
- Interface Performance
- Processing speed
- 程序执行效率
5.数据库

每个架构服务与服务之间都或多或少有一些关联, 我们需要将整个架构进行分层, 找到对应系统或服务的短板, 然后进行优化

4.系统性能优化

文件句柄, Linux一切皆文件，文件句柄可以理解为就是一个索引
- 文件句柄会随着我们进程的调用频繁增加
- 系统默认对文件句柄有限制，不能让一个进程无限的调用
- 需要限制每个进程和每个服务使用多大的文件句柄
- 文件句柄是必须要调整的优化参数
设置方式
- 系统全局性修改
- 用户局部性修改
- 进程局部性修改

vim /etc/security/limits.conf //针对root用户 root soft nofile 65535 root hard nofile 65535 //所有用户, 全局 * soft nofile 25535 * hard nofile 25535 //对于Nginx进程 worker_rlimit_nofile 65535; //root用户 //soft提醒 //hard限制 //nofile文件数配置项 //65535最大大小

备注：可以使用ulimit -a 命令查看open files 65535 系统最大打开文件数的值。

5.Nginx性能优化

CPU亲和, 减少进程之间不断频繁迁移, 减少性能损耗

1.查看当前CPU物理状态

 
         [root@nginx ~] 
         # lscpu |grep "CPU(s)" 
        
         CPU(s):                24 
        
         On-line CPU(s) list:   0-23 
        
         NUMA node0 CPU(s):     0,2,4,6,8,10,12,14,16,18,20,22 
        
         NUMA node1 CPU(s):     1,3,5,7,9,11,13,15,17,19,21,23

//2颗物理cpu,没颗cpu12核心, 总共24核心

2.将Nginx worker进程绑到不同的核心上

 
         // 
         启动多少worker进程, 官方建议和cpu核心一致, 第一种绑定组合方式 
        
         #worker_processes 24; 
        
         #worker_cpu_affinity 000000000001 000000000010 000000000100 000000001000 000000010000 000000100000 000001000000 000010000000 000100000000 001000000000 010000000000 10000000000; 
        
         // 
         第二种方式 
        
         #worker_processes 2; 
        
         #worker_cpu_affinity 101010101010 010101010101; 
        
         // 
         最佳方式绑定方式 
        
         worker_processes auto; 
        
         worker_cpu_affinity auto;

3.查看nginx worker进程绑定至对应cpu

ps -eo pid,args,psr|grep [n]ginx

4.Nginx通用优化配置文件

 
         [root@nginx ~] 
         # cat nginx.conf 
        
         user nginx; 
        
         worker_processes auto; 
        
         worker_cpu_affinity auto; 
        
         error_log  
         /var/log/nginx/error 
         .log warn; 
        
         pid  
         /run/nginx 
         .pid; 
        
         #调整至1w以上,负荷较高建议2-3w以上 
        
         worker_rlimit_nofile 65535; 
        
         events { 
        
         use epoll; 
        
         #限制每个进程能处理多少个连接请求,10240x16 
        
         worker_connections 10240; 
        
         } 
        
         http { 
        
         include              
         /etc/nginx/mime 
         .types; 
        
         default_type        application 
         /octet-stream 
         ; 
        
         # 统一使用utf-8字符集 
        
         charset utf-8; 
        
         log_format  main   
         '$remote_addr - $remote_user [$time_local] "$request" ' 
        
         '$status $body_bytes_sent "$http_referer" ' 
        
         '"$http_user_agent" "$http_x_forwarded_for"' 
         ; 
        
         access_log   
         /var/log/nginx/access 
         .log  main; 
        
         # Core module 
        
         sendfile            on; 
        
         # 静态资源服务器建议打开 
        
         tcp_nopush          on; 
        
         # 动态资源服务建议打开,需要打开keepalived 
        
         tcp_nodelay         on; 
        
         keepalive_timeout   65; 
        
         # Gzip module 
        
         gzip  
         on; 
        
         gzip_disable  
         "MSIE [1-6]\." 
         ; 
        
         gzip_http_version 1.1; 
        
         # Virtal Server 
        
         include  
         /etc/nginx/conf 
         .d/*.conf; 
        
         }