openresty 性能优化，

文章由LinuxBoy分享于2023-03-25 02:03:03热评（626）

openresty 性能优化，

openresty 性能优化

ab 压测

命令格式：ab [options] url

huli@hudeMacBook-Pro ~ % ab -h
Usage: ab [options] [http[s]://]hostname[:port]/path

# 可选参数
Options are:
    -n requests     Number of requests to perform
                    * 发送的请求总数
 
    -c concurrency  Number of multiple requests to make at a time
                    * 请求并发数

    -t timelimit    Seconds to max. to spend on benchmarking
                    This implies -n 50000
                    * 压测最多执行的时间

    -s timeout      Seconds to max. wait for each response
                    Default is 30 seconds
                    * 响应超时时间，默认30s

    -b windowsize   Size of TCP send/receive buffer, in bytes
    -B address      Address to bind to when making outgoing connections
    -p postfile     File containing data to POST. Remember also to set -T
    -u putfile      File containing data to PUT. Remember also to set -T
    -T content-type Content-type header to use for POST/PUT data, eg.
                    'application/x-www-form-urlencoded'
                    Default is 'text/plain'
    -v verbosity    How much troubleshooting info to print
    -w              Print out results in HTML tables
    -i              Use HEAD instead of GET
    -x attributes   String to insert as table attributes
    -y attributes   String to insert as tr attributes
    -z attributes   String to insert as td or th attributes
    -C attribute    Add cookie, eg. 'Apache=1234'. (repeatable)
    -H attribute    Add Arbitrary header line, eg. 'Accept-Encoding: gzip'
                    Inserted after all normal header lines. (repeatable)
    -A attribute    Add Basic WWW Authentication, the attributes
                    are a colon separated username and password.
    -P attribute    Add Basic Proxy Authentication, the attributes
                    are a colon separated username and password.
    -X proxy:port   Proxyserver and port number to use
    -V              Print version number and exit
    -k              Use HTTP KeepAlive feature
    -d              Do not show percentiles served table.
    -S              Do not show confidence estimators and warnings.
    -q              Do not show progress when doing more than 150 requests
    -l              Accept variable document length (use this for dynamic pages)
    -g filename     Output collected data to gnuplot format file.
    -e filename     Output CSV file with percentages served
    -r              Don't exit on socket receive errors.
    -m method       Method name
    -h              Display usage information (this message)
    -I              Disable TLS Server Name Indication (SNI) extension
    -Z ciphersuite  Specify SSL/TLS cipher suite (See openssl ciphers)
    -f protocol     Specify SSL/TLS protocol
                    (TLS1, TLS1.1, TLS1.2 or ALL)
    -E certfile     Specify optional client certificate chain and private key

使用示例

huli@hudeMacBook-Pro ~ % ab -n 100 https://www.taobao.com/
This is ApacheBench, Version 2.3 <$Revision: 1879490 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking www.taobao.com (be patient).....done


Server Software:        Tengine
Server Hostname:        www.taobao.com
Server Port:            443
SSL/TLS Protocol:       TLSv1.2,ECDHE-ECDSA-AES128-GCM-SHA256,256,128
Server Temp Key:        ECDH X25519 253 bits
TLS Server Name:        www.taobao.com

Document Path:          /
Document Length:        86840 bytes

Concurrency Level:      1                   # 并发数
Time taken for tests:   24.865 seconds      # 完成测试花费的时间
Complete requests:      100                 # 完成发送的请求数
Failed requests:        0                   # 失败的请求数
Total transferred:      8796159 bytes       # 总传输数据量大小
HTML transferred:       8684000 bytes       # html数据量大小
Requests per second:    4.02 [#/sec] (mean) # 平均每秒执行的请求数
Time per request:       248.646 [ms] (mean) # 平均每个请求执行时间
Time per request:       248.646 [ms] (mean, across all concurrent requests)
                                            # 所有并发请求，平均每个请求执行时间
Transfer rate:          345.47 [Kbytes/sec] received       # 传输速率

Connection Times (ms)     # 连接耗时：连接、处理、响应等待
              min  mean[+/-sd] median   max
Connect:       96  154 137.2    120    1034
Processing:    74   95  36.1     86     396
Waiting:       26   63  38.6     58     372
Total:        178  248 154.4    207    1125

Percentage of the requests served within a certain time (ms)
                          # 请求执行时间分布
  50%    207              # 50%的请求在207ms内执行完成
  66%    218
  75%    235
  80%    245
  90%    328
  95%    467
  98%   1117
  99%   1125
 100%   1125 (longest request)

连接数优化

*************

连接数限制

连接数超出系统限制，通常会报出如下错误：

too many open files

用户连接数：临时操作

# 查看当前用户描述符限制
[root@7623085b4b89 /]# ulimit -n
1048576

# 临时修改限制，重启后失效
[root@7623085b4b89 /]# ulimit -n 200000 
[root@7623085b4b89 /]# ulimit -n       
200000

用户连接数：永久操作

vi /etc/ssh/sshd_config， 添加如下内容：
* hard nofile 102400
* soft nofile 102400
保存后退出，重新登录，即可生效

# 相关说明
*：表示所有用户
nofile：最大文件打开数

系统连接数（所有用户连接数总和）：临时操作

设置连接数：sysctl -w fs.file-max=102400
查看连接数：sysctl fs.file-max


# 示例
[root@ab103f74b163 /]# sysctl -w fs.file-max=102400
fs.file-max = 102400
[root@ab103f74b163 /]# sysctl fs.file-max
fs.file-max = 102400

系统连接数：永久操作

# 设置文件描述符术
vi /etc/sysctl.conf
==> 添加 fs.file-max=102400
退出后，执行 ==>  sysctl -p

# 查看文件描述符数
sysctl fs.file-max
cat /proc/sys/fs/file-max


# 示例
[root@ab103f74b163 /]# sysctl -w fs.file-max=102400
fs.file-max = 102400
[root@ab103f74b163 /]# sysctl fs.file-max
fs.file-max = 102400
[root@ab103f74b163 /]# cat /proc/sys/fs/file-max
102400

[root@ab103f74b163 /]# vi /etc/sysctl.conf
[root@ab103f74b163 /]# sysctl -p
fs.file-max = 100000

[root@ab103f74b163 /]# cat /proc/sys/fs/file-max
100000
[root@ab103f74b163 /]# sysctl fs.file-max
fs.file-max = 100000

*************

openresty 设置

nginx.conf

worker_processes auto;          # 工作进程个数，on：根据cpu数自动确定
wprker_rlimit_nofile 65535;     # 最多打开的文件数量

events {
    worker_connections 65535;   # 每个工作进程可接受的连接数
    multi_accept on;            # 是否允许一个工作进程响应多个请求
}

*************

示例

default.conf

server {
    listen       80;
    server_name  localhost;

    location / {
        root   /usr/local/openresty/nginx/html;
        index  index.html index.htm;
    }

    location /test {
        echo $server_name $remote_addr;
        echo "test";
    }

    error_page   500 502 503 504  /50x.html;
    location = /50x.html {
        root   /usr/local/openresty/nginx/html;
    }

}

nginx.conf

...

events {
    worker_connections  1024;
}

...

创建openresty容器

docker run -it -d --net fixed --ip 172.18.0.2 -p 9000:80 \
-v /Users/huli/lua/openresty/conn/nginx.conf:/usr/local/openresty/nginx/conf/nginx.conf \
-v /Users/huli/lua/openresty/conn/default.conf:/etc/nginx/conf.d/default.conf \
--name open-conn lihu12344/openresty

ab 测试：ab -n 2000 -c 1000 http://localhost:9000/test

# 请求全部执行成功
huli@hudeMacBook-Pro conn % ab -n 2000 -c 1000 http://localhost:9000/test
This is ApacheBench, Version 2.3 <$Revision: 1879490 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking localhost (be patient)
Completed 200 requests
Completed 400 requests
Completed 600 requests
Completed 800 requests
Completed 1000 requests
Completed 1200 requests
Completed 1400 requests
Completed 1600 requests
Completed 1800 requests
Completed 2000 requests
Finished 2000 requests


Server Software:        openresty/1.21.4.1
Server Hostname:        localhost
Server Port:            9000

Document Path:          /test
Document Length:        31 bytes

Concurrency Level:      1000
Time taken for tests:   1.760 seconds
Complete requests:      2000               # 发送请求2000
Failed requests:        0                  # 失败的请求数为0
Total transferred:      348000 bytes
HTML transferred:       62000 bytes
Requests per second:    1136.21 [#/sec] (mean)
Time per request:       880.123 [ms] (mean)
Time per request:       0.880 [ms] (mean, across all concurrent requests)
Transfer rate:          193.07 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   20  13.0     19      51
Processing:    53  719 479.1    437    1434
Waiting:        5  695 471.5    413    1429
Total:         56  739 471.4    459    1446

Percentage of the requests served within a certain time (ms)
  50%    459
  66%   1177
  75%   1211
  80%   1319
  90%   1342
  95%   1366
  98%   1388
  99%   1395
 100%   1446 (longest request)

ab 测试：ab -n 2000 -c 2000 http://localhost:9000/test

# 部分请求执行失败
huli@hudeMacBook-Pro conn % ab -n 2000 -c 2000 http://localhost:9000/test
This is ApacheBench, Version 2.3 <$Revision: 1879490 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking localhost (be patient)
Completed 200 requests
Completed 400 requests
Completed 600 requests
Completed 800 requests
Completed 1000 requests
Completed 1200 requests
Completed 1400 requests
Completed 1600 requests
Completed 1800 requests
Completed 2000 requests
Finished 2000 requests


Server Software:        openresty/1.21.4.1
Server Hostname:        localhost
Server Port:            9000

Document Path:          /test
Document Length:        31 bytes

Concurrency Level:      2000
Time taken for tests:   0.856 seconds
Complete requests:      2000              # 发送请求2000
Failed requests:        448               # 448个请求执行失败
   (Connect: 0, Receive: 0, Length: 448, Exceptions: 0)
Total transferred:      270048 bytes
HTML transferred:       48112 bytes
Requests per second:    2337.75 [#/sec] (mean)
Time per request:       855.525 [ms] (mean)
Time per request:       0.428 [ms] (mean, across all concurrent requests)
Transfer rate:          308.25 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   63  21.6     64     104
Processing:   108  488 150.2    383     730
Waiting:        0  401 254.8    378     724
Total:        109  550 131.9    470     779

Percentage of the requests served within a certain time (ms)
  50%    470
  66%    646
  75%    695
  80%    710
  90%    727
  95%    744
  98%    762
  99%    764
 100%    779 (longest request)

nginx.conf：修改worker_connections = 2000，重启应用

...

events {
    worker_connections  2000;
}

...

ab 测试：ab -n 2000 -c 2000 http://localhost:9000/test

# 修改worker_connections =2000后，请求全部执行成功
huli@hudeMacBook-Pro conn % ab -n 2000 -c 2000 http://localhost:9000/test
This is ApacheBench, Version 2.3 <$Revision: 1879490 $>
Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Licensed to The Apache Software Foundation, http://www.apache.org/

Benchmarking localhost (be patient)
Completed 200 requests
Completed 400 requests
Completed 600 requests
Completed 800 requests
Completed 1000 requests
Completed 1200 requests
Completed 1400 requests
Completed 1600 requests
Completed 1800 requests
Completed 2000 requests
Finished 2000 requests


Server Software:        openresty/1.21.4.1
Server Hostname:        localhost
Server Port:            9000

Document Path:          /test
Document Length:        26 bytes

Concurrency Level:      2000
Time taken for tests:   1.126 seconds
Complete requests:      2000              # 请求发送数2000
Failed requests:        0                 # 请求失败数为0
Total transferred:      338000 bytes
HTML transferred:       52000 bytes
Requests per second:    1776.36 [#/sec] (mean)
Time per request:       1125.901 [ms] (mean)
Time per request:       0.563 [ms] (mean, across all concurrent requests)
Transfer rate:          293.17 [Kbytes/sec] received

Connection Times (ms)
              min  mean[+/-sd] median   max
Connect:        0   69  25.4     72     112
Processing:   117  607 216.5    586     993
Waiting:        5  602 217.6    583     982
Total:        118  676 199.1    672    1042

Percentage of the requests served within a certain time (ms)
  50%    672
  66%    788
  75%    827
  80%    875
  90%    969
  95%    991
  98%   1002
  99%   1013
 100%   1042 (longest request)

客户端限制

限制同一个ip并发数

# 同一个ip地址限制10个并发连接
http {
    limit_conn_zone $binary_remote_addr zone=perip:10m;
    limit_conn perip 10;
}

# 相关说明
limit_conn_zone：开辟一块内存空间保存客户端ip，空间名称perip，空间大小10m
limit_conn：限制客户端请求数
$binary_remote_addr：以二进制形式保存客户端ip

limit_conn_zone、limit_conn也可设置在server、location中，实现不同级别的并发控制

限制虚拟主机的并发数

# 对虚拟主机限制10个并发连接
http {
    limit_conn_zone $server_name zone=perserver:10m;

    server {
        listen 80;
        server_name localhost;

        limit_conn perserver 10;
    }
}

# 相关说明
limit_conn_zone：开辟一块内存空间保存虚拟主机，空间名称perserver，空间大小10m
limit_conn：限制客户端请求数
$server_name：获取虚拟主机

限制响应的传输速率

# 限制响应传输速率
http {
    limit_rate 100k;
    limit_rate_after 10m;

    ...
}

# 相关说明
limit_rate：限制传输速率为100kb/s
limit_rate_after：如果设置了该指令，表示传输制定数据量（10m）后，开始限速
                  如果没有设置，则一开始就会进行限速

浏览器缓存

nginx 缓存响应头

Last_Modified：nginx自动生成
ETag：nginx自动生成

Cache-Control：缓存控制响应头
* Cache-Control:max-age=N   ==> 缓存有效期N秒
* Cache-Control:no-cache    ==> 本地保存缓存，如果没有更新使用本地缓存（304）；
                                如果有更新，去服务器获取最新的数据（200），并更新本地缓存
* Cache-Control:no-store    ==> 不存储缓存，每次请求都会访问服务器
* Cache-Control:max-age=N,must-revalidate
                            ==> 缓存有效期N秒，过期前直接使用本地缓存
                                过期后，如果没有更新，直接使用本地缓存（304）
                                过期后，如果有更新，获取最新数据（200），并更新本地缓存

Expires：缓存过期时间

expires 指令设置缓存有效期

server {

    ...

    location \.(gif|jpg|jpeg|png|bmp|swf)$ {

        ...

        expires 10d;
    }

    location \.(css|js)$ {

        ...

        expires 12h;
    }
}


# 相关说明
即使mginx没有设置expires，浏览器也会自动缓存常见的静态资源（根据拓展名识别静态资源）；
使用expires可控制浏览器缓存的有效时间
当浏览器需要更新静态资源时，可设置动态路径（*css?arg=random_int）更新资源

推荐文章：

openresty 性能优化，