Apache ab的使用

简介

ab的全称是ApacheBench,是 Apache 附带的一个小工具,专门用于 HTTP Server 的benchmark testing,可以同时模拟多个并发请求。前段时间看到公司的开发人员也在用它作一些测试,看起来也不错,很简单,也很容易使用,所以今天花一点时间看了一下。

通过下面的一个简单的例子和注释,相信大家可以更容易理解这个工具的使用。

一个简单的例子

/*在这个例子的一开始,我执行了这样一个命令 ab -n 10 -c 10 http://www.google.com/这个命令的意思是启动 ab ,向 www.google.com 发送10个请求(-n 10) ,并每次发送10个请求(-c 10)——也就是说一次都发过去了。跟着下面的是 ab 输出的测试报告,红色部分是我添加的注释。*/

C:\Program Files\Apache Software Foundation\Apache2.2\bin>ab -n 10 -c 10 http

://www.google.com/

This is ApacheBench, Version 2.0.40-dev <$Revision: 1.146 $> apache-2.0

Copyright 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/

Copyright 1997-2005 The Apache Software Foundation, http://www.apache.org/

 

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

 

 

Server Software:        GWS/2.1

Server Hostname:        www.google.com

Server Port:            80

 

Document Path:          /

Document Length:        230 bytes

 

Concurrency Level:      10

/*整个测试持续的时间*/

Time taken for tests:   3.234651 seconds

/*完成的请求数量*/

Complete requests:      10

/*失败的请求数量*/

Failed requests:        0

Write errors:           0

Non-2xx responses:      10

Keep-Alive requests:    10

/*整个场景中的网络传输量*/

Total transferred:      6020 bytes

/*整个场景中的HTML内容传输量*/

HTML transferred:       2300 bytes

/*大家最关心的指标之一,相当于 LR 中的 每秒事务数 ,后面括号中的 mean 表示这是一个平均值*/

Requests per second:    3.09 [#/sec] (mean)

/*大家最关心的指标之二,相当于 LR 中的 平均事务响应时间 ,后面括号中的 mean 表示这是一个平均值*/

Time per request:       3234.651 [ms] (mean)

/*这个还不知道是什么意思,有知道的朋友请留言,谢谢 ^_^ */

Time per request:       323.465 [ms] (mean, across all concurrent requests)

/*平均每秒网络上的流量,可以帮助排除是否存在网络流量过大导致响应时间延长的问题*/

Transfer rate:          1.55 [Kbytes/sec] received

/*网络上消耗的时间的分解,各项数据的具体算法还不是很清楚*/

Connection Times (ms)

              min  mean[+/-sd] median   max

Connect:       20  318 926.1     30    2954

Processing:    40 2160 1462.0   3034    3154

Waiting:       40 2160 1462.0   3034    3154

Total:         60 2479 1276.4   3064    3184

 

/*下面的内容为整个场景中所有请求的响应情况。在场景中每个请求都有一个响应时间,其中 50 的用户响应时间小于 3064 毫秒,60  的用户响应时间小于 3094 毫秒,最大的响应时间小于 3184 毫秒*/

Percentage of the requests served within a certain time (ms)

  50%   3064

  66%   3094

  75%   3124

  80%   3154

  90%   3184

  95%   3184

  98%   3184

  99%   3184

 100%   3184 (longest request)

 

更多信息

ab 不像 LR 那么强大,但是它足够轻便,如果只是在开发过程中想检查一下某个模块的响应情况,或者做一些场景比较简单的测试,ab 还是一个不错的选择——至少不用花费很多时间去学习 LR 那些复杂的功能,就更别说那 License 的价格了。

下面是 ab 的详细参数解释,大家有兴趣的可以研究一下,最近没有足够多的时间研究,如果哪位朋友有兴趣希望可以帮忙翻译一下每个参数的含义,有问题讨论也欢迎在这里回帖 ^_^

ab [ -A auth-username:password ] [ -c concurrency ] [ -C cookie-name=value ] [ -d ] [ -e csv-file ] [ -g gnuplot-file ] [ -h ] [ -H custom-header ] [ -i ] [ -k ] [ -n requests ] [ -p POST-file] [ -P proxy-auth-username:password ] [ -q ] [ -s ] [ -S ] [ -t timelimit ] [ -T content-type ] [ -v verbosity] [ -V ] [ -w ] [ -x 

-attributes ] [ -X proxy[:port] ] [ -y -attributes] [ -z 
-attributes ] [http://]hostname[:port]/path

 

-A auth-username:password

Supply BASIC Authentication credentials to the server. The username and password are separated by a single : and sent on the wire base64 encoded. The string is sent regardless of whether the server needs it (i.e., has sent an 401 authentication needed).

-c concurrency

Number of multiple requests to perform at a time. Default is one request at a time.

-C cookie-name=value

Add a Cookie: line to the request. The argument is typically in the form of a name=value pair. This field is repeatable.

-d

Do not display the "percentage served within XX [ms] table". (legacy support).

-e csv-file

Write a Comma separated value (CSV) file which contains for each percentage (from 1% to 100%) the time (in milliseconds) it took to serve that percentage of the requests. This is usually more useful than the 'gnuplot' file; as the results are already 'binned'.

-g gnuplot-file

Write all measured values out as a 'gnuplot' or TSV (Tab separate values) file. This file can easily be imported into packages like Gnuplot, IDL, Mathematica, Igor or even Excel. The labels are on the first line of the file.

-h

Display usage information.

-H custom-header

Append extra headers to the request. The argument is typically in the form of a valid header line, containing a colon-separated field-value pair (i.e."Accept-Encoding: zip/zop;8bit").

-i

Do HEAD requests instead of GET.

-k

Enable the HTTP KeepAlive feature, i.e., perform multiple requests within one HTTP session. Default is no KeepAlive.

-n requests

Number of requests to perform for the benchmarking session. The default is to just perform a single request which usually leads to non-representative benchmarking results.

-p POST-file

File containing data to POST.

-P proxy-auth-username:password

Supply BASIC Authentication credentials to a proxy en-route. The username and password are separated by a single : and sent on the wire base64 encoded. The string is sent regardless of whether the proxy needs it (i.e., has sent an 407 proxy authentication needed).

-q

When processing more than 150 requests, ab outputs a progress count on stderr every 10% or 100 requests or so. The -q flag will suppress these messages.

-s

When compiled in (ab -h will show you) use the SSL protected https rather than the http protocol. This feature is experimental and very rudimentary. You probably do not want to use it.

-S

Do not display the median and standard deviation values, nor display the warning/error messages when the average and median are more than one or two times the standard deviation apart. And default to the min/avg/max values. (legacy support).

-t timelimit

Maximum number of seconds to spend for benchmarking. This implies a -n 50000 internally. Use this to benchmark the server within a fixed total amount of time. Per default there is no timelimit.

-T content-type

Content-type header to use for POST data.

-v verbosity

Set verbosity level - 4 and above prints information on headers, 3 and above prints response codes (404, 200, etc.), 2 and above prints warnings and info.

-V

Display version number and exit.

-w

Print out results in HTML tables. Default table is two columns wide, with a white background.

-x 

-attributes

String to use as attributes for 

. Attributes are inserted here >.

-X proxy[:port]

Use a proxy server for the requests.

-y 

-attributes

String to use as attributes for 

.

-z 

-attributes

String to use as attributes for 

.

 

参数格式: -n number_of_requests

用来标明在本次测试中将要发送的请求的总数

 

 

参数格式: -c Number_of_requests_at_one_time

用来标明每次发送多少请求

 

e.g.  ab -n 100 -c 1 http://jackei.cnblogs.com/

表示向 http://jackei.cnblogs.com/ 这个 url 发送 100 次请求,每次发送 1 个。

 

 

参数格式: -e csv_file_full_path

表示把响应时间存放到一个 csv 文件中——CSV文件表示用逗号来分隔的文件。

 

e.g.  ab -n 100 -c 1 -e f:\ab.csv http://jackei.cnblogs.com/

测试完成之后可以可以在 f 盘的根目录下看到一个名为 ab.csv 的文件,你可以直接用 EXCEL 或者记事本打开,看到的结果就像下面这样:

Percentage served

Time in ms

 

0

881.00

 

1

921.00

 

2

971.00

 

3

1011.00

 

4

1021.00

 

5

1071.00

 

6

1071.00

 

7

1091.00

 

8

1101.00

 

9

1111.00

 

 

……

 

 

此处删节数百字 ^_^

 

 

……

 

 

92

3825.00

 

93

4165.00

 

94

4426.00

 

95

4826.00

 

96

4987.00

 

97

5447.00

 

98

5598.00

 

99

8131.00

 

可以看到在这次测试中响应时间的分布情况。

 

 

参数格式: -g tsv_file_full_path

表示把每个请求的响应情况存放到一个 tsv 文件中——TSV文件表示用 Tab 健来分隔的文件。

 

e.g.  ab -n 100 -c 1 -g f:\ab.xls http://jackei.cnblogs.com/

测试完成之后可以可以在 f 盘的根目录下看到一个名为 ab.xls 的文件,你可以直接用 EXCEL 打开,看到的结果就像下面这样:

starttime

seconds

ctime

dtime

ttime

wait

 

Thu Jul 20 19:50:28 2006

1.1534E+15

190

1792

1982

490

 

Thu Jul 20 19:50:30 2006

1.1534E+15

140

2043

2183

300

 

Thu Jul 20 19:50:32 2006

1.1534E+15

130

1822

1952

270

 

Thu Jul 20 19:50:34 2006

1.1534E+15

140

3144

3284

340

 

Thu Jul 20 19:50:37 2006

1.1534E+15

140

1262

1402

270

 

Thu Jul 20 19:50:39 2006

1.1534E+15

150

1522

1672

320

 

Thu Jul 20 19:50:40 2006

1.1534E+15

140

1832

1972

290

 

Thu Jul 20 19:50:42 2006

1.1534E+15

3134

2594

5728

290

 

Thu Jul 20 19:50:48 2006

1.1534E+15

140

2994

3134

300

 

 

……

 

 

 

 

 

 

此处删节数百字 ^_^

 

 

 

 

 

 

……

 

 

 

 

 

 

Thu Jul 20 19:54:50 2006

1.1534E+15

180

4717

4897

370

 

Thu Jul 20 19:54:55 2006

1.1534E+15

190

2854

3044

390

 

Thu Jul 20 19:54:58 2006

1.1534E+15

190

3244

3434

390

 

Thu Jul 20 19:55:01 2006

1.1534E+15

190

3244

3434

420

 

Thu Jul 20 19:55:05 2006

1.1534E+15

190

3455

3645

400

 

Thu Jul 20 19:55:09 2006

1.1534E+15

180

2033

2213

360

 

Thu Jul 20 19:55:11 2006

1.1534E+15

190

3855

4045

390

 

Thu Jul 20 19:55:15 2006

1.1534E+15

200

4516

4716

390

 

其中 starttime 比较容易理解,后面的几个参数分别表示

secondes: 还没有搞明白什么意思,有知道的朋友帮忙解答一下吧 ^_^

ctime:connection time

dtime: processing time

ttime: total time, = connection time + processing time

wait:wait time

 

相对于上面的那个 -e 的参数,-g 参数可以提供更多的信息,帮助我们来了解更具体的系统响应情况。

另外,文件中的记录默认是按照 starttime 排序的,你也可以在 EXCEL 中自己随意编辑。v




相关链接

ab 是 Apache 的一个安装组件,所以需要下载 Apache 安装后才能使用,可以访问 Apache 的项目主页来下载 http://httpd.apache.org/download.cgi

ab 的更多信息可以参加 Apache 主页上的描述

http://httpd.apache.org/docs/2.0/programs/ab.html

你可能感兴趣的:(Apache ab的使用)