1、 nGrinder是什么
nGrinder是一个免费的、开放源代码的Web性能测试平台。运行在应用中间件服务器中运行。它由一个控制端和多个代理端组成。通过控制端（浏览器访问）建立测试场景，然后通过分发到代理端进行压力测试，是一个分布式的测试场景。
测试过程是基于Jython脚本来模拟浏览器或其他终端使用HTTP协议访问站点来完成的。测试过程中采集的数据将通过动态图和数据表的形式展示出来。用户可以方便的看到tps、以及测试目标机器的cpu和内存等情况。
nGrinder依赖的其他开源项目：
Grinder一个开源的Java负载测试框架，它通过很多负载注射器来为分布式测试提供了便利。支持用于执行测试脚本的Jython脚本引擎。
Cubrid是一个全面开源，且完全免费的关系数据库管理系统。CUBRID为高效执行Web应用进行了高度优化，特别是需要处理大数据量和高并发请求的复杂商务服务。通过提供独特的最优化特性，CUBRID可以支持更多的并发请求，更少的响应时间。
Jython是一种完整的语言，而不是一个Java翻译器或仅仅是一个Python编译器，它是一个Python语言在Java中的完全实现。 Jython也有很多从CPython中继承的模块库。最有趣的事情是Jython不像CPython或其他任何高级语言，它提供了对其实现语言的一切存 取。所以Jython不仅给你提供了Python的库，同时也提供了所有的Java类。这使其有一个巨大的资源库。
Spring是一个开源框架，是为了解决企业应用程序开发复杂性由Rod Johnson创建的。框架的主要优势之一就是其分层架构，分层架构允许使用者选择使用哪一个组件，同时为 J2EE 应用程序开发提供集成的框架。Spring使用基本的JavaBean来完成以前只可能由EJB完成的事情。然而，Spring的用途不仅限于服务器端的 开发。从简单性、可测试性和松耦合的角度而言，任何 Java应用都可以从Spring中受益。
2、 windows平台的nGrinder配置
2.1 前提条件
nGrinder下在：http://sourceforge.net/projects/ngrinder/files/ngrinder-3.2.3/。本次演示使用了tomcat因此需要下载ngrinder-controller-3.2.3-with-tomcat.zip和ngrinder-core-3.2.3-agent-package.zip。
jdk1.6：下载地址http://www.java.net/download/jdk6/6u10/promoted
/b32/binaries/jdk-6u10-rc2-bin-b32-windows-i586-p-12_sep_2008.exe
tomcat 6 下载地址：http://tomcat.apache.org/
2.2 配置控制端
安装jkd1.6和tomcat 在这就不再累述，配置JAVA_HOME和JRE_HOME，放到系统环境变量的最前面。然后将ngrinder-controller-3.2.3- with-tomcat.zip加压缩后，放入webapps文件夹下，运行该文件夹下的 startup.bat。将如下图启动成功：
在浏览器中输入http://localhost:8080/ ngrinder-controller ，进入登录页面，输入用户名密码（用户名admin，密码admin），进入系统，点击右上角admin下拉菜单，选择系统配置，如下图所示：
修改配置如下：
# Set verbose to print the detailed log #verbose=true # testmode makes some several constraints not applied. # if testmode is true, security and pluginsupport is disabled. #testmode=false # true if enabling security manager. The default value is false #security=true # false if disabling ngrinder usage report. it will send host ip and executed tests number to Goolge Analytics. # The default value is true #usage.report=false # true if the plugin is enabled. # This is not the option applied on the fly. You need to restart the controller. #pluginsupport=true # false if you want to make the user profile fields not mandatory. # Default value is true. #user.security=false # the maximum number of agents which can be attached into one test. #agent.max.size=10 # the maximum number of vusers which can be initiated per one agent #agent.max.vuser=2000 # the maximum runcount of test per one agent #agent.max.runcount=10000 # the maximum running hour of one test. #agent.max.runhour=8 # the listener port of monitor. monitor.listen.port=13243 #provide the base URL of the ngrinder. #ngrinder.http.url= # the starting port number of consoles which will be mapped to each test. # This is not the option applied on the fly. You need to reboot to apply this. #ngrinder.console.portbase=12000 # the port number for the agent connection. ngrinder.agent.control.port=8088 # The count of allowed concurrent tests. #ngrinder.max.concurrenttest=10 # How many milliseconds console will wait until all agents are connected. #ngrinder.max.waitingmilliseconds=5000 # select the ip for controller to use. ngrinder.controller.ipaddress=192.168.68.134 # validation timeout in the unit of sec. #ngrinder.validation.timeout=100 # To be compatible with old agents (before 3.1.1), please set following to true. # If you use the latest version agent, just set it false. ngrinder.dist.logback=false # From 3.1.1, nGrinder doesn't check the file distribution result to speed up the test execution. # If your agent is located in the far places or you distribute big files everyday, you'd better to change this to true. ngrinder.dist.safe=false # If some region has the slow network, it may need the safe distribution by force #ngrinder.dist.safe.region= # Set the safe distribution threshold to enable safe distribution for specific transfer size by force. # ngrinder.dist.safe.threshold=1000000 # You can point your own rss page on the ngrinder first page. #ngrinder.frontpage.rss=http://www.cubrid.org/wiki_ngrinder/rss # If you want to provide your own custom help page. please modify this #ngrinder.help.url=http://www.cubrid.org/wiki_ngrinder/entry/user-guide # language default en/kr/cn is available. #ngrinder.langauge.default=en # The monitor port is configurable. The default port is 13243. #monitor.listen.port=13243 ###################################################################################### # clustering configuration. # This is not the option applied on the fly. You need to reboot to apply this. ###################################################################################### # These should be very carefully set. # You can refer http://www.cubrid.org/wiki_ngrinder/entry/controller-clustering-guide # if you want to enable ngrinder controller clustering. please enable below. #ngrinder.cluster.mode=false # cache cluster uri of all clustered controller servers. #ngrinder.cluster.uris=192.168.1.1;192.168.2.2;192.168.3.3 # cache cluster listener port. #ngrinder.cluster.listener.port=40003 # region setting of this controller. # When running ngrinder as cluster mode, the ${NGRINDER_HOME} should be shared by multiple controllers. # We expect this to be done by running ${NGRINDER_HOME} on NFS. # However this option is not to set here. Because each controller in cluster must have system-ex.conf # which should be stored in ${NGRINDER_EX_HOME} in the local so that each node cluster looks different value. # Please locate followings this on the ${NGRINDER_EX_HOME}(by default .ngrinder_ex)/system-ex.conf # ngrinder.cluster.region=Beijing |
2.3 配置代理端
将ngrinder-core-3.2.3-agent-package.zip解压，建立环境变量NGRINDER_AGENT_HOME，变量值为C:\DOCUME~1\ADMINI~1\.grinder_agent。
运行代理文件夹下的startuo.bat后，进入C:\Documents and Settings\Administrator\.ngrinder_agent，修改agent.conf如下：
#start.mode=monitor #monitor.listen.port=13243 # If you want to monitor bind to the different local ip not automatically selected ip. Specify below field. #monitor.host=hostname_or_ip start.mode=agent agent.console.ip=192.168.68.134 agent.console.port=8088 #agent.region= #agent.hostid= #agent.servermode=true # provide more agent java execution option if necessary. #agent.javaopt= # set following false if you want to use more than 1G Xmx memory per a agent process. #agent.useXmxLimit=true #agent.same.console.host=true # please uncomment the following option if you want to send all logs to the controller. #agent.send.all.logs=true |
其中agent.console.ip为控制端ip，agent.console.port为控制端监听端口。
