HL AsySocket 服务开发框架 - 一般性测试1

一 概述

• Socket服务器性能要好就要经过无数次的测试，来保证，以下是记录一次的测试经过。
• 机器配置：Inter(R) Core(TM) i3-2310m CPU 2.10GHz RAM 6.00GB,64位系统
• 今天的测试主要是5个客服端连接和连续发送2000数据包，每个包大概在55192大小。

二 测试准备工作

 

在测试之前需要准备哪些呢。

启动多客服端.bat 代码如下：

@echo off
echo 请稍等......
for /l %%i in (1,2,12) do  start HLAsySocketClient.exe %%i
echo 完毕！

 

三 开始测试

 

第一步我们先配置好相关的配置启动一个Socket服务如下配置

<?xml version="1.0" encoding="gb2312" ?>
<socketServer  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema">
  <servers>
    <server Disabled="" bufferSize="" maxConnectionNumber="" />
    <server serviceName="" bufferSize="" maxConnectionNumber="" />
    <server serviceName="" bufferSize="" maxConnectionNumber="" />
  </servers>
</socketServer>

HLAsySocketServer.exe.config配置相关的目前把数据库相关的配置先存放在这里，还有运行时目录的查找。

<?xml version="1.0"?>
<configuration>
  <appSettings>
    <add key="server" value="LUOMG-PC"/>
    <add key="uid" value="sa"/>
    <add key="pwd" value="123456"/>
    <add key="db" value="3983"/>
  </appSettings>
  <runtime>
    <assemblyBinding xmlns="urn:schemas-microsoft-com:asm.v1">
      <probing privatePath="Servers"/>
    </assemblyBinding>
  </runtime>
<startup><supportedRuntime version="v4.0" sku=".NETFramework,Version=v4.0"/></startup></configuration>

第二步启动Socket服务 如下：

目前只能显示连接数，缓冲区还没有开发所以还不能显示。

第三步启动启动多客服端.bat 如下：

HLAsySocketClient.exe 在启动的时候连接到服务器端。

全部启动完成并且DataGrid里有数据也就是说连接成功，并第一次向服务器发送请求，并回报成功了。

第四步，我们逐个点击客服端的测试按钮，旁边文本框显示的是向服务发送请求的次数，lab里是成功回报的次数，如上图。

一共启动了5个客服端，每个客服端连续向服务器发送2000次请求，也就是10000次。

 

四 测试结果

 

我们先来看一下SystemLog文件夹生成了那些文件

里有错误日志，性能日志，一般性日志。

生成日志对应的代码，如下

public class DeviceSocket : ICommand<AsyncSocketSession>
    {
        public void ExecuteCommand(AsyncSocketSession session, SocketSendData commandData)
        {
            Device dev = new Device();
            if ("list".Equals(commandData.StrMessage, StringComparison.OrdinalIgnoreCase))
            {
              DataTable dt = dev.GetList("").Tables[0];
              commandData.ObjMessage = dt;
            }
            else if ("add".Equals(commandData.StrMessage, StringComparison.OrdinalIgnoreCase))
            { 

            }
            session.SendResponse("", commandData);
            LogManage.AddActionLog(string.Format("BusinessLogicLayer.SocketCommand 业务对象 {0} 业务类型 {1} ",this.GetType().Name,commandData.StrMessage));
        }
    }

也就是说这个日志是业务处理完，并反馈给客服端，后成功的日志了。

在错误日志文件里面发现了很多:

==============================================

[2013-7-17 22:19:59]:System.OutOfMemoryException: 引发类型为“System.OutOfMemoryException”的异常。

   在 System.Text.StringBuilder.ToString()

   在 System.IO.StringWriter.ToString()

   在 System.Data.DataSet.GetRemotingDiffGram(DataTable table)

   在 System.Data.DataTable.SerializeDataTable(SerializationInfo info, StreamingContext context, Boolean isSingleTable, SerializationFormat remotingFormat)

   在 System.Data.DataTable.GetObjectData(SerializationInfo info, StreamingContext context)

   在 System.Runtime.Serialization.Formatters.Binary.WriteObjectInfo.InitSerialize(Object obj, ISurrogateSelector surrogateSelector, StreamingContext context, SerObjectInfoInit serObjectInfoInit, IFormatterConverter converter, ObjectWriter objectWriter, SerializationBinder binder)

   在 System.Runtime.Serialization.Formatters.Binary.ObjectWriter.Write(WriteObjectInfo objectInfo, NameInfo memberNameInfo, NameInfo typeNameInfo)

   在 System.Runtime.Serialization.Formatters.Binary.ObjectWriter.Serialize(Object graph, Header[] inHeaders, __BinaryWriter serWriter, Boolean fCheck)

   在 System.Runtime.Serialization.Formatters.Binary.BinaryFormatter.Serialize(Stream serializationStream, Object graph, Header[] headers, Boolean fCheck)

   在 System.Runtime.Serialization.Formatters.Binary.BinaryFormatter.Serialize(Stream serializationStream, Object graph)

   在 HLAsySocket.Serialization.BinaryFormatterWrapper.Serialze(Object obj)

   在 HLAsySocket.SocketTreaty.OrganizeTreaty(String strData, SocketSendData objData)

   在 HLAsySocket.AsyncSocketSession.SendResponse(String strData, SocketSendData objData)

   在 BusinessLogicLayer.SocketCommand.DeviceSocket.ExecuteCommand(AsyncSocketSession session, SocketSendData commandData)

   在 HLAsySocket.AsyncSocketSession.ExecuteCommand(SocketSendData cmdInfo)

   在 HLAsySocket.AsyncSocketSession.ProcessReceive(MySocketAsyncEventArgs e)

==============================================

性能日志如下：

[2013-7-17 22:12:54]:---------------------------------------------------

CPU Usage: 0.00%, Physical Memory Usage: 49,524,736.00, Total Thread Count: 13

AvailableWorkingThreads: 1021, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:13:24]:---------------------------------------------------

CPU Usage: 23.22%, Physical Memory Usage: 402,817,024.00, Total Thread Count: 25

AvailableWorkingThreads: 1019, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:13:54]:---------------------------------------------------

CPU Usage: 38.94%, Physical Memory Usage: 828,608,512.00, Total Thread Count: 28

AvailableWorkingThreads: 1018, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:14:24]:---------------------------------------------------

CPU Usage: 35.66%, Physical Memory Usage: 1,123,102,720.00, Total Thread Count: 26

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:14:54]:---------------------------------------------------

CPU Usage: 27.43%, Physical Memory Usage: 1,332,174,848.00, Total Thread Count: 26

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:15:24]:---------------------------------------------------

CPU Usage: 23.68%, Physical Memory Usage: 1,523,118,080.00, Total Thread Count: 23

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:15:54]:---------------------------------------------------

CPU Usage: 25.20%, Physical Memory Usage: 1,654,398,976.00, Total Thread Count: 23

AvailableWorkingThreads: 1016, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:16:24]:---------------------------------------------------

CPU Usage: 25.25%, Physical Memory Usage: 1,654,448,128.00, Total Thread Count: 22

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:16:54]:---------------------------------------------------

CPU Usage: 31.06%, Physical Memory Usage: 1,654,358,016.00, Total Thread Count: 20

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:17:24]:---------------------------------------------------

CPU Usage: 30.66%, Physical Memory Usage: 1,654,378,496.00, Total Thread Count: 20

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:17:54]:---------------------------------------------------

CPU Usage: 33.64%, Physical Memory Usage: 1,654,394,880.00, Total Thread Count: 20

AvailableWorkingThreads: 1017, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:18:24]:---------------------------------------------------

CPU Usage: 34.98%, Physical Memory Usage: 1,654,403,072.00, Total Thread Count: 20

AvailableWorkingThreads: 1019, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:18:54]:---------------------------------------------------

CPU Usage: 33.10%, Physical Memory Usage: 1,654,484,992.00, Total Thread Count: 21

AvailableWorkingThreads: 1019, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:19:25]:---------------------------------------------------

CPU Usage: 38.04%, Physical Memory Usage: 1,654,562,816.00, Total Thread Count: 23

AvailableWorkingThreads: 1019, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:19:54]:---------------------------------------------------

CPU Usage: 38.59%, Physical Memory Usage: 1,654,542,336.00, Total Thread Count: 22

AvailableWorkingThreads: 1019, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

[2013-7-17 22:20:24]:---------------------------------------------------

CPU Usage: 5.80%, Physical Memory Usage: 1,654,280,192.00, Total Thread Count: 14

AvailableWorkingThreads: 1021, AvailableCompletionPortThreads: 1000

MaxWorkingThreads: 1023, MaxCompletionPortThreads: 1000

==============================================

查看Windows任务管理器：

 

五 后续工作

 

根据以下的测试结果，还有很长的一段路要走啊。

性能相关：

1：CPU过高，内存过高，把CPU和内存全部消耗完了。

2：查询了10000多次的数据库，在这里需要加入缓存，以减少一点消耗。

3：多个客服端连续发送相对大的数据量，处理太慢，好多都卡着不走了。

4：有时候会出现丢包，粘包 现象，数据包解析不过直接报数据解析错误，在这里需要加入一个发送队列会不会好一点。

5：加入发送超时，接收超时机制。但是如果设置是3秒超时，而执行数据库操作要5秒完成那不是？??!!!!。

6：以上功能加完在这个测试基础上在加一个客服端专门做连接关闭操作来干扰其他客服端发送接收看看她的稳定性。