Examples of solving problems with TCP sticky packets in C#

This article mainly introduces in detail the solution to the TCP sticky problem in C#, which has certain reference value. Interested friends can refer to it

1. TCP Sticky The principle of packet generation

1.TCP sticky packet means that several packets of data sent by the sender are glued into one packet when the receiver receives it. From the receiving buffer, the header of the latter packet of data follows immediately The end of the previous packet of data. There are many reasons for the sticky phenomenon. It may be caused by the sender or the receiver.

2. The sticky packet caused by the sender is caused by the TCP protocol itself. In order to improve the transmission efficiency of TCP, the sender often has to collect enough data before sending a packet of data. If the data sent several times in a row is very small, usually TCP will combine the data into one packet according to the optimization algorithm and send it out at once, so that the receiver receives the sticky packet data. The sticky packet caused by the receiver is because the user process of the receiver does not receive the data in time, which leads to the sticky packet phenomenon.

3. This is because the receiver first puts the received data in the system receiving buffer, and the user process takes the data from the buffer. If the next packet of data arrives, the previous packet of data has not been received by the user process. Take it away, then the next packet of data is placed in the system receiving buffer after receiving the previous packet of data, and the user process fetches data from the system receiving buffer according to the preset buffer size, so that multiple packets are fetched at one time data. ,

2. Solution principle and code implementation

1. Use the header (fixed length, which contains the length of the package body, dynamically obtained when sending) + package body transmission mechanism. As shown in the figure

HeaderSize stores the length of the packet body, and the HeaderSize itself is a fixed length of 4 bytes;

A complete data packet (L) = HeaderSize+BodySize;

2. Subpackaging algorithm

The basic idea is to first forcibly convert the received data stream to be processed, that is, the system buffer data (length set to M) into a predetermined structure data form, and extract the structural data length field L from it, and then calculate the first packet data length according to the packet header.

      M=System buffer size; L=Data packet sent by the user=HeaderSize+BodySize;

1) If L64634aa8278f87a05e1bcebcc7c0ce2eM, it means that the content of the data stream is not enough to form a complete structured data and needs to be merged with the next packet of data before processing.

4) The following is the code implementation (the server side of the HP-SOCKET framework receives data)

int headSize = 4;//包头长度 固定4
  byte[] surplusBuffer = null;//不完整的数据包，即用户自定义缓冲区
  /// <summary>
  /// 接收客户端发来的数据
  /// </summary>
  /// <param name="connId">每个客户的会话ID</param>
  /// <param name="bytes">缓冲区数据</param>
  /// <returns></returns>
  private HandleResult OnReceive(IntPtr connId, byte[] bytes) 
  {
   //bytes 为系统缓冲区数据
   //bytesRead为系统缓冲区长度
   int bytesRead = bytes.Length;
   if (bytesRead > 0)
   {
    if (surplusBuffer == null)//判断是不是第一次接收，为空说是第一次
     surplusBuffer = bytes;//把系统缓冲区数据放在自定义缓冲区里面
    else
     surplusBuffer = surplusBuffer.Concat(bytes).ToArray();//拼接上一次剩余的包
    //已经完成读取每个数据包长度
    int haveRead = 0;
    //这里totalLen的长度有可能大于缓冲区大小的(因为 这里的surplusBuffer 是系统缓冲区+不完整的数据包)
    int totalLen = surplusBuffer.Length;
    while (haveRead <= totalLen)
    {
     //如果在N此拆解后剩余的数据包连一个包头的长度都不够
     //说明是上次读取N个完整数据包后，剩下的最后一个非完整的数据包
     if (totalLen - haveRead < headSize)
     {
      byte[] byteSub = new byte[totalLen - haveRead];
      //把剩下不够一个完整的数据包存起来
      Buffer.BlockCopy(surplusBuffer, haveRead, byteSub, 0, totalLen - haveRead);
      surplusBuffer = byteSub;
      totalLen = 0;
      break;
     }
     //如果够了一个完整包，则读取包头的数据
     byte[] headByte = new byte[headSize];
     Buffer.BlockCopy(surplusBuffer, haveRead, headByte, 0, headSize);//从缓冲区里读取包头的字节
     int bodySize = BitConverter.ToInt32(headByte, 0);//从包头里面分析出包体的长度

     //这里的 haveRead=等于N个数据包的长度 从0开始；0,1,2,3....N
     //如果自定义缓冲区拆解N个包后的长度 大于 总长度，说最后一段数据不够一个完整的包了，拆出来保存
     if (haveRead + headSize + bodySize > totalLen)
     {
      byte[] byteSub = new byte[totalLen - haveRead];
      Buffer.BlockCopy(surplusBuffer, haveRead, byteSub, 0, totalLen - haveRead);
      surplusBuffer = byteSub;
      break;
     }
     else
     {
      //挨个分解每个包，解析成实际文字
      String strc = Encoding.UTF8.GetString(surplusBuffer, haveRead + headSize, bodySize);
      //AddMsg(string.Format(" > [OnReceive] -> {0}", strc));
      //依次累加当前的数据包的长度
      haveRead = haveRead + headSize + bodySize;
      if (headSize + bodySize == bytesRead)//如果当前接收的数据包长度正好等于缓冲区长度，则待拼接的不规则数据长度归0
      {
       surplusBuffer = null;//设置空 回到原始状态
       totalLen = 0;//清0
      }
     }
    }
   }
   return HandleResult.Ok;
  }

At this time, the work of unpacking and parsing text is completed. But it's not actually finished yet. If this code is for the client to receive data from the server, it will be fine.

Look carefully at the IntPtr connId session ID of each connection

private HandleResult OnReceive(IntPtr connId, byte[] bytes)
{
}

But the server side also needs to distinguish which session each data packet is generated, because the server side is multi-threaded and multi-user. mode, the first packet and the second may be data from different sessions, so the above code only works in single session mode.

I am going to solve this problem below.

Using c#safetyConcurrentDictionary,

latest code

//线程安全的字典
  ConcurrentDictionary<IntPtr, byte[]> dic = new ConcurrentDictionary<IntPtr, byte[]>();
  int headSize = 4;//包头长度 固定4
  /// <summary>
  /// 接收客户端发来的数据
  /// </summary>
  /// <param name="connId">每个客户的会话ID</param>
  /// <param name="bytes">缓冲区数据</param>
  /// <returns></returns>
  private HandleResult OnReceive(IntPtr connId, byte[] bytes) 
  {
   //bytes 为系统缓冲区数据
   //bytesRead为系统缓冲区长度
   int bytesRead = bytes.Length;
   if (bytesRead > 0)
   {
    byte[] surplusBuffer = null;
    if (dic.TryGetValue(connId, out surplusBuffer))
    {
     byte[] curBuffer = surplusBuffer.Concat(bytes).ToArray();//拼接上一次剩余的包
     //更新会话ID 的最新字节
     dic.TryUpdate(connId, curBuffer, surplusBuffer);
     surplusBuffer = curBuffer;//同步
    }
    else
    {
     //添加会话ID的bytes
     dic.TryAdd(connId, bytes);
     surplusBuffer = bytes;//同步
    }

    //已经完成读取每个数据包长度
    int haveRead = 0;
    //这里totalLen的长度有可能大于缓冲区大小的(因为 这里的surplusBuffer 是系统缓冲区+不完整的数据包)
    int totalLen = surplusBuffer.Length;
    while (haveRead <= totalLen)
    {
     //如果在N此拆解后剩余的数据包连一个包头的长度都不够
     //说明是上次读取N个完整数据包后，剩下的最后一个非完整的数据包
     if (totalLen - haveRead < headSize)
     {
      byte[] byteSub = new byte[totalLen - haveRead];
      //把剩下不够一个完整的数据包存起来
      Buffer.BlockCopy(surplusBuffer, haveRead, byteSub, 0, totalLen - haveRead);
      dic.TryUpdate(connId, byteSub, surplusBuffer);
      surplusBuffer = byteSub;
      totalLen = 0;
      break;
     }
     //如果够了一个完整包，则读取包头的数据
     byte[] headByte = new byte[headSize];
     Buffer.BlockCopy(surplusBuffer, haveRead, headByte, 0, headSize);//从缓冲区里读取包头的字节
     int bodySize = BitConverter.ToInt32(headByte, 0);//从包头里面分析出包体的长度

     //这里的 haveRead=等于N个数据包的长度 从0开始；0,1,2,3....N
     //如果自定义缓冲区拆解N个包后的长度 大于 总长度，说最后一段数据不够一个完整的包了，拆出来保存
     if (haveRead + headSize + bodySize > totalLen)
     {
      byte[] byteSub = new byte[totalLen - haveRead];
      Buffer.BlockCopy(surplusBuffer, haveRead, byteSub, 0, totalLen - haveRead);
      dic.TryUpdate(connId, byteSub, surplusBuffer);
      surplusBuffer = byteSub;
      break;
     }
     else
     {
      //挨个分解每个包，解析成实际文字
      String strc = Encoding.UTF8.GetString(surplusBuffer, haveRead + headSize, bodySize);
      AddMsg(string.Format(" > {0}[OnReceive] -> {1}", connId, strc));
      //依次累加当前的数据包的长度
      haveRead = haveRead + headSize + bodySize;
      if (headSize + bodySize == bytesRead)//如果当前接收的数据包长度正好等于缓冲区长度，则待拼接的不规则数据长度归0
      {
       byte[] xbtye=null;
       dic.TryRemove(connId, out xbtye);
       surplusBuffer = null;//设置空 回到原始状态
       totalLen = 0;//清0
      }
     }
    }
   }
   return HandleResult.Ok;
  }

This solves the problem, multiple clients Reception confusion caused by sessions. At this point all work is completed. The above code is just for reference and learning, if you really don’t want to go to such trouble. You can directly use the PACK model of the HP-SOCKET communication framework, which automatically solves the problem of sticky packets.

The above is the detailed content of Examples of solving problems with TCP sticky packets in C#. For more information, please follow other related articles on the PHP Chinese website!