WO2015188770A1 - 一种发送通路及发送通路的数据处理方法 - Google Patents

一种发送通路及发送通路的数据处理方法 Download PDF

Info

Publication number
WO2015188770A1
WO2015188770A1 PCT/CN2015/081284 CN2015081284W WO2015188770A1 WO 2015188770 A1 WO2015188770 A1 WO 2015188770A1 CN 2015081284 W CN2015081284 W CN 2015081284W WO 2015188770 A1 WO2015188770 A1 WO 2015188770A1
Authority
WO
WIPO (PCT)
Prior art keywords
frame
data
amount
module
stored
Prior art date
Application number
PCT/CN2015/081284
Other languages
English (en)
French (fr)
Inventor
游泳
吴锡攸
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2015188770A1 publication Critical patent/WO2015188770A1/zh

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/30Flow control; Congestion control in combination with information about buffer occupancy at either end or at transit nodes
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/26Flow control; Congestion control using explicit feedback to the source, e.g. choke packets
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/26Flow control; Congestion control using explicit feedback to the source, e.g. choke packets
    • H04L47/263Rate modification at the source after receiving feedback
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/36Flow control; Congestion control by determining packet size, e.g. maximum transfer unit [MTU]

Definitions

  • the present invention relates to the field of network device technologies, and in particular, to a data processing method for a transmission path and a transmission path.
  • the Transmission Control Protocol is a communication protocol designed to provide a reliable end-to-end communication service over an unreliable internetwork.
  • TCP Transmission Control Protocol
  • the sending of the packet by the sending end is generally triggered by the received TCP acknowledgement (ACK) frame from the receiving end, and is generally not received if the ACK frame sent by the receiving end is not received.
  • ACK TCP acknowledgement
  • a message will be sent (except for timeout retransmissions).
  • Some TCP-based applications such as File Transfer Protocol (FTP) download, if the ACK frame sent by the client to the server is blocked during network transmission, the delay of the ACK frame reaching the server will increase. The delay in starting the next message transmission will also increase accordingly, resulting in a decrease in FTP download efficiency.
  • FTP File Transfer Protocol
  • an ACK frame and a message containing normal service data are sent to the link by a same first input first output (FIFO) structure transmission path.
  • the transmission path (Tx_path) It mainly includes a frame descriptor first-in first-out (FD_FIFO) unit, a transmission processing (Tx_Prcs) unit, and a transmission buffer (TxBuffer) unit.
  • the traffic management (TM) module of the pre-stage dispatches a frame descriptor (FD) to the transmission path, and the FD records information such as the buffer address and frame length of the frame.
  • the scheduling of the TM has a certain burst and delay.
  • the FD_FIFO is designed to store the FD.
  • the Tx_Prcs reads the corresponding packet data from the packet buffer according to the cache address information of the frame recorded in the FD.
  • the Packet Buffer is a memory, which may be shared.
  • the random access memory (RAM) the message data read back from the Packet Buffer will be cached in the TxBuffer.
  • the transmission is initiated when there is an entire frame in the TxBuffer or the buffered data reaches the set byte transmission threshold, which prevents the long packet from being interrupted.
  • the FD_FIFO buffers the FD and the TxBuffer buffers the packets, and the cached data constitutes the head of the FD that newly enters the transmission path.
  • the newly entered FD corresponds to all the packets in the queue. After the transmission is completed, it can be sent out.
  • the message a indicates that the FD newly enters the transmission path, a before the a, a message d is waiting to be transmitted, and the message a is an ACK frame with a high priority.
  • the strict priority scheduling in the TM can ensure that the TM preferentially schedules the high priority ACK frame into the transmission path without the back pressure signal. There are multiple long packets to be sent in front of the ACK frame, which will cause a large blocking delay for the ACK frame. If there are more short packets in the message in front of the message a, the blocking delay will be lower.
  • the TM since the number of messages that can be buffered in the transmission path is limited, when the FD_FIFO in the transmission path has filled the message, the TM will receive the back pressure signal of the transmission path, and the data cannot be scheduled to enter the transmission path until The transmission path completes sending a message to the link, and cancels the back pressure signal to the TM. In this case, the blocking delay of the ACK frame will be larger.
  • the number of long packets buffered in the transmission path may be large, that is, the amount of data stored in the transmission path may be large, and all packets including the ACK frame may be higher in the transmission path.
  • the blocking delay however, the prior art has not proposed a solution that can effectively control the amount of data stored in the transmission path, thereby controlling the blocking delay of the message in the transmission path.
  • the embodiment of the invention provides a data processing method for a transmission path and a transmission path, which is used for controlling the amount of data stored in the transmission path, thereby controlling the blocking delay of the message.
  • an embodiment of the present invention provides a transmission path, including:
  • a frame descriptor FD storage module configured to receive and store a frame descriptor FD of a first frame sent by the traffic management TM module, where the FD of the first frame includes a length value of the first frame;
  • a calculating module configured to determine, according to a length value of the first frame stored in the FD storage module, an amount of data stored in the sending path, where the amount of data stored in the sending path includes a data amount of a frame corresponding to the stored FD The amount of data of the stored frame;
  • a determining module configured to determine whether the amount of data stored in the sending path determined by the calculating module is greater than a preset data amount threshold
  • the indicating module is configured to send a back pressure signal to the TM module when the determining module determines that the amount of data stored in the sending path is greater than a preset data amount threshold, where the back pressure signal is used to indicate the TM module Stop transmitting the FD of the second frame to the transmission path.
  • the computing module is specifically configured to:
  • the sending path further includes:
  • a sending module configured to send first data to the link, where the first data is data of a frame stored in the sending path;
  • the calculating module is further configured to subtract the data amount of the first data by the amount of data stored in the sending path to update the amount of data stored in the sending path.
  • the calculating module is further configured to determine the number of FDs stored by the FD storage module;
  • the determining module is further configured to determine whether the number of FDs determined by the calculating module reaches a preset threshold
  • the indicating module sends the back pressure signal to the TM module when the determining module determines that the number of FDs determined by the calculating module reaches a preset threshold.
  • the FD of the first frame further includes a starting storage address of the first frame in the memory
  • the sending path further includes:
  • a frame storage module configured to acquire the first frame from the memory according to a length value of a first frame stored in the FD storage module and a starting storage address of the first frame, and store the first frame One frame.
  • an embodiment of the present invention provides a data processing method for a transmission path, including:
  • the sending path receives and stores a frame descriptor FD of the first frame sent by the traffic management TM module, where the FD of the first frame includes a length value of the first frame;
  • the transmission path determines the amount of data stored in the transmission path according to the length value of the first frame, and the data amount stored in the transmission path includes a data amount of a frame corresponding to the stored FD and a data amount of the stored frame;
  • the sending path determines whether the amount of data stored in the sending path is greater than a preset data amount threshold
  • the sending path sends a back pressure signal to the TM module, where the back pressure signal is used to indicate that the TM module stops to the sending path. Send the FD of the second frame.
  • the sending path determines, according to the length value of the first frame, the amount of data stored in the sending path, including:
  • the transmission path determines the sum of the amount of data stored before storing the FD of the first frame and the length value of the first frame as the amount of data stored in the transmission path.
  • the method further includes:
  • the transmission path subtracts the amount of data of the first data by the amount of data stored in the transmission path to update the amount of data stored in the transmission path.
  • the method further includes:
  • the sending path determines the number of FDs stored in the sending path
  • the transmitting path sends the back pressure signal to the TM module when it is determined that the number of stored FDs reaches a preset number threshold.
  • the FD of the first frame further includes the first frame in the memory
  • the initial storage address in the method the method further includes:
  • the transmitting path acquires the first frame from the memory according to a length value of the first frame and a starting storage address of the first frame, and stores the first frame.
  • the sending path when the amount of data stored in the transmission path is greater than a preset data amount threshold, the sending path sends a back pressure signal to the TM module, so that the amount of data stored in the sending path is controlled within a certain range, and correspondingly The blocking delay of the new packet entering the transmission path is also controlled.
  • FIG. 1 is a schematic structural diagram of a data transmitting apparatus in the prior art
  • FIG. 2 is a schematic structural diagram of a transmission path according to an embodiment of the present invention.
  • FIG. 3 is a schematic structural diagram of a data transmitting apparatus according to another embodiment of the present invention.
  • FIG. 4 is a schematic diagram showing the amount of data accumulated in a transmission path in the prior art
  • FIG. 5 is a schematic diagram of data amount accumulated in a transmission path according to an embodiment of the present invention.
  • FIG. 6 is a schematic flowchart of a data processing method of a transmission path according to an embodiment of the present invention.
  • the embodiment of the invention provides a data processing method for the transmission path and the transmission path, which is used to control the blocking delay of the message by the sending path, and avoid the situation that the blocking delay of the message is too high.
  • the transmission path stores the FD of the first frame in the FD queue until the FD of the first frame enters the head position, and the transmission path takes out the FD of the first frame, according to The extracted FD information is obtained from the memory to the first frame, and the first frame is stored to the frame queue. Finally, the transmission path sequentially transmits the frames in the order of the frame queue.
  • the amount of data stored in the transmission path refers to the sum of the data amounts of the frames to be sent by the transmission path. Since the FD is stored in the transmission path and the frame is stored, the FD is used to obtain the corresponding frame. And sending out, therefore, the amount of data stored in the transmission path includes the amount of data of the frame corresponding to the stored FD and the amount of data of the stored frame.
  • the sending path provided by the embodiment of the present invention includes:
  • the FD storage module 210 is configured to receive and store an FD of the first frame sent by the traffic management module, where the FD of the first frame includes a length value of the first frame.
  • the calculating module 220 is configured to determine, according to a length value of the first frame stored in the FD storage module, an amount of data stored in the sending path, where the amount of data stored in the sending path includes a data amount of a frame corresponding to the stored FD And the amount of data stored in the frame;
  • the determining module 230 is configured to determine whether the amount of data stored in the sending path determined by the calculating module is greater than a preset data amount threshold;
  • the indicating module 240 is configured to send a back pressure signal to the TM module when the determining module determines that the amount of data stored in the sending path is greater than a preset data amount threshold, where the back pressure signal is used to indicate the TM
  • the module stops transmitting the FD of the second frame to the transmission path.
  • the calculation module 220 is specifically configured to:
  • the calculating module is further configured to determine the number of FDs stored by the FD storage module;
  • the determining module is further configured to determine whether the number of FDs determined by the calculating module reaches a preset threshold
  • the indicating module sends the back pressure signal to the TM module when the determining module determines that the number of FDs determined by the calculating module reaches a preset threshold.
  • the FD is stored in units of the transmission path, and the number of FDs that can be stored in the transmission path is limited.
  • the back pressure signal is transmitted to the TM module.
  • the number of FDs stored in the current transmission path is monitored, and when the number of FDs reaches a preset threshold, the back pressure signal is sent to the TM module.
  • the threshold can be set to any value not greater than the maximum number of storages, so as to achieve more flexible control of the data processing of the transmission path.
  • the transmission path may send packets to the network faster than the scheduling speed of the TM, reducing the efficiency of the transmission path.
  • the FD of the first frame further includes a starting storage address of the first frame in the memory, where the sending path further includes:
  • a frame storage module configured to acquire the first frame from the memory according to a length value of a first frame stored in the FD storage module and a starting storage address of the first frame, and store the first frame One frame.
  • the sending path further includes:
  • a sending module configured to send first data to the link, where the first data is data of a frame stored in the sending path;
  • the calculating module is further configured to subtract the data amount of the first data by the amount of data stored in the sending path to update the amount of data stored in the sending path.
  • the transmission path keeps recording the amount of data currently stored, and when there is FD entry, the amount of recorded data is increased, and when there is a message, the amount of recorded data is reduced, and the amount of stored data of the transmission path record is ensured to be accurate. Sex.
  • the first data sent by the sending module is 2 bytes or 16 bytes of data in a frame of the head of the transmission path frame storage module.
  • the sending path provided by the embodiment of the present invention specifically includes:
  • a frame descriptor FIFO unit 310 configured to receive and store an FD of a first frame sent by the TM module, and obtain a length value of the first frame according to the FD;
  • the sending processing unit 320 is configured to extract, according to the FD of the first frame from the frame descriptor FIFO unit 310, the initial storage address of the first frame in the memory, and store according to the start of the first frame. The address and the length value of the first frame are obtained from the memory, and the first frame is written to the transmission buffer unit 330;
  • the sending buffer unit 330 is configured to send the first frame to the data and send the link to the link;
  • the counting unit (BYTE_COUNTER) 340 is configured to add the amount of data stored in the recorded transmission path to the length value of the first frame acquired from the frame descriptor first-in first-out unit 310, and whenever the transmission buffer is detected When the unit 330 sends the data in the packet to the link, the amount of data stored in the recorded transmission path is subtracted from the amount of data sent by the transmission buffer unit 330; and it is determined whether the amount of data stored in the recorded transmission path is greater than a preset value.
  • the data volume threshold is configured to send a back pressure signal to the TM module to stop the sending of the FD by the TM module when the amount of data stored in the recorded transmission path is greater than a preset data amount threshold.
  • the amount of data sent by the sending buffer unit 330 to the link is a fixed value that can be preset in advance, and is generally 2 bytes or 16 bytes. Whenever the counting unit detects that the sending buffer unit 330 sends the data once, In the case of the link, the data amount of the currently recorded data to be transmitted is directly subtracted from the preset fixed value to update the amount of data stored in the transmission path, and to avoid detecting the amount of data sent by the transmission buffer unit 330 to the link.
  • the storage depth of the entire transmission path (the maximum number of FDs and frames that can be stored) is TXBUF_DEPTH packets
  • the maximum length of the message data is MTU
  • the data volume threshold (BP_TH) of the transmit path configuration is BYTE_COUNTER max .
  • FIG. 4 and FIG. 5 are schematic diagrams showing the effect of the data volume backlog of the transmission path in the prior art and the embodiment of the present invention.
  • the transmission path sends the inverse to the TM only when the number of stored packets reaches the maximum value.
  • the pressure signal, the number of accumulated bytes in the transmission path may have a very large peak, which is because of the long packet length in a period of time; in the embodiment of the present invention, the total number of bytes in the transmission path can be limited to Within a certain range.
  • the maximum number of bytes that can be stored in the transmission path is that when the amount of data stored in the transmission path is just one byte lower than BP_TH, the TM sends a FD into the transmission path with the recording frame length of MTU, that is, the maximum stored byte.
  • the number is (BP_TH-1)+MTU, and the maximum blocking delay for subsequent packets entering the transmission path can be calculated as:
  • Latency max [(BP_TH-1)+MTU]*8bit/BW
  • BW is the link bandwidth
  • the blocking delay of the packet can be controlled within a certain range (less than Latencymax), ensuring that the packet does not introduce excessive delay due to blocking of the device at the current level.
  • a certain range less than Latencymax
  • the storage path has a storage depth of 10 packets and a link bandwidth of 2 Mbps.
  • the FTP download and upload services are performed simultaneously.
  • the device shown in FIG. 1 outputs a backpressure signal to the TM when the number of packets stored in the transmission path reaches 10, and the blocking delay of the subsequent ACK message entering the TM is:
  • Latency max [1518*(10+1)]*8/2Mbps ⁇ 67ms
  • Latency max [1000-1+1518]*8/2Mbps ⁇ 10ms
  • the embodiment of the present invention can significantly reduce the blocking delay of the ACK frame.
  • the device shown in FIG. 1 is used, and when the number of packets stored in the transmission path reaches 10, the back pressure signal is output to the TM.
  • the blocking delay for subsequent ACK messages entering the TM is:
  • Latency max [64*(10+1)]*8/2Mbps ⁇ 3ms
  • BP_TH is set to 1000 Byte
  • the calculation unit sends a back pressure signal to the TM module when the number of bytes stored in the transmission path is (BP_TH-1)+MTU.
  • two back pressure mechanisms are generally used at the same time, that is, when the FD number is greater than a preset number threshold, the FD_FIFO sends a back pressure signal to the TM module, and the counter stores data stored in the path.
  • Latency max [64*(10+1)]*8/2Mbps ⁇ 3ms
  • a data processing method for a transmission path includes:
  • the sending path receives and stores a frame descriptor FD of the first frame sent by the traffic management module, where the FD of the first frame includes a length value of the first frame.
  • the sending path determines, according to a length value of the first frame, that the sending path is stored.
  • the amount of data, the amount of data stored in the transmission path includes the amount of data of the frame corresponding to the stored FD and the amount of data of the stored frame;
  • the sending path determines whether the amount of data stored in the sending path is greater than a preset data amount threshold
  • step S620 includes:
  • the transmission path determines the sum of the amount of data stored before storing the FD of the first frame and the length value of the first frame as the amount of data stored in the transmission path.
  • the method further includes:
  • the transmission path subtracts the amount of data of the first data by the amount of data stored in the transmission path to update the amount of data stored in the transmission path.
  • the method further includes:
  • the sending path determines the number of FDs stored in the sending path
  • the transmitting path sends the back pressure signal to the TM module when it is determined that the number of stored FDs reaches a preset number threshold.
  • the FD of the first frame further includes a starting storage address of the first frame in the memory, and the method further includes:
  • the transmitting path acquires the first frame from the memory according to a length value of the first frame and a starting storage address of the first frame, and stores the first frame.
  • the embodiments of the present invention provide a data processing method and a transmission path of a transmission path, and control the blocking of packets newly entering the transmission path by controlling the amount of data of the to-be-sent packet that is accumulated in the transmission path.
  • the delay reduces the transmission delay of the message.
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) including computer usable program code.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明实施例提供了一种发送通路及发送通路的数据处理方法,用以控制发送通路存储的数据量,进而控制报文的阻塞延时,避免报文的发送延时过高的情况,该发送通路包括:FD存储模块,用于接收并存储流量管理TM模块发送的第一帧的FD,该第一帧的FD包括第一帧的长度值;计算模块,用于根据该FD存储模块中存储的第一帧的长度值确定该发送通路存储的数据量;判断模块,用于判断该计算模块确定的该发送通路存储的数据量是否大于预设的数据量门限;指示模块,在该判断模块判断该发送通路存储的数据量大于预设的数据量门限时,用于向该TM模块发送反压信号,该反压信号用于指示该TM模块停止向该发送通路发送第二帧的FD。

Description

一种发送通路及发送通路的数据处理方法 技术领域
本发明涉及网络设备技术领域,尤其涉及一种发送通路及发送通路的数据处理方法。
背景技术
传输控制协议(Transmission Control Protocol,TCP)是专门为了在不可靠的互联网络上提供一个可靠的端到端通信服务而设计的通信协议。基于TCP协议传输报文时,发送端发送报文的动作一般是由所接收的来自接收端的TCP确认(Acknowledgment,ACK)帧触发的,而如果没有收到接收端发来的ACK帧时一般不会发送报文(超时重传的情况除外)。一些基于TCP的应用,譬如文件传输协议(File Transfer Protocol,FTP)下载,如果客户端发出到服务器端的ACK帧在网络传输的过程中受到阻塞,ACK帧到达服务器端的延时就会加大,服务器端启动下一次报文发送的延时也就会相应加大,从而导致FTP下载效率下降。下面对发送端内部产生ACK帧发送延时的原因进行分析。
一般的网络设备中,ACK帧和包含正常业务数据的报文都由同一个先进先出(First Input First Output,FIFO)结构的发送通路发往链路,由图1可知,发送通路(Tx_path)主要包括帧描述符先进先出(FD_FIFO)单元、发送处理(Tx_Prcs)单元和发送缓存(TxBuffer)单元等。前级的流量管理(Traffic Management,TM)模块向发送通路调度输出帧描述符(Frame Descriptor,FD),FD记录了帧的缓存地址和帧长等信息,TM的调度具有一定突发和延时,为了吸收这个突发和延时,加上适配Tx_Prcs单元的处理,设计FD_FIFO来存放FD。Tx_Prcs根据FD中记录的帧的缓存地址信息从包缓存(Packet Buffer)中将对应的报文数据读出,其中,Packet Buffer为存储器,具体可以为共享 的随机存取存储器(Random Access Memory,RAM),从Packet Buffer读回来的报文数据会先缓存到TxBuffer中。当TxBuffer里有一个整帧或缓存数据量达到设定的字节发送门限时才启动发送,这样能够防止长包断流。
根据以上发送通路的结构分析可知,FD_FIFO缓存FD,TxBuffer缓存报文,这些缓存数据构成了新进入发送通路的FD的队头,新进入的FD所对应的报文只有在队头的所有报文发送完毕之后,才能够发送出去。如图1所示,报文a表示新进入发送通路的FD,a前面有若干报文d等待发送,而报文a是具有高优先级的ACK帧。若排在报文a前面的报文d中长包较多,尽管TM里通过严格优先级调度可以确保TM在无反压信号的情况下优先调度高优先级的ACK帧进入发送通路,但由于ACK帧的前面有多个长包待发送,会对ACK帧产生很大的阻塞延时;而若排在报文a前面的报文中短包较多,阻塞延时便会较低。另外,由于发送通路能够缓存的报文的个数是有限的,当发送通路内的FD_FIFO已经填满了报文时,TM会收到发送通路的反压信号,不能调度数据进入发送通路,直到发送通路完成发送一个报文到链路,撤销对TM的反压信号为止,这种情况下,ACK帧的阻塞延时会更加大。
基于以上分析,可以得知,由于发送通路内缓存的长包可能较多,即发送通路存储的数据量可能会很大,包括ACK帧在内的所有报文在发送通路内都可能受到较高的阻塞延时;而现有技术还没有提出一种解决方案,能够有效控制发送通路中存储的数据量,进而控制报文在发送通路中的阻塞延时。
发明内容
本发明实施例提供了一种发送通路及发送通路的数据处理方法,用以控制发送通路中存储的数据量,进而控制报文的阻塞延时。
第一方面,本发明实施例提供了一种发送通路,包括:
帧描述符FD存储模块,用于接收并存储流量管理TM模块发送的第一帧的帧描述符FD,所述第一帧的FD包括第一帧的长度值;
计算模块,用于根据所述FD存储模块中存储的第一帧的长度值确定所述发送通路存储的数据量,所述发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量;
判断模块,用于判断所述计算模块确定的所述发送通路存储的数据量是否大于预设的数据量门限;
指示模块,在所述判断模块判断所述发送通路存储的数据量大于预设的数据量门限时,用于向所述TM模块发送反压信号,所述反压信号用于指示所述TM模块停止向所述发送通路发送第二帧的FD。
结合第一方面,在第一种可能的实现方式中,所述计算模块具体用于:
将所述FD存储模块在存储所述第一帧的FD之前所述发送通路存储的数据量与所述FD存储模块中存储的第一帧的长度值之和,确定为所述发送通路存储的数据量。
结合第一方面或第一方面的第一种可能的实现方式,在第二种可能的实现方式中,所述发送通路还包括:
发送模块,用于向链路发送第一数据,所述第一数据为所述发送通路中存储的帧的数据;
所述计算模块,还用于将所述发送通路存储的数据量减去所述第一数据的数据量以更新所述发送通路存储的数据量。
结合第一方面,第一方面的第一种或第二种可能的实现方式,在第三种可能的实现方式中,
所述计算模块,还用于确定所述FD存储模块存储的FD的个数;
所述判断模块,还用于判断所述计算模块确定的FD的个数是否到达预设的个数门限;
所述指示模块,在所述判断模块判断所述计算模块确定的FD的个数达到预设的个数门限时,向所述TM模块发送所述反压信号。
结合第一方面,第一方面的第一种至第三种可能的实现方式中的任一种, 在第四种可能的实现方式中,所述第一帧的FD还包括第一帧在存储器中的起始存储地址,所述发送通路还包括:
帧存储模块,用于根据所述FD存储模块中存储的第一帧的长度值和所述第一帧的起始存储地址,从所述存储器中获取所述第一帧,并且存储所述第一帧。
第二方面,本发明实施例提供了一种发送通路的数据处理方法,包括:
所述发送通路接收并存储流量管理TM模块发送的第一帧的帧描述符FD,所述第一帧的FD包括第一帧的长度值;
所述发送通路根据所述第一帧的长度值确定所述发送通路存储的数据量,所述发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量;
所述发送通路判断所述发送通路存储的数据量是否大于预设的数据量门限;
在所述发送通路存储的数据量大于预设的数据量门限时,所述发送通路向所述TM模块发送反压信号,所述反压信号用于指示所述TM模块停止向所述发送通路发送第二帧的FD。
结合第二方面,在第一种可能的实现方式中,所述发送通路根据所述第一帧的长度值确定所述发送通路存储的数据量,包括:
所述发送通路将在存储第一帧的FD之前存储的数据量与所述第一帧的长度值之和,确定为所述发送通路存储的数据量。
结合第二方面或者第二方面的第一种可能的实现方式,在第二种可能的实现方式中,所述方法还包括:
所述发送通路向链路发送第一数据,所述第一数据为所述发送通路中存储的帧的数据;
所述发送通路将所述发送通路存储的数据量减去所述第一数据的数据量以更新所述发送通路存储的数据量。
结合第二方面,第二方面的第一种或第二种可能的实现方式,在第三种可能的实现方式中,所述方法还包括:
所述发送通路确定所述发送通路存储的FD的个数;
当确定存储的FD的个数达到预设的个数门限时,所述发送通路向所述TM模块发送所述反压信号。
结合第二方面,第二方面的第一种至第三种可能的实现方式中的任一种,在第四种可能的实现方式中,所述第一帧的FD还包括第一帧在存储器中的起始存储地址,所述方法还包括:
所述发送通路根据所述第一帧的长度值和所述第一帧的起始存储地址,从所述存储器中获取所述第一帧,并且存储所述第一帧。
本发明实施例中,在发送通路存储的数据量大于预设的数据量门限时,该发送通路向TM模块发送反压信号,使得该发送通路存储的数据量大小控制在一定范围内,相应地,新进入该发送通路的报文的阻塞延时也得到了控制。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作一简单地介绍,显而易见地,下面描述中的附图是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。
图1是现有技术中的数据发送装置的结构示意图;
图2是本发明实施例提供的发送通路的结构示意图;
图3是本发明另一实施例提供的数据发送装置的结构示意图;
图4是现有技术中的发送通路内积压的数据量的示意图;
图5是本发明实施例提供的发送通路内积压的数据量的示意图;
图6是本发明实施例提供的发送通路的数据处理方法的流程示意图。
具体实施方式
本发明实施例提供了一种发送通路及发送通路的数据处理方法,用以控制发送通路对报文的阻塞延时,避免报文的阻塞延时过高的情况。
TM模块向发送通路模块发送第一帧的FD后,发送通路将第一帧的FD存储在FD队列中,直到第一帧的FD进入队首位置,发送通路再取出第一帧的FD,根据取出的FD的信息从存储器中获取到第一帧,存储第一帧至帧队列,最后发送通路按照帧队列的顺序将各帧依次发送出去。本发明实施例中,发送通路存储的数据量,指的是发送通路将要发送出的各帧数据量之和,由于发送通路中既存储有FD又存储有帧,FD用于获取到对应的帧并发送出去,因此,发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量。
参见图2,本发明实施例提供的发送通路包括:
FD存储模块210,用于接收并存储流量管理TM模块发送的第一帧的FD,所述第一帧的FD包括第一帧的长度值;
计算模块220,用于根据所述FD存储模块中存储的第一帧的长度值确定所述发送通路存储的数据量,所述发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量;
判断模块230,用于判断所述计算模块确定的所述发送通路存储的数据量是否大于预设的数据量门限;
指示模块240,在所述判断模块判断所述发送通路存储的数据量大于预设的数据量门限时,用于向所述TM模块发送反压信号,所述反压信号用于指示所述TM模块停止向所述发送通路发送第二帧的FD。
其中,所述计算模块220具体用于:
将所述FD存储模块在存储所述第一帧的FD之前所述发送通路存储的数据量与所述FD存储模块中存储的第一帧的长度值之和,确定为所述发送通路 存储的数据量。
可选地,所述计算模块,还用于确定所述FD存储模块存储的FD的个数;
所述判断模块,还用于判断所述计算模块确定的FD的个数是否到达预设的个数门限;
所述指示模块,在所述判断模块判断所述计算模块确定的FD的个数达到预设的个数门限时,向所述TM模块发送所述反压信号。
FD在发送通路中是以个为单位进行存储,发送通路能够存储的FD的个数是有限的,当发送通路存储FD的个数达到最大存储个数时,即向TM模块发送反压信号。而本发明实施例中,通过对当前发送通路存储的FD的个数进行监测,在FD的个数达到预设的个数门限时,即向所述TM模块发送所述反压信号,个数门限可设置为任一不大于最大存储个数的值,达到对发送通路的数据处理进行更加灵活的控制的目的。需要注意的是,将允许进入发送通路的FD个数限制在较低值尽管能够降低阻塞延时,但是会对发送通路的抗突发性能会造成一定的影响,即连续多个FD对应的报文均为短包时,可能导致发送通路向网络发送报文的速度大于TM的调度速度,降低发送通路的工作效率。
可选地,所述第一帧的FD还包括第一帧在存储器中的起始存储地址,所述发送通路还包括:
帧存储模块,用于根据所述FD存储模块中存储的第一帧的长度值和所述第一帧的起始存储地址,从所述存储器中获取所述第一帧,并且存储所述第一帧。
可选地,所述发送通路还包括:
发送模块,用于向链路发送第一数据,所述第一数据为所述发送通路中存储的帧的数据;
所述计算模块,还用于将所述发送通路存储的数据量减去所述第一数据的数据量以更新所述发送通路存储的数据量。
即是说,发送通路保持记录当前存储的数据量,当有FD进入时,增加记录的数据量,当有报文发出时,减少记录的数据量,确保发送通路记录的存储的数据量的准确性。
进一步地,所述发送模块发送的第一数据,为发送通路帧存储模块队首的帧中的2个字节或16个字节的数据。
下面结合实际发送通路硬件结构,对本发明实施例进行进一步说明。参见图3,本发明实施例提供的发送通路具体包括:
帧描述符先进先出单元310,用于接收并存储所述TM模块发送的第一帧的FD,根据所述FD获取所述第一帧的长度值;
所述发送处理单元320,用于根据从所述帧描述符先进先出单元310取出第一帧的FD,解析出第一帧在存储器中的起始存储地址,根据第一帧的起始存储地址和第一帧的长度值从所述存储器中获取所述第一帧,并将所述第一帧写入发送缓存单元330;
所述发送缓存单元330,用于将所述第一帧发送至所述数据发送至链路;
所述计数单元(BYTE_COUNTER)340,用于将记录的发送通路存储的数据量加上从帧描述符先进先出单元310获取的第一帧的长度值,以及,每当检测到所述发送缓存单元330发送一次报文中的数据至链路时,将记录的发送通路存储的数据量减去所述发送缓存单元330发送的数据量;判断记录的发送通路存储的数据量是否大于预设的数据量门限,当记录的发送通路存储的数据量大于预设的数据量门限时,向所述TM模块发送反压信号,用以指示所述TM模块停止发送FD。
其中,发送缓存单元330每次发送至链路的数据量为可以预先设置的固定值,一般为2个字节或16个字节,则每当计数单元检测到发送缓存单元330发送一次数据至链路时,直接将当前记录的待发送数据的数据量与预先设置的固定值相减,以更新发送通路存储的数据量,并避免去检测发送缓存单元330发送至链路的数据量。
假设整个发送通路的存储深度(能够存储的FD以及帧的最大个数)为TXBUF_DEPTH个报文,报文数据最大长度为MTU,计数单元可统计的发送通路存储的数据量BYTE_CONTER的最大值为BYTE_CONTERmax=TXBUF_DEPTH*MTU,发送通路配置的数据量门限(BP_TH)为BYTE_COUNTERmax。当计数单元发现BYTE_CONTER大于BP_TH时,向前级的TM模块输出反压信号,当BYTE_CONTER不大于BP_TH时,撤销对前级TM的反压信号,TM开始重新调度出新的FD发送给发送通路。
图4和图5分别是现有技术与本发明实施例的发送通路的数据量积压状况的效果示意图,现有技术中,发送通路仅在存储的报文个数达到最大值时向TM发送反压信号,发送通路内积压字节数可能会出现非常大的峰值,这是因为一段时间内长包较多的原因;而本发明实施例中,可以将发送通路内的总字节数限制在一定的范围之内。
进一步地,发送通路内可以存储的最大字节数是在发送通路存储的数据量刚低于BP_TH一个字节时,TM发送一个记录帧长为MTU的FD进入发送通路,即最大存储的字节数是(BP_TH-1)+MTU,则对后续进入发送通路的报文的最大阻塞延时可以计算为:
Latencymax=[(BP_TH-1)+MTU]*8bit/BW
其中,BW为链路带宽。
因此,通过设置BP_TH值的大小,可以对报文的阻塞延时控制在一定的范围之内(少于Latencymax),确保报文不会因为本级设备的阻塞而引入过大的延时,可以提高TCP类的应用的传输效率和性能,提升服务质量(Quality of Service,QoS)。
下面结合具体数据对本发明实施例的效果进行进一步说明:
假设发送通路的存储深度为10个报文,链路带宽为2Mbps,FTP下载和上传业务同时进行。
假如FTP业务的ACK帧和上传的数据报文的帧长都为1518Byte,采用 图1所示装置,则当发送通路存储报文数达到10个时,向TM输出反压信号,此时对后续进入TM的ACK报文的阻塞延时为:
Latencymax=[1518*(10+1)]*8/2Mbps≈67ms
而采用图3所示装置,设置BP_TH为1000Byte,那么根据前面的分析,最坏的情况是发送通路存储的字节数为(BP_TH-1)+MTU,则最大的延时为:
Latencymax=[1000-1+1518]*8/2Mbps≈10ms
因此,在ACK帧队头的长包较多的情况下,本发明实施例能够明显降低ACK帧的阻塞延时。
假如下载业务回应的ACK帧和上传的数据报文的帧长都为64Byte,采用图1所示装置,则当发送通路存储的报文数达到10个时,向TM输出反压信号,此时对后续进入TM的ACK报文的阻塞延时为:
Latencymax=[64*(10+1)]*8/2Mbps≈3ms
而采用图3所示装置,设置BP_TH为1000Byte,那么根据前面的分析,计算单元在发送通路存储的字节数为(BP_TH-1)+MTU时向TM模块发送反压信号。但是,考虑到实际使用中通常同时使用两种反压机制,即FD_FIFO在FD个数大于预设的个数门限时,向所述TM模块发送反压信号,并且,计数器在发送通路存储的数据量超过BP_TH时,向所述TM模块发送反压信号;由于1000Byte的报文数据对应于1000Byte/64Byte≈15个短包,而当FD个数达到10条时FD_FIFO已经发出反压信号,因此,发送通路缓存的字节数不会达到(BP_TH-1)+MTU,最大延时仍为:
Latencymax=[64*(10+1)]*8/2Mbps≈3ms
同时采用这两种反压机制,既可以保证短包时的突发性又可以保证不会在发送通路引入过大的阻塞延时。
参见图6,本发明实施例提供的一种发送通路的数据处理方法包括:
S610:所述发送通路接收并存储流量管理TM模块发送的第一帧的帧描述符FD,所述第一帧的FD包括第一帧的长度值;
S620:所述发送通路根据所述第一帧的长度值确定所述发送通路存储的 数据量,所述发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量;
S630:所述发送通路判断所述发送通路存储的数据量是否大于预设的数据量门限;
S640:在所述发送通路存储的数据量大于预设的数据量门限时,所述发送通路向所述TM模块发送反压信号,所述反压信号用于指示所述TM模块停止向所述发送通路发送第二帧的FD。
具体地,步骤S620包括:
所述发送通路将在存储第一帧的FD之前存储的数据量与所述第一帧的长度值之和,确定为所述发送通路存储的数据量。
可选地,该方法还包括:
所述发送通路向链路发送第一数据,所述第一数据为所述发送通路中存储的帧的数据;
所述发送通路将所述发送通路存储的数据量减去所述第一数据的数据量以更新所述发送通路存储的数据量。
可选地,该方法还包括:
所述发送通路确定所述发送通路存储的FD的个数;
当确定存储的FD的个数达到预设的个数门限时,所述发送通路向所述TM模块发送所述反压信号。
可选地,所述第一帧的FD还包括第一帧在存储器中的起始存储地址,该方法还包括:
所述发送通路根据所述第一帧的长度值和所述第一帧的起始存储地址,从所述存储器中获取所述第一帧,并且存储所述第一帧。
综上所述,本发明实施例提供了一种发送通路的数据处理方法及发送通路,通过控制发送通路内积压的待发送报文的数据量,实现了控制新进入发送通路的报文的阻塞延时,降低了报文的发送延时。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中,使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品,该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
本领域内的技术人员应明白,本发明的实施例可提供为方法、系统、或计算机程序产品。因此,本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且,本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器和光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器,使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上,使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理,从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
显然,本领域的技术人员可以对本发明进行各种改动和变型而不脱离本发明的精神和范围。这样,倘若本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内,则本发明也意图包含这些改动和变型在内。

Claims (10)

  1. 一种发送通路的数据处理方法,其特征在于,包括:
    所述发送通路接收并存储流量管理TM模块发送的第一帧的帧描述符FD,所述第一帧的FD包括第一帧的长度值;
    所述发送通路根据所述第一帧的长度值确定所述发送通路存储的数据量,所述发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量;
    所述发送通路判断所述发送通路存储的数据量是否大于预设的数据量门限;
    在所述发送通路存储的数据量大于预设的数据量门限时,所述发送通路向所述TM模块发送反压信号,所述反压信号用于指示所述TM模块停止向所述发送通路发送第二帧的FD。
  2. 如权利要求1所述的方法,其特征在于,所述发送通路根据所述第一帧的长度值确定所述发送通路存储的数据量,包括:
    所述发送通路将在存储第一帧的FD之前存储的数据量与所述第一帧的长度值之和,确定为所述发送通路存储的数据量。
  3. 如权利要求1或2所述的方法,其特征在于,还包括:
    所述发送通路向链路发送第一数据,所述第一数据为所述发送通路中存储的帧的数据;
    所述发送通路将所述发送通路存储的数据量减去所述第一数据的数据量以更新所述发送通路存储的数据量。
  4. 如权利要求1-3任一权项所述的方法,其特征在于,还包括:
    所述发送通路确定所述发送通路存储的FD的个数;
    当确定存储的FD的个数达到预设的个数门限时,所述发送通路向所述TM模块发送所述反压信号。
  5. 如权利要求1-4任一权项所述的方法,其特征在于,所述第一帧的FD 还包括第一帧在存储器中的起始存储地址,所述方法还包括:
    所述发送通路根据所述第一帧的长度值和所述第一帧的起始存储地址,从所述存储器中获取所述第一帧,并且存储所述第一帧。
  6. 一种发送通路,其特征在于,包括:
    帧描述符FD存储模块,用于接收并存储流量管理TM模块发送的第一帧的帧描述符FD,所述第一帧的FD包括第一帧的长度值;
    计算模块,用于根据所述FD存储模块中存储的第一帧的长度值确定所述发送通路存储的数据量,所述发送通路存储的数据量包括存储的FD所对应的帧的数据量和存储的帧的数据量;
    判断模块,用于判断所述计算模块确定的所述发送通路存储的数据量是否大于预设的数据量门限;
    指示模块,在所述判断模块判断所述发送通路存储的数据量大于预设的数据量门限时,用于向所述TM模块发送反压信号,所述反压信号用于指示所述TM模块停止向所述发送通路发送第二帧的FD。
  7. 如权利要求6所述的发送通路,其特征在于,所述计算模块具体用于:
    将所述FD存储模块在存储所述第一帧的FD之前所述发送通路存储的数据量与所述FD存储模块中存储的第一帧的长度值之和,确定为所述发送通路存储的数据量。
  8. 如权利要求6或7所述的发送通路,其特征在于,还包括:
    发送模块,用于向链路发送第一数据,所述第一数据为所述发送通路中存储的帧的数据;
    所述计算模块,还用于将所述发送通路存储的数据量减去所述第一数据的数据量以更新所述发送通路存储的数据量。
  9. 如权利要求6-8任一权项所述的发送通路,其特征在于,
    所述计算模块,还用于确定所述FD存储模块存储的FD的个数;
    所述判断模块,还用于判断所述计算模块确定的FD的个数是否到达预设的 个数门限;
    所述指示模块,在所述判断模块判断所述计算模块确定的FD的个数达到预设的个数门限时,向所述TM模块发送所述反压信号。
  10. 如权利要求6-9任一权项所述的发送通路,其特征在于,所述第一帧的FD还包括第一帧在存储器中的起始存储地址,所述发送通路还包括:
    帧存储模块,用于根据所述FD存储模块中存储的第一帧的长度值和所述第一帧的起始存储地址,从所述存储器中获取所述第一帧,并且存储所述第一帧。
PCT/CN2015/081284 2014-06-13 2015-06-11 一种发送通路及发送通路的数据处理方法 WO2015188770A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410265683.3A CN104052676B (zh) 2014-06-13 2014-06-13 一种发送通路装置及发送通路的数据处理方法
CN201410265683.3 2014-06-13

Publications (1)

Publication Number Publication Date
WO2015188770A1 true WO2015188770A1 (zh) 2015-12-17

Family

ID=51505058

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/081284 WO2015188770A1 (zh) 2014-06-13 2015-06-11 一种发送通路及发送通路的数据处理方法

Country Status (4)

Country Link
US (1) US9813349B2 (zh)
EP (1) EP2958283B1 (zh)
CN (1) CN104052676B (zh)
WO (1) WO2015188770A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116192759A (zh) * 2022-12-13 2023-05-30 网易(杭州)网络有限公司 数据传输量的调整方法、装置、电子设备及存储介质

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104052676B (zh) 2014-06-13 2017-12-05 华为技术有限公司 一种发送通路装置及发送通路的数据处理方法
ES2711354T3 (es) * 2014-09-24 2019-05-03 Alcatel Lucent Aparatos, métodos y programas informáticos para un sistema de comunicación móvil que comprende un transceptor de estación base y un transceptor móvil
CN106357554A (zh) * 2015-07-13 2017-01-25 中兴通讯股份有限公司 一种设备内部处理器网口收包方法及装置
CN106878171B (zh) * 2016-12-30 2020-03-13 晶赞广告(上海)有限公司 一种用于多数据源的流式数据处理方法及装置
JP7251075B2 (ja) * 2018-09-03 2023-04-04 株式会社オートネットワーク技術研究所 中継装置、中継方法及びコンピュータプログラム

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1488105A (zh) * 2001-01-31 2004-04-07 �Ҵ���˾ 控制数据处理系统间经由存储器的数据流的方法和装置
CN101013935A (zh) * 2007-01-31 2007-08-08 华为技术有限公司 数据传输的方法和接收装置及系统
CN101984594A (zh) * 2010-10-25 2011-03-09 华为技术有限公司 交换网流量控制方法和装置
CN104052676A (zh) * 2014-06-13 2014-09-17 华为技术有限公司 一种发送通路及发送通路的数据处理方法

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6405258B1 (en) 1999-05-05 2002-06-11 Advanced Micro Devices Inc. Method and apparatus for controlling the flow of data frames through a network switch on a port-by-port basis
US7586909B1 (en) * 2002-03-06 2009-09-08 Agere Systems Inc. Striping algorithm for switching fabric
US7948880B2 (en) * 2004-10-29 2011-05-24 Broadcom Corporation Adaptive dynamic thresholding mechanism for link level flow control scheme
JP2006325054A (ja) * 2005-05-20 2006-11-30 Seiko Epson Corp Tcp/ip受信処理回路及びそれを具備する半導体集積回路
US7571247B2 (en) 2005-12-12 2009-08-04 International Business Machines Corporation Efficient send socket call handling by a transport layer
US8037213B2 (en) * 2007-05-30 2011-10-11 International Business Machines Corporation Replenishing data descriptors in a DMA injection FIFO buffer
US8121038B2 (en) * 2007-08-21 2012-02-21 Cisco Technology, Inc. Backward congestion notification
US7773519B2 (en) * 2008-01-10 2010-08-10 Nuova Systems, Inc. Method and system to manage network traffic congestion
US8910168B2 (en) * 2009-04-27 2014-12-09 Lsi Corporation Task backpressure and deletion in a multi-flow network processor architecture
US8737417B2 (en) * 2010-11-12 2014-05-27 Alcatel Lucent Lock-less and zero copy messaging scheme for telecommunication network applications
US9106545B2 (en) * 2011-12-19 2015-08-11 International Business Machines Corporation Hierarchical occupancy-based congestion management

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1488105A (zh) * 2001-01-31 2004-04-07 �Ҵ���˾ 控制数据处理系统间经由存储器的数据流的方法和装置
CN101013935A (zh) * 2007-01-31 2007-08-08 华为技术有限公司 数据传输的方法和接收装置及系统
CN101984594A (zh) * 2010-10-25 2011-03-09 华为技术有限公司 交换网流量控制方法和装置
CN104052676A (zh) * 2014-06-13 2014-09-17 华为技术有限公司 一种发送通路及发送通路的数据处理方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116192759A (zh) * 2022-12-13 2023-05-30 网易(杭州)网络有限公司 数据传输量的调整方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN104052676A (zh) 2014-09-17
EP2958283B1 (en) 2018-11-21
US20150365332A1 (en) 2015-12-17
EP2958283A1 (en) 2015-12-23
CN104052676B (zh) 2017-12-05
US9813349B2 (en) 2017-11-07

Similar Documents

Publication Publication Date Title
WO2015188770A1 (zh) 一种发送通路及发送通路的数据处理方法
US11277349B2 (en) Minimal buffer network arbiter
US10075382B2 (en) Communication device, relay device, and communication method for a plurality of packets
US20230171301A1 (en) Monitoring Network Conditions
US9930097B2 (en) Transport accelerator systems and methods
US20090268747A1 (en) Communication apparatus
US7643420B2 (en) Method and system for transmission control protocol (TCP) traffic smoothing
CN112104562B (zh) 拥塞控制方法及装置、通信网络、计算机存储介质
US10606492B2 (en) Detecting and handling solicited IO traffic microbursts in a fibre channel storage area network
US20120054362A1 (en) Mechanism for autotuning mass data transfer from a sender to a receiver over parallel connections
WO2015142913A1 (en) Transport accelerator implementing request manager and connection manager functionality
WO2017161967A1 (zh) 包每秒流量监管方法、装置和计算机存储介质
US7355976B2 (en) Method and apparatus for providing retry control, buffer sizing and management
WO2018076641A1 (zh) 一种减少时延的方法、装置及存储介质
US20120158957A1 (en) Relay apparatus and communication method
JP4698645B2 (ja) フロー制御装置およびフロー制御方法
JP4146375B2 (ja) 処理制御装置、処理制御方法及び処理制御プログラム
JP4930275B2 (ja) 通信システム、通信方法、送信機、受信機、レート計算方法およびプログラム
JP4382830B2 (ja) パケット転送装置
CN115514710A (zh) 一种基于自适应滑动窗的弱连接流量管控方法
US20060221827A1 (en) Tcp implementation with message-count interface
WO2023195171A1 (ja) 通信システム、中継機、通信方法、及び非一時的なコンピュータ可読媒体
JP2012100165A (ja) データ伝送方式、及び、複数拠点データ配信方式
JP6766817B2 (ja) データ通信装置、データ通信制御方法及びプログラム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15806613

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15806613

Country of ref document: EP

Kind code of ref document: A1