WO2012119414A1

WO2012119414A1 - Method and device for controlling traffic of switching network

Info

Publication number: WO2012119414A1
Application number: PCT/CN2011/078853
Authority: WO
Inventors: 雷春; 项能武
Original assignee: 华为技术有限公司
Priority date: 2011-08-24
Filing date: 2011-08-24
Publication date: 2012-09-13
Also published as: CN102356609A

Abstract

Provided are a method and device for controlling the traffic of a switching network, relating to the technical field of communications, reducing the packet loss ratio, and improving the efficiency of the switching network. The method includes: obtaining the rate at which at least one destination end receives a data packet; and sending the data packet corresponding to the destination end to the destination end, wherein the rate at which the data packet is sent is lower than or equal to the rate at which the destination end receives the data packet. The device includes: a processor for obtaining the rate at which at least one destination end receives a data packet; a sender for sending the data packet corresponding to the destination end to the destination end, wherein the rate at which the data packet is sent is lower than or equal to the rate at which the destination end receives the data packet.

Description

Traffic control method and device for switching network

The present invention relates to the field of communications technologies, and in particular, to a flow control method and apparatus for a switching network. BACKGROUND In a typical switching network system composed of a plurality of line cards and a switching network board, data exchange is performed between a line card and a line card. For example, as shown in FIG. 1, in the first line card 11, The Ethernet packet 21 of the second line card 12 is switched to the second line card 12 through the switching network board 3. The Ethernet packet 21 whose destination end is the third line card 13 is exchanged by the switching network board 3 and reaches the third line card 13. The first line card 11 has a rate of 10 Gbps, the second line card 12 has a rate of 1 Gbps, and the third line card 13 has a rate of 10 Gbps. When the Ethernet packet 21 in the first line card 11 is sent to the second line card 12 at a rate of 10 Gbps. If the egress buffer 31 of the switching network board 3 connected to the second line card 12 is not large enough, all the Ethernet packets 21 sent from the first line card 11 to the second line card 12 cannot be stored, and the switching network board 3 is connected to the second line card 12 Packet loss occurs at the outbound cache 31.

In order to solve the above problem, the prior art sets the egress buffer 31 of the switching network board 3 connected to the second line card 12 to a non-drop mode. When the input buffer 31 has an input traffic greater than the output traffic, after a period of time, the egress buffer 31, a back pressure occurs, and the back pressure is transmitted to the switching network board 3 to connect to the ingress buffer 33 of the first line card 11, which is expressed as the first line when the first line card 11 receives the pause frame of the switching network board 3. The card 11 stops transmitting the Ethernet packet 21, that is, the Ethernet packet 21 of the first line card 11 is no longer sent to the outgoing buffer 31. After the first line card 11 receives the pause frame, since the buffer queue of the first line card 11 is a single queue, all the Ethernet packets 21 are blocked, even if the switching network board 3 is connected to the outbound buffer of the third line card 13. 32 is idle, because the Ethernet packet 21 whose destination end is the second line card 12 in the inbound buffer 33 blocks the subsequent Ethernet packet 21 whose destination end is the third line card 13, so that it cannot be scheduled, Low data transmission efficiency. SUMMARY OF THE INVENTION The technical problem to be solved by the embodiments of the present invention is to provide a method and an apparatus for controlling traffic of a switching network, which can reduce packet loss rate and improve data transmission efficiency.

In one aspect, an embodiment of the present invention provides a method for controlling a traffic of a switching network, including: obtaining a rate at which at least one destination receives a data packet;

Sending a data packet corresponding to the destination end to the destination end, where a rate at which the data packet is sent is less than or equal to a rate at which the destination end receives the data packet.

In another aspect, the embodiment of the present invention further provides a traffic control device for a switching network, including: a processor, configured to obtain a rate at which at least one destination receives a data packet;

And a transmitter, configured to send a data packet corresponding to the destination end to the destination end, where a rate at which the data packet is sent is less than or equal to a rate at which the destination end receives the data packet.

After the foregoing technical solution, the flow control method and apparatus for the switching network provided by the embodiment of the present invention reduces the burst data packet received by the same destination end by controlling the data packet sending rate to be less than or equal to the data receiving rate of the destination end. At the same time, the transmission of the sender data packet does not stop, thereby reducing the packet loss rate and improving the data transmission efficiency. BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, and obviously, in the following description The drawings are only some of the embodiments of the present invention, and other drawings may be obtained from those skilled in the art without departing from the drawings.

1 is a schematic diagram of a typical switching network system in the prior art;

2 is a flowchart of a method for controlling traffic of a switching network according to an embodiment of the present invention; FIG. 3 is a schematic diagram of a switching network system according to an embodiment of the present disclosure;

4 is a schematic diagram of a switching network system applied to a base station controller according to an embodiment of the present invention; FIG. 5 is a schematic diagram of an apparatus for controlling a traffic of a switching network according to an embodiment of the present invention;

Figure 6 is a schematic diagram of the transmitter of Figure 5. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The technical solutions in the embodiments of the present invention will be clearly and completely described in the following with reference to the accompanying drawings.

It should be understood that the described embodiments are only a part of the embodiments of the invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

Embodiment 1

The embodiment of the present invention provides a traffic control method for a switching network. As shown in FIG. 2, the method is a flow control performed by a data packet sending end of a switching network. The method specifically includes:

Step 101: Obtain a rate at which at least one destination end receives data.

This step specifically includes:

Step 102: Send a data packet corresponding to the destination end to the destination end, where a rate at which the data packet is sent is less than or equal to a rate at which the destination end receives the data packet.

The data packet is buffered in the data stream corresponding to the destination end according to the rate at which the destination end of the data packet receives the data, and the data packet in the data stream is sent at a rate less than or equal to the rate at which the destination end corresponding to the data packet receives the data. Optionally, when the data packet is cached in the data stream corresponding to the destination end, the preset priority may be used, so that the data stream can send the data packet according to the priority.

The total transmission rate of the data packet is less than or equal to the transmission data packet rate of the transmitting end, so as to avoid the rate of receiving the data packet by controlling the transmitting end to be less than or equal to the rate at which the destination end receives the data. There is a case where the destination end is lost due to insufficient buffering, and since the transmission of the data packet at the transmitting end is not stopped, the data transmission efficiency is also improved.

Embodiment 2

As shown in FIG. 3, the technical solution of the present invention will be described below by taking a switching system composed of a plurality of line cards and a switching network board 3 as an example.

The data packet transmitted in this embodiment is the Ethernet packet 21. It is assumed that the first line card 11 is a transmitting end, and the rate at which the Ethernet packet 21 is transmitted is 10 Gbps; the second line card 12 and the third line card 13 are destination ends, and the receiving rate of the second line card 12 is 1 Gbps, and the receiving of the third line card 13 is At a rate of 10 Gbps, the first line card 11 transmits the Ethernet packets 21 to the second line card 12 and the third line card 13, respectively.

The specific data interaction is:

The first line card 11 first obtains the rate of receiving the Ethernet packet of the second line card 12 and the third line card 13; and then identifies the destination information contained in the destination address of the Ethernet packet 21 that the first line card 11 needs to transmit, The Ethernet packet 21 of the first line card 11 with the destination end being the second line card 12 can be cached in the first buffer queue 14 as a data stream; optionally, the ethernet with the destination line being the second line card 12 When the network packet 21 is cached in the first cache queue 14, the cache may be cached in a preset priority order. Similarly, for the Ethernet packet 21 of the first line card 11 whose destination end is the third line card 13 is cached in the second buffer queue 15 as another data stream according to a preset priority, optionally, in the purpose When the Ethernet packet 21 of the third line card 13 is buffered in the second buffer queue 15, the buffering may be performed in a preset priority order. Considering that the cache of the switching network board is small, the first cache queue 14 and the second buffer queue 15 described above can be implemented on the first line card 11.

The cache queue is then scheduled, and the Ethernet packets 21 in the first cache queue 14 are sent to the second line card 12 in order of priority, and the rate at which the first buffer queue 14 transmits the Ethernet packet 21 is less than or equal to the second line card 12 Receiving the rate of the Ethernet packet 21, for example, setting the rate of transmitting the Ethernet packet 21 of the first buffer queue 14 to 1 Gbps; and transmitting the Ethernet packet 21 in the second buffer queue 15 to the third line in order of priority Card 13, and causing the second buffer queue 1 to transmit the Ethernet packet 21 at a rate less than or equal to the rate at which the third line card 13 receives the Ethernet packet 21, for example, The rate at which the Ethernet packet 21 of the queue 15 is transmitted is set to 10 Gbps. The rate of the total transmission Ethernet packet 21 of the first buffer queue 14 and the second buffer queue 15 is set to the transmission rate of the first line card 11 by 10 Gbps, so as to prevent the transmission rate of the Ethernet packet 21 from being greater than the transmission of the first line card 11. The rate is lost.

The scheduling of the cache queue may be performed by using a polling manner. For example, the Ethernet packet 21 in the first buffer queue 14 is sent in the first time period, and the Ethernet packet 21 in the second buffer queue 15 is sent in the second time. The Ethernet packet 21 in the first buffer queue 14 is sent again in three periods, and thus alternates.

The traffic control method of the switching network provided by the embodiment of the present invention is compared with the prior art. When the first line card 11 has a continuous destination end, the Ethernet packet 21 of the second line card 12 needs to be sent, because the first cache is used. The rate at which the queue 14 sends the Ethernet packet 21 is controlled, so that the first line card 11 transmits the Ethernet packet 21 to the second line card 12 at a rate of 1 Gbps, and does not connect the outgoing line buffer of the second line card 12 because the switching network board 3 is connected. 31 is not large enough to cause congestion and packet loss, and since there are two columns of buffer queues for transmitting the Ethernet packet 21, the second line card 12 and the third line card 13 can receive the Ethernet packet 21 sent by the first line card 11 relatively evenly, The third line card 13 may be idle for a long time and the Ethernet packet 21 of the first line card 11 whose destination end is the third line card 13 cannot be transmitted, thereby improving the efficiency of transmitting the Ethernet packet.

Embodiment 3

As shown in FIG. 4, a switching network system applied to a base station controller includes two interface boards 16, a switching network board 3, and ten processing boards 17, and the processing capacity of the interface board 16 is 10 Gbps, and the processing board 17 The processing capacity is 1 Gbps, the switching capacity of the switching board 3 is greater than 40 Gbps, and the processing board 17 is used as a resource pool, and the total processing capacity is 10 Gbps. The data transmitted between the interface board 16 and the processing board 17 is a message, and the interface board 16 as a transmitting end transmits a packet as a data packet to the processing board 17 as a destination.

The specific data interaction process is:

First, the rate at which the processing board 17 receives the message, that is, the processing capability of the processing board 17, is obtained, which is 1 Gbps;

Then identifying the destination information contained in the destination address of the packet that the interface board 16 needs to send, according to According to the destination information, the packets in the two interface boards 16 are respectively buffered in the Double Rate Synchronous Dynamic Random Access Memory (DDR) according to the destination end, and the corresponding destinations are formed according to the preset priority. A plurality of cache queues 18 of the board 17, each of the processing boards 17 corresponding to a column or sequence of cache queues 18 formed according to a preset priority, and a column or sequence of cache queues 18 corresponding to each processing board 17 as a data stream, so that each strip The data stream is sent according to the priority, and the transmission rate of each data stream is less than or equal to the processing capability of the processing board 17 according to the processing capability of the processing board 17, preferably 0.5 Gbps;

Then, a Field-Programmable Gate Array (FPGA) schedules the buffer queues 18 in the interface board 16, and transmits the packets to the processing board 17 at a rate of 0.5 Gbps in order of priority. The total rate of the packets sent by the interface board 16 is limited according to the processing capability of the interface board. For example, the rate of sending packets is limited to 10 Gbps to avoid the packet transmission rate being greater than the transmission rate of the interface board 16. Lose the package.

The scheduling of the buffer queue 18 in each data stream or each data stream may be performed in a polling manner, and each data stream or each buffer queue 18 is sequentially transmitted.

The flow control method of the switching network provided by the embodiment of the present invention is compared with the prior art. When the transmission rate of each data stream is 0.5 Gbps, when two interface boards 16 simultaneously send a message to a processing board 17, If the data flow is scheduled by polling, the processing board 17 is sufficient to process the packets sent by the two interface boards 16 at the same time, and no traffic burst occurs, because the outgoing port buffer of the interface between the switching network board 3 and the processing board 17 31 is less, it is not possible to cache the packets in the burst time and thus congest the packet loss situation.

It should be noted that the sending rate of each data stream can be set according to the situation. For example, when one of the two interface boards 16 is broken and cannot send a message, the sending rate of each data stream is set to 1 Gbps. Improve the efficiency of the switching network.

When the transmission rate of each data stream is set to 1 Gbps, when two interface boards 16 simultaneously send a message to one processing board 17, the transmission rate of the data stream corresponding to each processing board 17 is limited, which is reduced. The burst message received by the same processing board 17 reduces the packet loss rate. The interface board 16 can receive the packets sent by the interface board 16 on a relatively even average basis, and the interface board 16 does not stop sending the packets, thereby improving the data. The efficiency of the transmission.

Embodiment 4

The embodiment of the present invention further provides an apparatus for using the flow control method of the above switching network. As shown in FIG. 5, the apparatus includes: a processor 51 and a transmitter 52.

The processor 51 is configured to obtain a rate at which the at least one destination end receives the data packet, and the sender 52 is configured to send, to the destination end, the data packet corresponding to the destination end, where the rate of sending the data packet is less than or equal to The destination receives the rate of the data packet.

Further, as shown in FIG. 6, the transmitter 52 includes: a cache module 521 and a sending module 522. The buffering module 521 is configured to buffer the data packet in a data stream corresponding to the destination end according to a rate at which the destination end of the data packet is received, and the sending module 522 is configured to send the data packet in the data stream, where And sending a data packet in the data stream at a rate less than or equal to a rate at which the destination end receives the data packet.

Further, the destination end has multiple, and the transmitter 52 is further configured to enable the total rate of sending the data packet to be less than or equal to the rate of sending the data packet of the transmitting end.

Further, the destination end has multiple, and the sending module 522 is specifically configured to sequentially send data in a data stream corresponding to each destination end, where a rate of data packets in the sending data stream is less than or equal to the receiving of each destination end. The rate of the packet.

Further, the cache module 521 is specifically configured to cache the data packet in a data stream corresponding to the destination end according to a preset priority according to the rate at which the destination end of the data packet receives the data packet.

The data rate of the data packet transmitted by the transmitter 52 is less than or equal to the data receiving rate of the destination end, and the data stream of the corresponding destination end of the buffer module 521 is less than or equal to the transmitting end. The data packet transmission rate transmits the data packet therein, and may also sequentially send each data stream, and send the data packet buffered therein at the transmission rate of the data stream, thereby reducing burst data packets received by the same destination end, and simultaneously transmitting end Packet transmission The delivery will not stop, reducing the packet loss rate and improving the efficiency of data transmission.

In all the above embodiments, the sending end may be a line card or an interface board, etc., for sending a data packet; the destination end may be a port, a core in a CPU, a line card or a processing board, etc., for receiving a data packet.

A person skilled in the art can understand that all or part of the process of implementing the above method embodiments may be completed by using computer program related hardware, and the foregoing program may be stored in a computer readable storage medium, when executed, The foregoing steps include the steps of the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.

The above is only a specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Accordingly, the scope of the invention should be determined by the scope of the appended claims.

Claims

Claim

A flow control method for a switching network, characterized in that:

Obtaining a rate at which at least one destination receives a data packet;

Transmitting a data packet corresponding to the destination end to the destination end, where a rate at which the data packet is sent is less than or equal to a rate at which the destination end receives the data packet.

The method according to claim 1, wherein the data packet corresponding to the destination end is sent to the destination end, wherein a rate at which the data packet is sent is less than or equal to a rate at which the destination end receives a data packet. The rate further includes:

Decoding the data packet in a data stream corresponding to the destination end according to a rate at which the destination end receives the data packet, where a rate of sending the data packet in the data stream is less than or equal to a rate at which the destination end receives the data packet .

The method according to claim 1 or 2, wherein, when there are multiple destinations, the total rate at which the data packet is sent is less than or equal to the rate at which the data packet is sent by the sender.

The method according to claim 2, wherein the destination end has a plurality of, and the sending of the data packet corresponding to the destination end to the destination end comprises:

The data in the data stream corresponding to each destination is sent in sequence, wherein the rate at which the data packets in the data stream are sent is less than or equal to the rate at which each destination receives the data packet.

5. A method according to claim 2 or 4, characterized in that

The buffering the data packet to the data stream corresponding to the destination according to the rate at which the destination end receives the data packet includes:

And according to the rate at which the destination end receives the data packet, the data packet is cached in a data stream corresponding to the destination end according to a preset priority.

6. A flow control device for a switching network, comprising:

a processor, configured to obtain a rate at which at least one destination receives the data packet;

The device according to claim 6, wherein the transmitter comprises: a buffering module, configured to cache the data packet in a data stream corresponding to the destination end according to a rate at which the destination end receives a data packet Medium

And a sending module, configured to send a data packet in the data stream, where a rate of sending the data packet in the data stream is less than or equal to a rate at which the destination end receives the data packet.

The device according to claim 6 or 7, wherein the destination end has multiple, the transmitter is further configured to enable the total rate of sending the data packet to be less than or equal to a sending data packet of the transmitting end. rate.

The device according to claim 7, wherein the destination end has a plurality of, the sending module is specifically configured to sequentially send data in a data stream corresponding to each destination end, where the data stream is sent The rate of the data packet is less than or equal to the rate at which each destination receives the data packet.

10. Apparatus according to claim 7 or claim 9 wherein:

The cache module is specifically configured to cache the data packet in a data stream corresponding to the destination end according to a preset priority according to a rate at which the destination end receives a data packet.