CN109981471A - A kind of method, apparatus and system for alleviating congestion - Google Patents
A kind of method, apparatus and system for alleviating congestion Download PDFInfo
- Publication number
- CN109981471A CN109981471A CN201711449438.8A CN201711449438A CN109981471A CN 109981471 A CN109981471 A CN 109981471A CN 201711449438 A CN201711449438 A CN 201711449438A CN 109981471 A CN109981471 A CN 109981471A
- Authority
- CN
- China
- Prior art keywords
- equipment
- message
- data flow
- source device
- congestion
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/20—Traffic policing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/24—Traffic characterised by specific attributes, e.g. priority or QoS
- H04L47/2458—Modification of priorities while in transit
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/25—Flow control; Congestion control with rate being modified by the source upon detecting a change of network conditions
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
This application provides a kind of methods for alleviating congestion, this method comprises: the first equipment receives the first message of the second device forwards, the first message is the message that the second equipment congestion occurs in response to the exit port of second the first data flow of device forwards and generates.First equipment generates the second message according to first message.First equipment forwards the second message to source device, and the second message, which is used to indicate source device, reduces the rate that source device sends the first data flow to the first equipment.So that the network equipment that congestion occurs is sent back-pressure message to the source device for sending message by this method, reduces the rate that source device is sent, rapid recovery network congestion.
Description
Technical field
The present invention relates to field of communication technology more particularly to a kind of method, apparatus and system for alleviating congestion.
Background technique
Remote direct data access (Remote Direct Memory Access, RDMA) can make a computer
On direct memory access (DMA) another computer on memory.RDMA technology has been applied to data center network.In data
In heart network, a port of interchanger can receive the data flow that multiple network equipments are sent simultaneously, and data center network can deposit
The flow congestion the case where.
In order to solve above-mentioned congestion problems, congestion notification can be sent from receiving end to transmitting terminal.For example, TCP
Receiver sends congestion notification message to TCP sender, to reduce the rate that TCP sender sends message.And above-mentioned side
Method, transmitting terminal may not be able to receive congestion notification in time, aggravate network congestion.
Summary of the invention
The embodiment of the present application provides a kind of method, apparatus and system for alleviating congestion, so that in the process to E-Packet
In, the network equipment that congestion occurs can send back-pressure message to the source device for sending message, the rate that source device is sent is reduced,
Rapid recovery network congestion.
In a first aspect, this application provides a kind of methods for alleviating congestion.It is set this method comprises: the first equipment receives second
The first message that preparation is sent, wherein the first message is exit port of second equipment in response to second the first data flow of device forwards
The message for congestion occur and generating.The purpose Internet protocol (English: Internet Protocol, abbreviation: IP) of first message
Location is equal to the IP address for sending the source device of the first data flow.The first data flow that source device is sent is via the first equipment and the
Two equipment reach purpose equipment, wherein the first data flow reaches the first equipment before reaching the second equipment.First equipment is according to first
Message generates the second message.First equipment sends the second message to source device, wherein the second message is used to indicate source device reduction
Source device sends the rate of the first data flow to the first equipment.
In above-mentioned technical proposal, when congestion occurs for the exit port that the second equipment is used to send the first data flow, second is set
Standby to send the first message to the first equipment, the first equipment generates the second message according to the second message, and by sending to source device
Second message come reduce source device send the first data flow rate, in alleviate the second equipment exit port congestion.In addition,
Congestion triggering source device, which occurs, in the exit port of second equipment reduces the rate for sending the first data flow.Relative to passing through purpose equipment
Triggering source device reduces the technical solution for sending the rate of the first data flow, and the congestion feedback path of above-mentioned technical proposal is shorter.
Therefore, above-mentioned technical proposal facilitates rapid recovery congestion.
In a kind of possible design, it includes: to obtain the first message that the first equipment, which generates the second message according to the first message,
The information for being used to indicate the first priority of middle carrying, and based on first priority that is used to indicate carried in the first message
Information generates the second message.Second message carries the information for being used to indicate the first priority, wherein the forwarding of the first data flow is excellent
First grade is the first priority, and the information that carrying is used to indicate the first priority in the second message is used to indicate source device reduction source and sets
It is standby that the data flow that forwarding priority is the first priority is sent to the first equipment.
Using aforesaid way, the source device for sending the first data flow reduces data identical with the first data stream priority
Stream, not only can also be carried out transmission control to data stream according to priority, be guaranteed data flow with rapid recovery network congestion
Transmission quality.
In a kind of possible design, destination queue of first data stream packets containing the first data flow is to (English: queue
Pair, referred to as: QP).The second data flow that purpose equipment is sent reaches source device via the second equipment and the first equipment,
In, the second data flow reaches the second equipment before reaching the first equipment.The source IP address of second data flow is equal to the IP of purpose equipment
Address, the purpose IP address of the second data flow are equal to the IP address of source device.Second data stream packets contain the purpose of the second data flow
Ground QP, the second message do not include the destination QP of the second data flow.
Using the above scheme, so that source device and purpose equipment do not need to safeguard the source device when sending data flow
The corresponding relationship of the QP of QP and the purpose equipment, so that it may back-pressure is carried out to the source device for sending the first data flow, do not need for
Each data flow establishes QP connection tracking table, can also avoid passing through to table look-up and sends message, to realize to the quick anti-of source device
Pressure alleviates network overhead.
Second aspect, this application provides a kind of methods for alleviating congestion.It is set this method comprises: the second equipment receives first
First data flow of standby forwarding, wherein the first data flow is sent by source device and passed through the first equipment and the second device forwards
Reach purpose equipment.Second equipment determines that congestion occurs in the exit port of second the first data flow of device forwards.It is set in response to second
There is congestion in the exit port of the first data flow of standby forwarding, and the second equipment generates the first message, wherein the destination IP of the first message
Location is equal to the IP address for sending the source device of the first data flow.First message is sent to the first equipment by the second equipment, with instruction
First equipment generates the message for being used to indicate the rate that source device reduction source device sends the first data flow to the first equipment.
Through the above technical solutions, when there is congestion, Ke Yixiang in the exit port that second network equipment sends the first data flow
The source device for sending the first data flow sends message to indicate that source device reduces the rate of transmission data flow, without passing through mesh
Equipment send message to source device to indicate to reduce the rate for sending data flow, in this way shorten to source device and send message
Path is conducive to rapid recovery network congestion.
The third aspect, the application provide one kind the first equipment, and any one of execution first aspect or first aspect may
Method in implementation.Specifically, first equipment include can for executing any one of first aspect or first aspect
The unit of method in the implementation of energy.
Fourth aspect, the application provide one kind the second equipment, and any one of execution second aspect or second aspect may
Method in implementation.Specifically, second equipment include can for executing any one of second aspect or second aspect
The unit of method in the implementation of energy.
5th aspect, the application provide a kind of system for alleviating congestion, the system comprises: the first equipment and second is set
It is standby.The first message that first equipment is used to receive the first data flow of source device transmission and the second equipment is sent, according to first
Message generates the second message, and sends the second message to source device.First message is hair of second equipment in response to the second equipment
The message for sending the exit port of the first data flow congestion occur and generating.The purpose IP address of first message, which is equal to, sends the first data
The IP address of the source device of stream.The first data flow that source device is sent reaches purpose by the forwarding of the first equipment and the second equipment
Equipment.First data flow, which reaches to reach first the second message of equipment before the second equipment and be used to indicate source device, reduces source device to the
One equipment sends the rate of the first data flow.
Second equipment is used to receive the first data flow of the first equipment transmission, and determines that the second equipment sends the first data flow
Exit port there is congestion.There is congestion in the exit port that second equipment sends the first data flow in response to the second equipment, generates the
One message, and the first message is sent to the first equipment.
6th aspect, this application provides a kind of computer readable storage medium, in the computer readable storage medium
It is stored with instruction, when run on a computer, so that computer executes above-mentioned first aspect and each side of being able to achieve
The method of formula.
7th aspect, this application provides another computer readable storage medium, the computer readable storage medium
In be stored with instruction, when run on a computer, so that computer executes above-mentioned second aspect and each is able to achieve
The method of mode.
Eighth aspect, this application provides a kind of network equipment, the network equipment includes network interface, processor, deposits
Reservoir and the bus for connecting the network interface, processor and memory.The memory is for storing program, instruction or generation
Code, the processor are used to execute program, instruction or code in the memory, complete above-mentioned first aspect and it is each can
It is able to achieve the method for mode.
9th aspect, this application provides a kind of network equipment, the network equipment includes network interface, processor, deposits
Reservoir and the bus for connecting the network interface, processor and memory.The memory is for storing program, instruction or generation
Code, the processor are used to execute program, instruction or code in the memory, complete above-mentioned second aspect and it is each can
It is able to achieve the method for mode.
Detailed description of the invention
Fig. 1 is a kind of application scenarios schematic diagram of network system provided by the embodiments of the present application.
Fig. 2 is a kind of method flow schematic diagram for alleviating congestion provided by the embodiments of the present application.
Fig. 3 is the method flow schematic diagram of another alleviation congestion provided by the embodiments of the present application.
Fig. 4 is a kind of form schematic diagram of first message provided by the embodiments of the present application.
Fig. 5 is a kind of form schematic diagram of second message provided by the embodiments of the present application.
Fig. 6 is a kind of structural framing schematic diagram of first equipment provided by the embodiments of the present application.
Fig. 7 is a kind of structural framing schematic diagram of second equipment provided by the embodiments of the present application.
Fig. 8 is a kind of hardware configuration block schematic illustration of first equipment provided by the embodiments of the present application.
Fig. 9 is a kind of hardware configuration block schematic illustration of second equipment provided by the embodiments of the present application
Figure 10 is a kind of system structure diagram for alleviating congestion provided by the embodiments of the present application.
Specific embodiment
The description and claims of this application and term " first " in above-mentioned attached drawing, " second " etc. are for distinguishing
Similar object, without being used to describe a particular order or precedence order.It should be understood that the data used in this way are in appropriate feelings
It can be interchanged under condition, so that the embodiments described herein can be real with the sequence other than the content for illustrating or describing herein
It applies.In addition, term " includes " and " having " and their any deformation, it is intended that cover it is non-exclusive include, for example, packet
The process, method, system, product or equipment for having contained a series of steps or units those of be not necessarily limited to be clearly listed step or
Unit, but may include other steps being not clearly listed or intrinsic for these process, methods, product or equipment or
Unit.
Fig. 1 provides a kind of system architecture schematic diagram for controlling network congestion.Network system 100 in Fig. 1 can be
CLOS network system.Network system 100 includes: aggregation node 101 and aggregation node 102, rack interchanger 103, rack exchange
Machine 104, rack interchanger 105 and rack interchanger 106 and server 107 to 112.Wherein, server 107-115 can be with
It is the storage server or calculation server of data center.Server 107 to 115 is for sending and receiving message.Rack exchange
Machine 103 to 106 is used for the message that forwarding server 107 to 115 is sent.Aggregation node 101 and 102 can be to have and E-Packet
The switch or router of function.Aggregation node 101 and 102 is used for the message that forwarding server is sent.As shown in Figure 1, convergence
Node 101 is separately connected with rack interchanger 103-106.Aggregation node 102 is separately connected with rack interchanger 103-106.Machine
Frame interchanger 103 is separately connected with server 107, server 108 and server 109.Rack interchanger 104 and server 110,
Server 111 is separately connected.Rack interchanger 105 is separately connected with server 112 and server 113 respectively.Rack interchanger
106 are separately connected with server 114 and 115.Hereafter using server 107 as source device, equipment is server 115 as a purpose
Repeating process of the contract quotation text in data center for example.Message is sent to machine connected to it as source device by server 107
Frame interchanger 103.Rack interchanger 103 receives the message and the message is transmitted to aggregation node 101.Aggregation node
101 receive the message and the message are transmitted to rack interchanger 106.It receives the message and is transmitted to the message
Server 115 connected to it.Server 115 as a purpose equipment receiver frame interchanger 106 forward derive from server
107 message.
In order to reduce the time delay of server end transmitting-receiving message in network transmission, data center network can use RDMA skill
Art.In the data center network of application rdma protocol, server 107 described above can as the message that transmitting terminal is sent
Using the direct access protocol second edition sheet of the long-distance inner (English: remote direct memory based on fusion ether
Access over Converged Ethernet version 2, referred to as: RoCEv2) the CLOS network architecture shown in Fig. 1
Data center forwarding, for example, server 107 as transmitting terminal send using RoCEv2 agreement message give receiving end clothes
Business device 115, rack interchanger 103 receive the RoCEv2 message that transmitting terminal server 107 connected to it is sent, and rack is handed over
103 forwardings of the changing planes RoCEv2 message is to aggregation node 101.If aggregation node 101 sends going out for the RoCEv2 message
There is congestion in interface, such as congestion occurs in the interface that aggregation node 101 is connect with rack interchanger 106, then aggregation node 101 will
In the RoCEv2 message ecn (explicit congestion notification) (English: Explicit Congestion Notification, referred to as:
ECN) position is labeled as congestion state, and the message is continued to send to receiving end server 115, and receiving end server 115 connects
The RoCEv2 message that the ECN is labeled as congestion state is received, congestion is constructed according to the RoCEv2 message for carrying congestion marking
Notification packet (English: Congestion Notification Packet, referred to as: CNP), and by the CNP to the transmission
Server 107 is held to send.Transmitting terminal server 107 receives the CNP according to congestion avoidance algorithm reduction of speed, is gathered around with alleviating network
Plug.But the method for the reduction network congestion, congestion feedback path is longer, and may result in transmitting terminal, cannot to receive congestion in time logical
Know, aggravate network congestion, message is caused to be dropped.The method of another kind control network congestion is firstly the need of according to transmit queue
Establish QP with receiving queue, transmitting terminal server can in the data flow to be sent adding purpose QP, destination QP be used for
Indicate that (queue pair number) is checked numbers in the queue for being located at purpose equipment.Even if the equipment for therefore sending data flow needs
Congestion notification message is sent to the sending ending equipment for sending data flow, it is also desirable to the adding purpose ground QP in congestion notification message,
QP connection tracking table is thus needed, such as establishes the queue of transmitting terminal server 107 and checks numbers and receiving end server 115
The mapping table that queue is checked numbers, QP are for identifying virtual port number, when source device sends message to purpose equipment, usually
The QP of destination is carried in messages.This is because RoCEv2 message and CNP require to be reported according to RoCEv2 in RDMA network
The QP's of destination that carries in text or CNP sends message.Second of jamming control method by generation congestion the network equipment,
Such as the aggregation node 101, RoCEv2 message generates the first CNP and is sent to by rack interchanger 103 based on the received
Transmitting terminal server 107.Since the first CNP is generated according to RoCEv2 message, the destination QP carried in the first CNP,
And the first CNP is sent to transmitting terminal server 107 and then needs to replace with the purpose QP in the first CNP and destination QP
Corresponding source QP, therefore before the CNP is received by transmitting terminal server 107 also need that mesh will be carried in the first CNP
The QP on ground replace with the QP of the corresponding source device in destination, generate the 2nd CNP, such transmitting terminal server 107 can basis
The QP of the corresponding source device in destination carried in 2nd CNP receives the message, to control the transmission of transmitting terminal server 107
Message speed, to alleviate congestion.And not only needing to act on behalf of in the above process is that each data flow establishes QP connection tracking table, is also needed
The corresponding relationship of the QP between transmitting terminal server and receiving end server is searched, and QP connection tracks table size by network
The limitation of equipment storage resource, and time that will make the to table look-up when list item scale is excessive is longer, is unfavorable for rapid recovery and gathers around
Plug causes performance to decline.
In order to avoid agency be data flow establish QP connection tracking table and by table look-up the determining transmitting terminal network equipment with
The corresponding relationship of the QP of the receiving end network equipment, as shown in Fig. 2, providing a kind of method stream for alleviating congestion for the embodiment of the present application
Journey schematic diagram includes source device, purpose equipment, the first equipment and the second equipment in the network.The network can be adapted for Fig. 1 institute
In the scene shown, source device and purpose equipment can be any one in server 107-115 shown in FIG. 1, the second equipment
It can be any one in rack interchanger 103-106, the first equipment can be any one net other than server 107-115
Network equipment, such as the first equipment can be the aggregation node 101 in Fig. 1, be also possible to rack interchanger 105.This method is specific
Include:
S210, the second equipment receive the first data flow of the first device forwards.
Wherein, the first data flow that source device is sent reaches purpose equipment by the first equipment and the second equipment.
In one possible implementation, source device and purpose equipment can be the storage server of data center, meter
Calculate the network equipments such as server.
In one possible implementation, the first data flow is sent to the first equipment by source device, and the first equipment will connect
The first data flow received is sent to the second equipment, and the first data flow received is sent to purpose equipment by the second equipment,
In, there can also be other network equipments between the first equipment and the second equipment and between the second equipment and purpose equipment, for example,
It further include third equipment between first equipment and the second equipment, i.e. the first data flow is sent to by the first equipment by third equipment
Second equipment.
In one possible implementation, the first data flow uses RoCEv2 agreement, the IP head of first data flow
Ecn (explicit congestion notification) (English: Explicit Congestion Notification, abbreviation ECN) position is labeled as " 10 ", indicates
The message has the function of ecn (explicit congestion notification).The source IP address of first data flow is the IP address of source device, the first data flow
Purpose IP address is the IP address of purpose equipment.
S220, the second equipment determine that congestion occurs in the exit port of second the first data flow of device forwards.
There is congestion in response to the exit port of second the first data flow of device forwards in S230, and the second equipment generates the first report
Text.
In one possible implementation, if the second equipment determines that congestion occurs in the port for sending the first data flow,
Then the second equipment generates CNP.Fig. 4 provides a kind of form schematic diagram of second message for the embodiment of the present application, that is, provides a kind of CNP
Form schematic diagram.
In one possible implementation, the precedence information of the first data flow is stored in the first message by the first equipment
In.
For example, as shown in Figure 1, the second equipment is the aggregation node 102 in Fig. 1, the first equipment is rack interchanger
103.Aggregation node 102 receives the first data flow from rack interchanger 103, and is determined and sent according to the routing of the first data flow
The exit port of first data flow.When the aggregation node 102 determines that congestion occurs in the exit port for sending the first data flow, converge
Poly- node 102 generates CNP, and the purpose IP address of the CNP is identical as the source IP address of first data flow, and in the CNP
" the basic transport header of IB " (InfiniBand Base Transport Header) in reserved field save this first number
According to the precedence information of stream.
In one possible implementation, the purpose IP address of the first message is equal to the source device for sending the first data flow
IP address, i.e. the second equipment by the addition of the source device IP address of the first data flow the first message purpose IP address.
In one possible implementation, when congestion occurs in the exit port that the second equipment sends the first data flow, meeting
It triggers the second equipment and generates the first message.
In one possible implementation, the second equipment is using the purpose IP address of the first data flow as the first message
Source IP address, and using the source IP address of the first data flow as the purpose IP address of the first message, the second equipment is according to the first report
The purpose IP address of text will send the first message.
In one possible implementation, the first equipment is CNP according to the first message that the first data flow generates.
First message is sent to first equipment by S240, the second equipment.
First message be used to indicate the first equipment generation be used to indicate source device reduce source device sent out to the first equipment
Send the message of the rate of the first data flow.
As shown in figure 3, providing a kind of method flow schematic diagram for alleviating congestion for the embodiment of the present application, include in the network
Source device, purpose equipment, the first equipment and the second equipment.The network can be adapted in scene shown in FIG. 1, source device and mesh
Equipment can be any one in server 107-115 shown in FIG. 1, the first equipment can be rack interchanger 103-
Any one in 106, the second equipment can be any one network equipment other than server 107-115, such as the second equipment
It can be the aggregation node 101 in Fig. 1, be also possible to rack interchanger 105.This method specifically includes:
S310, the first equipment receive the first message that the second equipment is sent.
Wherein, the first message to be the second equipment in response to the exit port of second the first data flow of device forwards there is congestion and
The message of generation.The purpose IP address of first message is equal to the IP address for sending the source device of the first data flow.Source device is sent
The first data flow reach purpose equipment via the first equipment and the second equipment, the first data flow reaches before reaching the second equipment
First equipment.
In one possible implementation, the first data flow sends mesh by the first equipment and the second equipment by source device
Equipment, i.e. the first data flow is sent to the first equipment by source device, and the first data flow received is sent to by the first equipment
The data flow received is sent to purpose equipment by the second equipment, the second equipment.
In one possible implementation, the first data flow is sent to purpose equipment by first path by source device,
It can also include that at least one network is set in addition to including source device, purpose equipment, the first equipment and the second equipment in first path
It is standby, such as further include third equipment between the first equipment and the second equipment, i.e. the first data flow is sent to third by the first equipment
The first data flow received is sent to the second equipment by equipment, third equipment, is sent the first data flow to by the second equipment
Purpose equipment.
S320, the first equipment generate the second message according to the first message.
In one possible implementation, the first message can be CNP, which has congestion notification function, and first sets
The standby CNP that receives will trigger generation third message.
Optionally, it includes: to obtain in first message to carry that the first equipment, which generates the second message according to the first message,
It is used to indicate the information of the first priority and is generated based on the information for being used to indicate the first priority carried in the first message
Second message.Second message carries the information for being used to indicate the first priority.The forwarding priority of first data flow is first excellent
First grade carries in the second message and is used to indicate the information of the first priority and is used to indicate source device and reduces source device to the first equipment
Send the data flow that forwarding priority is the first priority.
In one possible implementation, the first equipment obtains the priority letter of the first data flow from the first message
It ceases, and sets the value of the priority of the first data flow of acquisition to the priority of third message, is i.e. the forwarding of the first message is excellent
First grade is identical as the forwarding priority of third message.
In one possible implementation, the first equipment generates the second message according to the quantity for receiving the first message,
That is threshold value is arranged in the first message that the first equipment interconnection receives, when the quantity that the first equipment receives the first message reaches or surpasses
When crossing threshold value, the first equipment of triggering generates the second message.
For example, as shown in Figure 1, equipment is rack interchanger 103, the first equipment is aggregation node 101, and source device is
Server 107, purpose equipment are server 115.First equipment receives the CNP that the second equipment is sent, and generates priority-based
Flow control (English: Priority-based Flow Control, referred to as: PFC) and message, the value of the priority of the PFC message
It is that equipment is obtained from CNP reserved field, the information of the priority of the CNP reserved field the first data flow of preserver, i.e. the PFC
The value of the priority of message is identical as the value of the priority of the first message.
For example, as shown in Figure 1, the first equipment is rack interchanger 103.Fig. 5 provides a kind of report of the second message
Literary form schematic diagram, i.e. PFC message format figure.Fig. 4 provides a kind of message format schematic diagram of the first message, i.e. CNP format
Schematic diagram.Rack interchanger 103 obtains the preferential of the first saved data flow from the reserved field of the BTH in Fig. 4 in CNP
The information of grade, the value from the priority obtained in CNP is stored in excellent in PFC message shown in fig. 5 by rack interchanger 103
First grade-is enabled-position vector (Priority-enable-vector) in, that is, be arranged the value of the priority of the PFC message with from Fig. 4
The value of the priority of the first message obtained in middle CNP is identical.
In one possible implementation, the time of control back-pressure source device can also be carried in the second message, such as
The time numerical value of " when vector " position (time-vector) addition back-pressure source device in Fig. 5.
Optionally, destination queue of first data stream packets containing the first data flow is to QP.The second number that purpose equipment is sent
Source device is reached by the second equipment and the first equipment according to flowing through.Second data flow reaches the second equipment before reaching the first equipment.
The source IP address of second data flow is equal to the IP address of purpose equipment.The purpose IP address of second data flow is equal to the IP of source device
Address.Second data stream packets contain the destination QP of the second data flow, and the second message does not include the destination QP of the second data flow.
In one possible implementation, QP is made of transmit queue and receiving queue.The destination of first data flow
QP is used to indicate the queue positioned at the purpose equipment and checks numbers, and the destination QP of the second data flow is used to indicate positioned at the source
The queue of equipment is checked numbers.Queue positioned at the purpose equipment check numbers be located at the source device queue check numbers it is corresponding
The same queue pair, and the application does not need to establish QP connection tracking table, is less used in when sending CNP to source device and needs to pass through
QP connection tracking table is searched to obtain the destination QP of CNP, reduces network overhead and the time for establishing QP connection tracking table, avoids
It tables look-up with rapid recovery network congestion.
S330, the first equipment send the second message to source device.
Wherein, the second message, which is used to indicate source device, reduces the rate that source device sends the first data flow to the first equipment.
In one possible implementation, the first equipment determines that source is set according to the purpose IP address carried in the first message
Standby be the next-hop of the first equipment, the first equipment by the target MAC (Media Access Control) address addition of the first message the second message purpose MAC
The second message is sent next-hop by location.
In one possible implementation, the first equipment is determined according to the target MAC (Media Access Control) address of the second message sends second
The port of message, and the second message is sent to source device by identified port.Second message is used for for sending out described in back-pressure
The sending end network equipment sends message identical with the priority of the third message.
For example, as shown in Figure 1, equipment is rack interchanger 103, the first equipment is aggregation node 101, transmitting terminal net
Network equipment is server 107, and the receiving end network equipment is server 115, and third message is PFC message.Rack interchanger 103
The port for sending the PFC is determined according to the purpose IP address of PFC, and the PFC is sent to by server 107 by the port.
In one possible implementation, it includes the first equipment that the second message is sent to purpose equipment by the first equipment
When determining that the quantity of the first equipment the first message of reception reaches threshold value, the first equipment of triggering sends the second message to purpose equipment.
In a kind of possible implementation, the setting of the first equipment receives the threshold value of the quantity of the first message, when the first equipment
When receiving the quantity of the first message more than or equal to threshold value, the first equipment sends second message to source device, which can
With flexible setting.
For example, as shown in Figure 1, the first equipment is rack interchanger 103, the second equipment is aggregation node 101, and source is set
Standby is server 107, and purpose equipment is server 115, and the first message is CNP, and the second message is PFC message.Rack interchanger
103 determine receive aggregation nodes 101 send CNP quantity be greater than or equal to setting threshold value when, the rack interchanger 103 to
Server 107 sends a PFC message.
Optionally, the second message, which also carries, determines back-pressure duration value, and the value of back-pressure duration is determined by the first equipment.
For example, as shown in Figure 1, equipment is rack interchanger 103, rack interchanger can determine that back-pressure source device is sent out
Send the duration of data flow.
By the above method, when congestion occurs for the second equipment in the forward-path of the first data flow, by gathering around
Second equipment of plug sends congestion notification message to the first equipment.First equipment receives the congestion notification message and generates for back-pressure
The flow control message priority-based of source device alleviates congestion state.The above method is not needed by by control message
Source device QP corresponding with purpose equipment is established, agency is not needed yet and establishes QP connection tracking table, does not need more to report by first
Text, which is sent to, could will continue to send the first message by the purpose QP that QP table replaces the first message in the transmission process of source device,
Network overhead is greatly reduced, the time tabled look-up is saved, it being capable of rapid recovery congestion.
As shown in fig. 6, for this application provides a kind of the first equipment 600 for alleviating congestion.First equipment 600 both can be with
It is any one in rack interchanger 107 to 115 in Fig. 1, can also be the first equipment in method flow Fig. 2 and Fig. 3, it can be with
Realize the function of the first equipment.First equipment 600 includes receiving unit 601, processing unit 602 and transmission unit 603.
Receiving unit 601 is used to receive the first message of the second equipment transmission, wherein the first message is the response of the second equipment
In the message that congestion occurs in the exit port of the first data flow of forwarding of the second equipment and generates.The purpose IP address etc. of first message
In the IP address for the source device for sending the first data flow.The first data flow that source device is sent is via the first equipment 600 and the
Two equipment reach purpose equipment, and the first data flow reaches the first equipment 600 before reaching the second equipment.
Processing unit 602 is used to generate the second message according to received first message of receiving unit 601.
Transmission unit 603 is used to send the second message that processing unit 602 generates to source device, wherein the second message is used
Rate of the source device to the first equipment 600 the first data flow of transmission is reduced in instruction source device.
Optionally, it is excellent that the processing unit 602 was specifically used for carrying in acquisition first message is used to indicate first
The information of first grade and described the is generated based on the information for being used to indicate first priority carried in first message
Two reports.Second message carries the information for being used to indicate first priority.The forwarding priority of first data flow
For first priority.The information that carrying is used to indicate first priority in second message is used to indicate the source
Equipment reduces the speed that the source device sends the data flow that forwarding priority is first priority to first equipment 600
Rate.
In this specific embodiment, the specific implementation of receiving unit 601, processing unit 602 and transmission unit 603 can be with
The function and implementation steps of the first equipment described in referring to figs. 2 and 3, for sake of simplicity, repeating no more.
As shown in fig. 7, for this application provides a kind of the second equipment 700 for alleviating congestion.Second equipment 700 both can be with
It is any one in rack interchanger 107 to 115 or any one in aggregation node 103 to 106 in Fig. 1, can also be
The second equipment energy in method flow Fig. 2 and Fig. 3.Second equipment 700 includes receiving unit 701, determination unit 702, processing
Unit 703 and transmission unit 704.
Receiving unit 701 is used to receive the first data flow of the first device forwards, wherein the first data flow is sent out by source device
It send and passes through the second equipment 700 and the first equipment reaches purpose equipment.
Determination unit 702 is for determining that congestion occurs in the exit port of first the first data flow of device forwards.
Processing unit 703 is used to congestion occur in response to the exit port of first the first data flow of device forwards, and generates the
One message.The purpose IP address of first message is equal to the IP address for sending the source device of the first data flow.
Transmission unit 704 is used to the first message being sent to the first equipment, to indicate that the generation of the first equipment is used to indicate source
Equipment reduces the message that source device sends the rate of the first data flow to the first equipment.
In this specific embodiment, receiving unit 701, determination unit 702, processing unit 703 and transmission unit 704
Specific implementation can referring to figs. 2 and 3 described in the second equipment function and implementation steps, for sake of simplicity, repeating no more.
As shown in figure 8, for this application provides another the first equipment 800 for alleviating congestion.First equipment 800 both may be used
With any one in rack interchanger 107 to 115 in Fig. 1, the first equipment in method flow Fig. 2 and Fig. 3 can also be, it can be with
Realize the function of first equipment.First equipment 800 includes network interface 801 and processor 802, can also include depositing
Reservoir 803.
Processor 802 can be central processing unit (English: central processing unit, abbreviation: CPU), network
Processor (English: network processor, referred to as: NP), specific integrated circuit (English: application-sPE equipment
Cific integrated circuit, referred to as: ASIC) or programmable logic device (English: programmable logic
Device, abbreviation: PLD).Above-mentioned PLD can be Complex Programmable Logic Devices (English: complex programmable
Logic device, abbreviation: CPLD), field programmable gate array (English: field-programmable gate
Array, abbreviation: FPGA), Universal Array Logic (English: generic array logic, abbreviation: GAL) or any combination thereof.
Processor 802 is responsible for management bus 804 and common processing, can also provide various functions, including timing, peripheral interface, electricity
Pressure is adjusted, power management and other control functions.Memory 803 can be used for storage processor 802 and execute operation when institute
The data used.
Network interface 801 can be wireline interface, such as Fiber Distributed Data Interface (English: Fiber
Distributed Data Interface, referred to as: FDDI), Ethernet (English: Ethernet) interface.Network interface 801
It can be wireless interface, such as wireless lan interfaces.
Memory 803 can be content adressable memory (English: content-addressable memory, referred to as:
), such as three-state content addressing memory (English: ternary CAM, referred to as: TCAM) or random access memory (English CAM
Text: random-access memory, referred to as: RAM).
Memory 803 also can integrate in processor 802.If memory 803 and processor 802 are independent from each other
Device, memory 803 are connected with processor 802, such as memory 803 and processor 802 can pass through bus communication.Network connects
Mouthfuls 801 and processor 802 can also be direct-connected with processor 802 by bus communication, network interface 801.
Bus 804 may include the bus and bridge of any number of interconnection, and bus 804 will include being represented by processor 802
One or more processors 802 and the various circuits of memory that represent of memory 803 link together.Bus 804 may be used also
To link together various other circuits of such as peripheral equipment, voltage-stablizer and management circuit or the like, these are all
It is known in the art, therefore, it will not be further described herein.
Network interface 801 is used to receive the first message of the second equipment transmission and for sending processing unit to source device
The second message generated.First message is that the second equipment is gathered around in response to the exit port of the first data flow of transmission of the second equipment
The message filled in and generated.The purpose Internet protocol IP address of first message is with being equal to the IP for sending the source device of the first data flow
Location.The first data flow that source device is sent reaches purpose equipment via the first equipment 800 and the second equipment.First data flow to
The first equipment 800 is reached before up to the second equipment.
Processor 802 is used to generate the second message according to received first message of receiving unit.
In this specific embodiment, the specific implementation of the processor 802 and the network interface 801 can be with reference to figure
The function and implementation steps of first equipment described in 2 and Fig. 3, for sake of simplicity, repeating no more.
As shown in figure 9, for this application provides another the second equipment 900 for alleviating congestion.Second equipment 900 both may be used
With any one in rack interchanger 107 to 115 in Fig. 1 or any one in aggregation node 103 to 106, can also be
The second equipment in method flow Fig. 2 and Fig. 3, may be implemented the function of the second equipment, and the function of second equipment may be implemented
Energy.Second equipment 900 includes network interface 901 and processor 902, can also include memory 903.
The including but not limited to central processing unit of processor 902 (English: central processing unit, referred to as:
CPU), network processing unit (English: network processor, abbreviation: NP), specific integrated circuit (English:
Application-sPE equipment cific integrated circuit, referred to as: ASIC) or programmable logic device (English
One or more of text: programmable logic device, abbreviation: PLD).Above-mentioned PLD can be complex programmable
Logical device (English: complex programmable logic device, abbreviation: CPLD), field programmable gate battle array
Column (English: field-programmable gate array, abbreviation: FPGA), Universal Array Logic (English: generic
Array logic, abbreviation: GAL) or any combination thereof.Processor 902 is responsible for management bus 904 and common processing, can be with
Various functions, including timing, peripheral interface are provided, voltage is adjusted, power management and other control functions.Memory 903 can
The used data when executing operation of processor 902 for storage.
Network interface 901 can be wireline interface, such as Fiber Distributed Data Interface (English: Fiber
Distributed Data Interface, referred to as: FDDI), Ethernet (English: Ethernet) interface.Network interface 901
It can be wireless interface, such as wireless lan interfaces.
Memory 903 can be including but not limited to content adressable memory (English: content-addressable
Memory, referred to as: CAM), such as three-state content addressing memory (English: ternary CAM, referred to as: TCAM), arbitrary access
Memory (English: random-access memory, referred to as: RAM).
Memory 903 also can integrate in processor 902.If memory 903 and processor 902 are independent from each other
Device, memory 903 are connected with processor 902, such as memory 903 and processor 902 can pass through bus communication.Network connects
Mouthfuls 901 and processor 902 can also be direct-connected with processor 902 by bus communication, network interface 901.
Bus 904 may include the bus and bridge of any number of interconnection, and bus 904 will include being represented by processor 902
One or more processors 902 and the various circuits of memory that represent of memory 903 link together.Bus 904 may be used also
To link together various other circuits of such as peripheral equipment, voltage-stablizer and management circuit or the like, these are all
It is known in the art, therefore, it will not be further described herein.
Network interface 901 is used for the first data flow for receiving the transmission of the first equipment and sends first message
To first equipment.First data flow is sent by source device and is passed through second equipment 900 and described first sets
Standby forwarding reaches purpose equipment.
There is congestion in the exit port that processor 902 is used to determine that the first equipment sends the first data flow, and for responding
There is congestion in the exit port that the first equipment sends the first data flow, second equipment 900 generates the first message.First message
Purpose IP address be equal to the IP address for sending the source device of first data flow.
In this specific embodiment, the specific implementation of the processor 902 and the network interface 901 can be with reference to figure
The function and implementation steps of first equipment described in 2 and Fig. 3, for sake of simplicity, repeating no more.
As shown in Figure 10, a kind of system 1000 for alleviating congestion is provided for the application.The system 1000 includes the first equipment
1001 and second equipment 1002.First equipment 1001 is either any one in rack interchanger 107 to 115 in Fig. 1, also
It can be the first equipment in method flow Fig. 2 and Fig. 3, the function of first equipment may be implemented.Second equipment 1002
Either any one in rack interchanger 107 to 115 or any one in aggregation node 103 to 106 in Fig. 1, also
It can be the second equipment in method flow Fig. 2 and Fig. 3, the function of the second equipment may be implemented.
The first message that first equipment 1001 is used to receive the first data flow of source device transmission and the second equipment is sent,
The second message is generated according to the first message, and sends the second message to source device.First message be the second equipment 1002 in response to
There is congestion and the message that generates in the exit port of the first data flow of transmission of second equipment 1002.The purpose IP address of first message
Equal to the IP address for the source device for sending the first data flow.First data flow is sent to the first equipment 1001 and by source device
The forwarding of two equipment 1002 reaches purpose equipment.First data flow reaches the first equipment 1,001 second before reaching the second equipment 1002
Message, which is used to indicate source device, reduces the rate that source device sends the first data flow to the first equipment 1001.
Second equipment 1002 is used to receive the first data flow of the first equipment 1001 transmission, and determines the second equipment 1002 hair
The exit port of the first data flow is sent congestion occur.Second equipment 1002 sends going out for the first data flow in response to the second equipment 1002
There is congestion in port, generates the first message, and the first message is sent to the first equipment 1001.
In this specific embodiment, first equipment 1001 specific implementation can referring to figs. 2 and 3 described in the
The function and implementation steps of one equipment.Second equipment 1002 specific implementation can referring to figs. 2 and 3 described in second set
Standby 1002 function and implementation steps, for sake of simplicity, repeating no more.
It should be understood that the size of the serial number of above-mentioned each method is not meant to execute suitable in the various embodiments of the application
Sequence it is successive, the execution of each method sequence should be determined by its function and internal logic, the implementation without coping with the embodiment of the present application
Process constitutes any restriction.
In several embodiments provided herein, it should be understood that disclosed method and apparatus can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the module, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple module or components can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit
Letter connection can be electrical property, mechanical or other forms.
The module as illustrated by the separation member may or may not be physically separated, aobvious as module
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, each functional module in each embodiment of the application can integrate in one processing unit
It is that modules physically exist alone, can also be integrated in one unit with two or more modules.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
If the integrated unit is realized in the form of combination of hardware software and sells or use as independent product
When, the software can store in a computer readable storage medium.Based on this understanding, the technical side of the application
Case the part that contributes to existing technology technical characteristic can be embodied in the form of software products, which produces
Product are stored in a storage medium, including some instructions are used so that a computer equipment (can be personal computer, take
Business device or the network equipment etc.) execute some or all of each embodiment the method for the application step.And storage above-mentioned
Medium can be that USB flash disk, mobile hard disk, read-only memory (referred to as: ROM, English: Read-Only Memory), arbitrary access deposits
Reservoir (referred to as: RAM, English: Random Access Memory), magnetic or disk.
Claims (12)
1. a kind of method for alleviating congestion, which is characterized in that the described method includes:
First equipment receives the first message that the second equipment is sent, and first message is second equipment in response to described the
There is congestion and the message that generates in the exit port of two the first data flows of device forwards, the purpose Internet protocol IP of first message
Address is equal to the IP address for sending the source device of first data flow, first data flow that the source device is sent via
First equipment and second equipment reach the purpose equipment, before first data flow reaches second equipment
Reach first equipment；
First equipment generates the second message according to first message；
First equipment sends second message to the source device, and second message is used to indicate the source device drop
The low source device sends the rate of first data flow to first equipment.
2. the method according to claim 1, wherein first equipment generates second according to first message
Message includes:
Obtain the information for being used to indicate the first priority carried in first message；And
Second message is generated based on the information for being used to indicate first priority carried in first message, it is described
Second message carries the information for being used to indicate first priority；
Wherein, the forwarding priority of first data flow is first priority, is carried in second message for referring to
Show that the information of first priority is used to indicate the source device and reduces the source device to first equipment transmission forwarding
Priority is the rate of the data flow of first priority.
3. method according to claim 1 or 2, which is characterized in that first data stream packets contain first data flow
Destination queue to QP, the second data flow that the purpose equipment is sent is via second equipment and first equipment
The source device is reached, second data flow reaches second equipment, second data before reaching first equipment
The source IP address of stream is equal to the IP address of the purpose equipment, and the purpose IP address of second data flow is equal to the source device
IP address, destination QP of second data stream packets containing second data flow, second message do not include described the
The destination QP of two data flows.
4. a kind of method for alleviating congestion, which is characterized in that the described method includes:
Second equipment receives the first data flow of the first device forwards, and first data flow is sent as source device and described in passing through
First equipment and second equipment reach purpose equipment；
Second equipment determines that congestion occurs in the exit port of the first data flow described in second device forwards；
There is congestion in response to the exit port of the first data flow described in second device forwards, second equipment generates first
Message, the purpose IP address of first message are equal to the IP address for sending the source device of first data flow；
First message is sent to first equipment by second equipment, to indicate that first equipment is generated for referring to
Show that the source device reduces the message that the source device sends the rate of first data flow to first equipment.
5. a kind of first equipment characterized by comprising
Receiving unit, for receive the second equipment transmission the first message, first message be second equipment in response to
There is congestion and the message that generates in the exit port of the first data flow of forwarding of second equipment, the purpose net of first message
Border Protocol IP address is equal to the IP address for sending the source device of first data flow, first number that the source device is sent
The purpose equipment is reached by the first equipment and second equipment according to flowing through, first data flow reaches described second and sets
Standby preceding arrival first equipment；
Processing unit, for generating the second message according to received first message of the receiving unit；
Transmission unit, for sending second message that the processing unit generates, second message to the source device
Being used to indicate the source device reduces the rate that the source device sends first data flow to first equipment.
6. the first equipment according to claim 5, which is characterized in that the processing unit is used for:
Obtain the information for being used to indicate the first priority carried in first message；And
Second message is generated based on the information for being used to indicate first priority carried in first message, it is described
Second message carries the information for being used to indicate first priority；
The forwarding priority of first data flow is first priority；
The information that carrying is used to indicate first priority in second message is used to indicate described in the source device reduction
Source device sends the rate for the data flow that forwarding priority is first priority to first equipment.
7. a kind of second equipment characterized by comprising
Receiving unit, for receiving the first data flow of the first device forwards, first data flow is sent and is passed through by source device
It crosses second equipment and first equipment reaches purpose equipment；
Determination unit, for determining that congestion occurs in the exit port of the first data flow described in second device forwards；
Processing unit generates for there is congestion in response to the exit port of the first data flow described in second device forwards
One message, the purpose IP address of first message are equal to the IP address for sending the source device of first data flow；
Transmission unit, for first message to be sent to first equipment, to indicate that the first equipment generation is used for
Indicate that the source device reduces the message that the source device sends the rate of first data flow to first equipment.
8. a kind of system for alleviating congestion, which is characterized in that the system comprises:
First equipment, the first message that the first data flow and the second equipment for receiving source device transmission are sent, according to institute
It states the first message and generates the second message, and send second message to the source device, the destination IP of first message
Location is equal to the IP address for sending the source device of first data flow, and first data flow that the source device is sent is via institute
It states the first equipment and second equipment reaches the purpose equipment, first data flow arrives before reaching second equipment
Up to first equipment, second message is used to indicate the source device reduction source device and sends to first equipment
The rate of first data flow；
Second equipment determines that second equipment turns for receiving first data flow of first device forwards
There is congestion in the exit port for sending out the first data flow described, in response to the exit port of the first data flow described in second device forwards
There is congestion and generate the first message, first message is sent to first equipment.
9. a kind of first equipment characterized by comprising processor and the memory coupled with processor, wherein
The memory, for storing program instruction；
The processor requires described in 1 to 3 any one for the program instruction perform claim by executing in the memory
Method.
10. a kind of second equipment characterized by comprising processor and the memory coupled with processor, wherein
The memory, for storing program instruction；
The processor, for by executing method described in the program instruction perform claim requirement 4 in the memory.
11. a kind of computer-readable medium, including instruction, when it is executed by a computer, so that the computer is executed as weighed
Benefit requires method described in 1 to 3 any one.
12. a kind of computer-readable medium, including instruction, when it is executed by a computer, so that the computer is executed as weighed
Benefit require 4 described in method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711449438.8A CN109981471B (en) | 2017-12-27 | 2017-12-27 | Method, equipment and system for relieving congestion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711449438.8A CN109981471B (en) | 2017-12-27 | 2017-12-27 | Method, equipment and system for relieving congestion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109981471A true CN109981471A (en) | 2019-07-05 |
| CN109981471B CN109981471B (en) | 2023-04-18 |
Family
ID=67071294
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711449438.8A Active CN109981471B (en) | 2017-12-27 | 2017-12-27 | Method, equipment and system for relieving congestion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109981471B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110647071A (en) * | 2019-09-05 | 2020-01-03 | 华为技术有限公司 | Method, device and storage medium for controlling data transmission |
| CN111404826A (en) * | 2020-03-23 | 2020-07-10 | 盛科网络（苏州）有限公司 | Flow planning method and device based on output port feedback |
| CN111614471A (en) * | 2020-04-29 | 2020-09-01 | 网络通信与安全紫金山实验室 | SDN-based DCQCN data transmission system and transmission method |
| CN112242956A (en) * | 2019-07-18 | 2021-01-19 | 华为技术有限公司 | Flow rate control method and device |
| WO2021012902A1 (en) * | 2019-07-24 | 2021-01-28 | 华为技术有限公司 | Method for processing network congestion, and related apparatus |
| CN112311685A (en) * | 2019-07-24 | 2021-02-02 | 华为技术有限公司 | Method and related device for processing network congestion |
| CN112714072A (en) * | 2019-10-25 | 2021-04-27 | 华为技术有限公司 | Method and device for adjusting sending rate |
| CN112751765A (en) * | 2019-10-30 | 2021-05-04 | 华为技术有限公司 | Method and device for adjusting transmission rate |
| CN113162864A (en) * | 2021-04-25 | 2021-07-23 | 中国工商银行股份有限公司 | RoCE network flow control method, device, equipment and storage medium |
| WO2021244240A1 (en) * | 2020-05-30 | 2021-12-09 | 华为技术有限公司 | Network congestion control method and apparatus, device, system, and storage medium |
| WO2022028342A1 (en) * | 2020-08-05 | 2022-02-10 | 华为技术有限公司 | Method for processing congestion flow and device |
| CN114827234A (en) * | 2022-04-29 | 2022-07-29 | 广东浪潮智慧计算技术有限公司 | Data transmission method, system, device and storage medium |
| CN116032852A (en) * | 2023-03-28 | 2023-04-28 | 新华三工业互联网有限公司 | Flow control method, device, system, equipment and storage medium based on session |
| WO2023122995A1 (en) * | 2021-12-28 | 2023-07-06 | 华为技术有限公司 | Packet transmission method and device |
| CN112714072B (en) * | 2019-10-25 | 2024-06-07 | 华为技术有限公司 | Method and device for adjusting sending rate |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101227495A (en) * | 2008-02-20 | 2008-07-23 | 中兴通讯股份有限公司 | Common telecommunication grouping data network system and congestion control method thereof |
| CN102025617A (en) * | 2010-11-26 | 2011-04-20 | 中兴通讯股份有限公司 | Method and device for controlling congestion of Ethernet |
| CN104303465A (en) * | 2013-03-29 | 2015-01-21 | 华为技术有限公司 | Network congestion processing method, network node, and network system |
| CN106330742A (en) * | 2015-06-23 | 2017-01-11 | 华为技术有限公司 | Flow control method and network controller |
| CN106657365A (en) * | 2016-12-30 | 2017-05-10 | 清华大学 | High concurrent data transmission method based on RDMA (Remote Direct Memory Access) |
| CN107493238A (en) * | 2016-06-13 | 2017-12-19 | 华为技术有限公司 | A kind of method for controlling network congestion, equipment and system |
-
2017
- 2017-12-27 CN CN201711449438.8A patent/CN109981471B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101227495A (en) * | 2008-02-20 | 2008-07-23 | 中兴通讯股份有限公司 | Common telecommunication grouping data network system and congestion control method thereof |
| CN102025617A (en) * | 2010-11-26 | 2011-04-20 | 中兴通讯股份有限公司 | Method and device for controlling congestion of Ethernet |
| CN104303465A (en) * | 2013-03-29 | 2015-01-21 | 华为技术有限公司 | Network congestion processing method, network node, and network system |
| CN106330742A (en) * | 2015-06-23 | 2017-01-11 | 华为技术有限公司 | Flow control method and network controller |
| CN107493238A (en) * | 2016-06-13 | 2017-12-19 | 华为技术有限公司 | A kind of method for controlling network congestion, equipment and system |
| CN106657365A (en) * | 2016-12-30 | 2017-05-10 | 清华大学 | High concurrent data transmission method based on RDMA (Remote Direct Memory Access) |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021008562A1 (en) * | 2019-07-18 | 2021-01-21 | 华为技术有限公司 | Flow rate control method and apparatus |
| CN112242956B (en) * | 2019-07-18 | 2024-04-26 | 华为技术有限公司 | Flow rate control method and device |
| EP3993330A4 (en) * | 2019-07-18 | 2022-08-10 | Huawei Technologies Co., Ltd. | Flow rate control method and apparatus |
| CN112242956A (en) * | 2019-07-18 | 2021-01-19 | 华为技术有限公司 | Flow rate control method and device |
| WO2021012902A1 (en) * | 2019-07-24 | 2021-01-28 | 华为技术有限公司 | Method for processing network congestion, and related apparatus |
| CN112311685A (en) * | 2019-07-24 | 2021-02-02 | 华为技术有限公司 | Method and related device for processing network congestion |
| US12010023B2 (en) | 2019-07-24 | 2024-06-11 | Huawei Technologies Co., Ltd. | Network congestion handling method and related apparatus |
| CN110647071A (en) * | 2019-09-05 | 2020-01-03 | 华为技术有限公司 | Method, device and storage medium for controlling data transmission |
| CN112714072A (en) * | 2019-10-25 | 2021-04-27 | 华为技术有限公司 | Method and device for adjusting sending rate |
| CN112714072B (en) * | 2019-10-25 | 2024-06-07 | 华为技术有限公司 | Method and device for adjusting sending rate |
| CN112751765A (en) * | 2019-10-30 | 2021-05-04 | 华为技术有限公司 | Method and device for adjusting transmission rate |
| CN111404826A (en) * | 2020-03-23 | 2020-07-10 | 盛科网络（苏州）有限公司 | Flow planning method and device based on output port feedback |
| CN111404826B (en) * | 2020-03-23 | 2022-04-22 | 苏州盛科通信股份有限公司 | Flow planning method and device based on output port feedback |
| CN111614471B (en) * | 2020-04-29 | 2022-06-07 | 网络通信与安全紫金山实验室 | DCQCN data transmission system and transmission method based on SDN |
| CN111614471A (en) * | 2020-04-29 | 2020-09-01 | 网络通信与安全紫金山实验室 | SDN-based DCQCN data transmission system and transmission method |
| WO2021244240A1 (en) * | 2020-05-30 | 2021-12-09 | 华为技术有限公司 | Network congestion control method and apparatus, device, system, and storage medium |
| WO2022028342A1 (en) * | 2020-08-05 | 2022-02-10 | 华为技术有限公司 | Method for processing congestion flow and device |
| CN113162864A (en) * | 2021-04-25 | 2021-07-23 | 中国工商银行股份有限公司 | RoCE network flow control method, device, equipment and storage medium |
| CN113162864B (en) * | 2021-04-25 | 2022-11-08 | 中国工商银行股份有限公司 | RoCE network flow control method, device, equipment and storage medium |
| WO2023122995A1 (en) * | 2021-12-28 | 2023-07-06 | 华为技术有限公司 | Packet transmission method and device |
| CN114827234A (en) * | 2022-04-29 | 2022-07-29 | 广东浪潮智慧计算技术有限公司 | Data transmission method, system, device and storage medium |
| CN116032852B (en) * | 2023-03-28 | 2023-06-02 | 新华三工业互联网有限公司 | Flow control method, device, system, equipment and storage medium based on session |
| CN116032852A (en) * | 2023-03-28 | 2023-04-28 | 新华三工业互联网有限公司 | Flow control method, device, system, equipment and storage medium based on session |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109981471B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109981471A (en) | A kind of method, apparatus and system for alleviating congestion | |
| CN102792646B (en) | Communication system, control device, communication means | |
| CN103746911B (en) | A kind of SDN structure and its communication means | |
| US9237098B2 (en) | Media access control (MAC) address summation in Datacenter Ethernet networking | |
| CN109391560A (en) | Notifying method, agent node and the computer equipment of network congestion | |
| CN102368727B (en) | Crossed IP network TRILL network communication method, system thereof and devices | |
| CN103069756A (en) | Communication system, controller, node controlling method and program | |
| CN107483337A (en) | Determine the method and its network equipment in path | |
| CN105556916B (en) | The information statistical method and device of network flow | |
| CN104243270A (en) | Tunnel setup method and tunnel setup device | |
| Sun et al. | Diamond: An Improved Fat-tree Architecture for Large-scale Data Centers. | |
| CN109861913A (en) | A kind of prefix mark notifying method and device across Interior Gateway Protocol | |
| CN108965121A (en) | Transmit method, host and the interchanger of data | |
| Zafar et al. | On implementation of DCTCP on three-tier and fat-tree data center network topologies | |
| CN107864099A (en) | A kind of flow control methods and system of isomery FC networks | |
| CN104702506B (en) | A kind of message transmitting method, network node and message transfer system | |
| CN110311865A (en) | A kind of transmission method and relevant device of video data | |
| CN109995664A (en) | A kind of method, apparatus and system sending data flow | |
| CN106464670B (en) | Network entity and service strategy management method | |
| WO2018004931A1 (en) | Techniques for virtual ethernet switching of a multi-node fabric | |
| CN107872385A (en) | A kind of SDN router-level topology and control method | |
| CN109327400A (en) | A kind of data communications method and data communication network | |
| Yao et al. | Performance evaluation of MPTCP incast based on queuing network | |
| CN109756412A (en) | A kind of data message forwarding method and equipment | |
| US11909546B2 (en) | Method and network node for sending and obtaining assert packet |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |