WO2022105731A1

WO2022105731A1 - Data packet transmission port refresh method and device

Info

Publication number: WO2022105731A1
Application number: PCT/CN2021/130838
Authority: WO
Inventors: 胡乐勇; 曾伟
Original assignee: 武汉绿色网络信息服务有限责任公司
Priority date: 2020-11-20
Filing date: 2021-11-16
Publication date: 2022-05-27
Also published as: CN112104451B; CN112104451A

Abstract

The present invention relates to the technical field of port polling refresh, and provides a data packet transmission port refresh method and device. The method comprises: acquiring a data packet-output port relationship table in a plurality of adjacent time periods; increasing a weight mode of continuous occurrence of an output port for a specified output port, and/or increasing an offset mode during the polling of the output port to adjust the data packet-output port relationship table; by means of a recursive manner, calculating the total number of data packets corresponding to each output port after each round of adjustment; after one or more rounds of adjustment and calculation, using a weight value and/or offset for which the average degree of the total amount of data packets of each corresponding output port reaches a preset condition as parameters to adjust the mapping relationship between the data packets and a target output port, so that a newly received data packet is transmitted according to the adjusted mapping relationship. The present invention can not only ensure the same source and same destination of a data message, but also ensure the maximum utilization of the bandwidth of a downlink output interface.

Description

A method and device for refreshing a data packet transmission port

【Technical field】

The present invention relates to the technical field of port polling and refreshing, in particular to a method and device for refreshing a data packet transmission port.

【Background technique】

In the information age with the rapid development of modern informatization, with the sharp increase of traffic in the existing network, there are higher requirements for network security equipment not only processing performance but also interface bandwidth. There are higher requirements for the size, heat dissipation, and power consumption of the equipment.

For network security equipment, energy saving and consumption reduction will be a problem that major manufacturers need to break through. For service boards or interface boards, there are also higher requirements for the density of physical interfaces of service boards. Therefore, the planning of uplink and downlink interfaces, including the utilization of output interface bandwidth, will also be the core consideration of major hardware manufacturers. technology.

In the prior art, the bottom-level table brushing method usually adopts the conventional index+hash value+port, and the traditional port polling table brushing method. This method of brushing the table has exposed many problems with the increase of traffic on the existing network, such as insufficient logic processing performance, packet loss due to excessive bandwidth on the interface, etc., many manufacturers have to consider upgrading and replacing equipment, or Increase the number of downstream output ports in the output port group to avoid the problem of packet loss at the egress.

Compared with the current service boards, the early ETH, GE ports, and 10G ports occupy relatively little space on service boards or interface boards, and the board density can be made higher. In recent years, high-performance service boards basically support 100G, and may support 400G or even higher in the future. As far as 100G physical interfaces are concerned, the cfp2 port occupies a large space. Even the physical port space of QSFP28 is twice that of the 10G port. A hash operation is required, and the traffic is relatively balanced when it is transferred out. The problem now is that when 5G devices are launched in batches and data packets need to be redistributed, if the interface of the service module cannot be fully utilized, the number of service boards or interface boards will have to be increased, which will greatly increase the production cost. Not to mention the green network.

In view of this, overcoming the defects of the prior art is an urgent problem to be solved in the technical field.

[Content of the invention]

The technical problem to be solved by the present invention is that in the prior art using the hash value for pairing, when the hash value of the underlying table entry and the port port are not correspondingly equal, the flow rate of a single output port or some output ports will be very large, even exceeding The bandwidth of the interface leads to packet loss, and the traffic on some interfaces is very small, so the bandwidth of the interface cannot be used reasonably. On the other hand, in the prior art using the polling method for pairing, the same source and the same sink for packets cannot be guaranteed, that is, the same session. The traffic cannot be forwarded to the same physical interface or reach the same server, and the end user cannot guarantee the accuracy of the traffic when collecting traffic.

The present invention adopts following technical scheme:

In the first aspect, the present invention provides a method for refreshing a data packet transmission port. After the obtained data packet obtains a corresponding hash value according to a specified hash calculation rule, it is designated to a target output port according to a preset mapping relationship for data packet processing. The output of the method includes:

Obtain the data packet-output port relationship table in multiple adjacent time periods; wherein, the data packet-output port relationship table records the mapping relationship between the number of data packets in each time period and each target output port ;

adding a weighting method for the continuous appearance of the output port to a designated output port, and/or adjusting the data packet-output port relationship table by adding an offset method when the output port is polled;

After each round of adjustment, the total amount of data packets corresponding to each output port is calculated recursively; after one or more rounds of adjustment and calculation, the average degree of the total amount of data packets corresponding to each output port reaches a preset level The weight value and/or offset of the condition is used as a parameter for adjusting the mapping relationship between the data packet and the target output port, so that the newly received data packet is transmitted according to the adjusted mapping relationship.

Preferably, the default weight value corresponding to each port is 1, and each corresponding port cooperates to complete the transmission of each data packet under the respective associated hash value according to a preset polling sequence;

The weight is a natural number. After setting the weight value to n for the first target port, when polling the mapping relationship between the corresponding hash value and the target port, the first target port is located at the first target port and the first target port after the polling sequence of the original first target port. The first target port under n-1 times of polling is inserted between the two target ports, so that the data packets originally polled and output from the second port, the third port, ..., and the n-1th port are all polled by the The first target port is output; further, when the n+1th polling is performed, it will be restored to the second target port for data packet output;

Among them, after completing the polling of the owned target port, when periodically returning to the first target port to output the data packet, the above-mentioned repeating n-1 times of the first target port polling is still performed according to the weight value of n. Then, poll the remaining target ports in sequence;

Wherein, the corresponding default weight value in the original polling sequence is 1, the polling interval between the second target port and the first target port is the default weight value 1; the m target ports whose original default weight value is 1 m polling operations for m target ports are replaced by m+n-1 polling operations for m target ports.

Preferably, the default offset between each output port is 1, and the increase of the offset during the polling of the output port is used for each time after polling for all output ports is performed, in the next cycle When the polling starts, the first target port in the polling of the current cycle is obtained by offsetting the corresponding offset from the first target port in the polling of the previous cycle; The output port forms an offset closed loop. Specifically, after the offset is increased, if the port number exceeds the maximum port number, the corresponding jump to the minimum port is performed to complete the operation of increasing the offset to find the first target port in the current cycle.

Preferably, the data packet-output port relationship table includes a column of index numbers, a column of data packet groups corresponding to each index number, and an output port for outputting each data packet included in the corresponding data group, specifically:

Described index number is the ID value of table item bit width, and its value completes the index establishment of table item from small to large, and as the index number of corresponding data packet grouping;

The data packet grouping is the total amount of data packets output from the same port, which is counted according to a preset time length or a preset total amount of data flow;

The output port for outputting each data packet included in the corresponding data packet is represented in the default original table as each output port number is periodically associated with each index number according to the polling sequence.

Preferably, the total amount of data packets corresponding to each output port after each round of adjustment is calculated recursively, specifically including:

Group the data packets in each data packet-output port relationship table and arrange them in descending order;

Extract one or more data packet groups and corresponding index numbers whose data packets contained in each data packet group are greater than the preset threshold from the arranged data packet groups; The amount of included data packets is used as a discrete relationship curve of the ordinate; thus a corresponding discrete relationship curve is obtained for each data packet-output port relationship table;

Wherein, the increase of the weight value for the specified output port is represented by the compression of the abscissa and the extension of the ordinate corresponding to the specified area of the corresponding discrete relationship curve; wherein the extension of the ordinate is the ordinate value corresponding to the compressed abscissa. to accumulate;

Among them, the increase of the offset is manifested by inserting a vacancy whose ordinate is 0 in the specified area of the discrete relationship curve;

After adjusting the discrete relationship curve corresponding to each data packet-output port relationship table according to the weight value and/or the offset, the discrete relationship corresponding to each analyzed data packet-output port relationship table The curve is fitted, and if the smoothness of the fitted curve reaches the preset condition, the corresponding weight value and/or offset adjustment parameter is used as the weight value and/or offset that meets the preset condition.

Preferably, the data packet-output port relationship table is based on the unit time or the total number of packets as a derivation basis; wherein, if the total number of recorded data packets reaches the total number of packets, then a table of the A data packet-output port relationship table; or, when the recorded data packet grouping time reaches the unit time, a data packet-output port relationship table is derived.

Preferably, the specified hash calculation rule includes one or more of a mac calculation method, a sip+dip binary calculation method, and a sip+dip+sport+dport quaternary calculation method.

Preferably, the timing when the refresh method is triggered includes:

When receiving network management alarms and/or syslog logs, determine whether to adjust the weight value and/or offset according to the bandwidth usage of the output port;

If adjustment is required, use the refresh method to configure the weight value and/or offset of the output port according to the generated optimal cache entry.

In a second aspect, the present invention also provides a device for refreshing a data packet transmission port, which is used to implement the method for refreshing a data packet transmission port described in the first aspect, and the device includes:

at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the processor for The method for refreshing the data packet transmission port described in the first aspect is performed.

In a third aspect, the present invention also provides a non-volatile computer storage medium storing computer-executable instructions, the computer-executable instructions being executed by one or more processors for completing the first The refresh method of the data packet transmission port described in the aspect.

Under the condition that the hash algorithm is fixed, the present invention can not only ensure the same source and the same sink of the data packets, but also ensure the bandwidth of the downlink output port interface by configuring different interface weights and bottom layer offset positions for different current network traffic. maximum use. Regardless of the current network traffic, you can configure and adjust the optimal balance effect through the interface weight + underlying offset brushing method according to the characteristics of the current network traffic.

【Description of drawings】

In order to describe the technical solutions of the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings that need to be used in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and for those of ordinary skill in the art, other drawings can also be obtained from these drawings without creative effort.

1 is a schematic flowchart of a method for refreshing a data packet transmission port according to an embodiment of the present invention;

2 is a schematic diagram of a data packet-output port relationship table provided by an embodiment of the present invention;

3 is a schematic diagram of a data packet-output port relationship table provided by an embodiment of the present invention;

4 is a schematic diagram of a data packet-output port relationship table provided by an embodiment of the present invention;

5 is a schematic diagram of a data packet-output port relationship table provided by an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a device for refreshing a data packet transmission port according to an embodiment of the present invention.

【Detailed ways】

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

In the description of the present invention, the orientation or positional relationship indicated by the terms "inner", "outer", "longitudinal", "lateral", "upper", "lower", "top", "bottom", etc. are based on the drawings The orientation or positional relationship shown is only for the convenience of describing the present invention rather than requiring the present invention to be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention.

Solution 1 of the prior art: through the method of index+hash value+port, traditional hash algorithms (such as mac algorithm, sip+dip binary algorithm, sip+dip+sport+dport quaternary algorithm or self-defined algorithm, etc.), regardless of The advantage of which algorithm is that it can ensure the same source and the same destination for data packets. The disadvantage is that when the current network traffic is relatively balanced (the balance here does not refer to the forwarding balance, the hash value calculated when the traffic is balanced is easy to calculate to a position) , in short, when the current network traffic passes the XOR algorithm, the calculated hash value is in the entry corresponding to the port port. It is very likely that the hash value corresponding to the port entry is uneven, so that when the traffic is transferred out, the Uneven occurs. When the hash value of the underlying table entry does not correspond to the port port, the traffic of a single output port or some of the output ports will be very large, or even exceed the interface bandwidth, resulting in packet loss, while the traffic of some interfaces is very small, resulting in the interface bandwidth cannot be reasonably used. .

Solution 2 of the prior art: the form of index+interface, this method does not look at the hash value, and polls the output. For example, there are 5 output ports, which are 1 to 5 respectively, and the data packets will be polled and output in the 1-5 ports. . The advantage can ensure that the packets are output in a balanced manner, and the interface bandwidth can be used to the maximum. The accuracy of the flow cannot be guaranteed.

In the description of the subsequent embodiments of the present invention, the port and the interface are not strictly distinguished, that is, the corresponding port can be understood as the interface involved in the background technology, and the corresponding interface and the output port in different embodiments are expressed , can be understood as two expressions of the same technical concept.

In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

Embodiment 1:

Embodiment 1 of the present invention provides a method for refreshing a data packet transmission port. After the obtained data packet obtains a corresponding hash value according to a specified hash calculation rule, it is designated to a target output port according to a preset mapping relationship for data packet processing. Output, wherein, the specified hash calculation rule includes one or more of the mac calculation method, the sip+dip binary calculation method, and the sip+dip+sport+dport quaternary calculation method, as shown in Figure 1, Methods include:

In step 201, a data packet-output port relationship table in multiple adjacent time periods is obtained; wherein, the data packet-output port relationship table records the number of data packets in each time period and each target output port the mapping relationship between them.

In step 202, the data packet-output port relationship table is adjusted by adding a weight mode of continuous occurrence of the output port to a designated output port, and/or adding an offset mode when the output port is polled.

In step 203, the total amount of data packets corresponding to each output port after each round of adjustment is calculated recursively; after one or more rounds of adjustment and calculation, the total amount of data packets corresponding to each output port is calculated The weight value and/or offset whose average degree reaches the preset condition is used as a parameter for adjusting the mapping relationship between the data packet and the target output port, so that the newly received data packet is transmitted according to the adjusted mapping relationship.

The timing of triggering the refresh method includes: when receiving a network management alarm and/or syslog log, judge whether the weight value and/or offset need to be adjusted according to the bandwidth usage of the output port; if necessary, according to the generated optimal cache table item, use the refresh method to configure the weight value and/or offset of the output port.

In the embodiment of the present invention, under the condition that the hash algorithm is fixed, for different current network traffic, by configuring different interface weights and bottom layer offset positions, it can not only ensure the same source and the same sink for data packets, but also ensure the downstream output port interface. Maximum utilization of bandwidth. Regardless of the current network traffic, you can configure and adjust the optimal balance effect through the interface weight + underlying offset brushing method according to the characteristics of the current network traffic.

In an optional solution, adding a switch control for whether to generate a packet-output port relationship table, and checking it all the time will consume resources. You can choose to enable the detection switch when the port goes online for the first time or when a port reaches the rate threshold.

As the weight value and offset of the two core elements in the implementation process of the embodiment of the present invention, the implementation mechanism will be explained through a relatively complete theoretical process next:

Aspect 1. Weight value:

The default weight value corresponding to each port is 1, and each corresponding port cooperates to complete the transmission of each data packet under the respective associated hash value according to the preset polling sequence;

Aspect 2. Offset:

The default offset between each output port is 1, and the increase of the offset when the output port is polled is used to start the polling in the next cycle after each polling for all output ports is performed. When the first target port in the polling of the previous cycle is offset by the corresponding offset, the new first target port in the polling of the current cycle is obtained; among them, all the output ports owned constitute An offset closed loop, specifically, if the port number exceeds the maximum port number after the offset is increased, the corresponding jump to the minimum port is performed to complete the operation of increasing the offset to find the first target port in the current cycle.

As shown in FIG. 2 , the data packet-output port relationship table includes a column of index numbers (index column in the figure), a column of data packet groups corresponding to each index number (hash value in the figure), which is used for outputting the corresponding data packets in the The output port of each data packet included (port column in the figure), specifically:

As the core principle used for calculation in the implementation process of the embodiment of the present invention, the total amount of data packets corresponding to each output port after each round of adjustment is calculated in a recursive manner, and the details are as follows:

In the embodiment of the present invention, as shown in FIG. 2 , the data packet-output port relationship table is derived based on the unit time or the total number of packets; wherein, the total number of recorded data packets arrives at the total number of packets number, then derive a data packet-output port relationship table; or, when the recorded data packet grouping time reaches the unit time, then derive a data packet-output port relationship table.

In the embodiment of the present invention, under the condition that the hash algorithm is fixed, for different current network traffic, by configuring different interface weights and bottom layer offset positions, it can not only ensure the same source and the same sink for data packets, but also ensure the downstream output port interface. Maximum utilization of bandwidth. Regardless of the traffic on the live network, you can configure and adjust the optimal balance effect by brushing the interface weight + bottom layer offset according to the traffic characteristics of the live network.

Example 2:

This embodiment of the present invention is a specific presentation of the data packet-output port relationship table involved in Embodiment 1, and for the associated data involved, it corresponds to the specific weight value and offset in the solution implementation process, A presentation example of how to reflect through a corresponding relationship table; in this embodiment of the present invention, the data packet-output port relationship table in Embodiment 1 is also referred to as an underlying table entry.

General example: If the traffic is N packets, the sum of the hash values corresponding to index 0 to 2047 is N, and the corresponding hash values of the port table are as follows. 0 to 4 in the port table correspond to ports 1 to 5 in the port group respectively, then all the hash values corresponding to all 0s in the port table are the traffic output from the 1 interface (here, the 1 interface is mapped with the corresponding port value of 0). , all the hash values corresponding to 1 correspond to the traffic output by interface 2... .

As shown in Figure 2, the flows corresponding to ports 1 to 5 are as follows:

1. The packets output by interface 1 correspond to the hash value corresponding to port table 0, then the number of packets output by 1 is 8123+1638+6782+1157+… (the sum of all positions corresponding to port table 0);

2. The packets output by interface 2 correspond to the hash value corresponding to port table 1, then the number of packets output by 2 is 216+25+14738+24657+… (corresponding to the sum of all positions in port table 1);

3. The packets output by interface 3 correspond to the hash value corresponding to port table 2, then the number of packets output by 3 is 3105+678+13456+… (the sum of all positions corresponding to port table 2);

4. The packets output by interface 4 correspond to the hash value corresponding to port table 3, then the number of packets output by 4 is 3105+678+6782+13456+… (corresponding to the sum of all positions in port table 3);

5. The packets output by interface 5 correspond to the hash value corresponding to port table 4, then the number of packets output by 5 is 2506+987+2345+… (the sum of all positions corresponding to port table 4);

Based on the refresh method given in steps 201-203 of Embodiment 1, it can be subdivided into at least the following three schemes.

Option 1: Configure Interface Weights

The bit width of the underlying table entry is 2048, and there are 5 downlink output ports in the port group. For example, configure the interface weight. The default interface weight is 1. The second table, the corresponding table entry of the bottom port port is 001234~001234~001234... A total of 2048 table entries are cycled;

For example, the configuration weight of port 2 is 3, and the weight of port 4 is configured to be 5, that is, the port table corresponding to port 2 will be refreshed 3 times, and the port table corresponding to port 4 will be refreshed 5 times, then the corresponding entry of the underlying port port is 01112344444~ 01112344444～01112344444...... A total of 2048 entries are circulated;

Scenario 2: Offset of underlying table entry

The bit width of the bottom table entry is 2048, and there are 5 downstream output ports in the port group. For example, if the bottom table entry offset is 2, that is, after each round of table entries is flushed, the next position will be shifted to the right by 2 If the table starts to be refreshed from the first position, the corresponding table entries of the bottom port are 01234~23401~43012~01234... A total of 2048 table entries are circulated.

The bit width of the bottom table entry is 2048, and there are 5 downstream output ports in the port group. For example, if the bottom table entry offset is 3, that is, after each round of table entries is flushed, the next position will be shifted to the right by 2 If the table starts to be refreshed from the first position, the corresponding table entries of the bottom port are 01234~34012~12340~40123... A total of 2048 table entries are circulated.

Scheme 3: Interface weight + underlying table entry offset (ie, the combination of weight value and offset in Example 1)

The bit width of the underlying table entry is 2048, and there are 5 downlink output ports in the port group as an example. For example, the port weight of port 1 is 3 and the offset is 2, that is, the port table corresponding to port 1 will be flushed 3 times, and each flush is completed. After one round of entries, the next position will start to refresh the table at two positions to the right, and the corresponding entries of the bottom port are 0001234～0123400～2340001～..... A total of 2048 entries are circulated.

When the

technical solutions

1 and 2 are used alone, some problems can be solved compared with the conventional method of brushing the table. When the

solutions

1 and 2 are used in combination, the effect will be better, and the balance of data packets can be better guaranteed.

As shown in Figure 4, the bit width of the bottom table entry is 2048, and there are 5 downlink output ports in the port group. The brush table is shown in Figure 4.

If the traffic is N packets, the sum of the hash values corresponding to indexes 0 to 2047 is N, and the corresponding hash values of the port table are as follows. 0 to 4 in the port table correspond to ports 1 to 5 in the port group respectively, then all the hash values corresponding to all 0s in the port table are the traffic output by port 1, and all hash values corresponding to 1 are the traffic output by port 2... …

As shown in Figure 5, the flows corresponding to ports 1 to 5 are as follows:

1. The packets output by interface 1 correspond to the hash value corresponding to port table 0, then the number of packets output by 1 is 8123+216+11003+456+222849+… (the sum of all positions corresponding to port table 0);

2. The packets output by interface 3 correspond to the hash value corresponding to port table 1, then the number of packets output by 2 is 3105+678+… (the sum of all positions corresponding to port table 1);

3. The packets output by interface 3 correspond to the hash value corresponding to port table 1, then the number of packets output by 2 is 2506+987+2345+… (the sum of all positions corresponding to port table 1);

4. The packets output by interface 4 correspond to the hash value corresponding to port table 2, then the number of packets output by 3 is 1638+6782+1157+… (the sum of all positions corresponding to port table 2);

5. The packets output by interface 5 correspond to the hash value corresponding to port table 4, then the number of packets output by 5 is 25+14738+24567+… (the sum of all positions corresponding to port table 4);

The solution of this embodiment of the present invention is based on the unbalanced forwarding of the outgoing interface traffic in the current network traffic, and when the total forwarding bandwidth is within 90% of the total bandwidth of the physical interface, (why is not 100%, because no matter how the adjustment is made, there will be deviations, including that the traffic may have Burst, forwarding delay, etc., 1 to 1 bandwidth forwarding cannot be guaranteed, 90% of which is already a very high bandwidth utilization of the output interface). , try to balance the output of each output port, ensure the reasonable utilization of the bandwidth of each physical output port to the greatest extent, and ensure that the current network traffic is forwarded in the same source and the same destination under the condition of balance. By adjusting the forwarding table entry, the problem of packet loss caused by uneven traffic can be effectively solved without increasing the output port. In short, the optimal balanced forwarding effect is adjusted through the underlying table items, and the bandwidth of all output interfaces is maximized to achieve the purpose of saving the number of output ports, and further achieve the purpose of saving boards, and ultimately achieve no waste and savings. cost purpose.

Description of the cache entries generated by the software in cooperation with the logic: the number of hash values corresponding to each port per unit time, the software calculates several optimal entries according to the number of hash values. The software calculates the permutation and combination that can be set for each interface weight. For example, the interface weight can be configured in the range of 1 to 5, and the interface weight is 1 1 1 1 1, 1 1 1 1 2, 1 1 1 1 3... all the way to 5 5 5 5 5 A total of 32 entries. Then brush the table in combination with the offset. For example, the configurable offset is in the range of 1 to 5. With the interface weight brushing table, the software generates a total of 32*5=160 table entries corresponding relationships. The output traffic conditions are compared, and the first N (N can be matched) optimal table entry caches are selected, and the intersection of multiple optimal table entries per unit time is combined to finally select the same optimal table entry. The last entry can be used as the basis for us to configure the interface weight and offset. This table is only used to find the relationship between the hash value and the port table, and it is not really sent to the bottom layer. Why not use automatic adjustment? Considering the large fluctuation of the current network traffic, if it is adjusted frequently, it will violate the packet synchronization. The forwarding principle is based on the same source and the same sink. The current solution uses the form of alarms or logs. When the difference between the maximum rate and the minimum traffic rate of the member interfaces of a port group exceeds the configured threshold, or the rate of a single interface exceeds the bandwidth, the device sends an alarm or records a log. When the operation and maintenance personnel receive the network management alarm and syslog log, they can judge whether the underlying table entry needs to be adjusted according to the bandwidth usage of the interface. Interface weight and underlying table entry offset. Whether or not to generate bottom-level entries can also be controlled by switches. When not needed, there is no need to waste software resources. The purpose of the invention is to make the bandwidth of all downlink output port members in the port group under the premise of ensuring the same source and same sink for data packets. Get maximum utilization, under the condition of a certain bandwidth, effectively solve the problem of excessive bandwidth packet loss on the interface caused by uneven traffic output, achieve the purpose of saving the number of back-end output ports, and achieve the purpose of saving the physical space of the service card, so as to achieve The purpose of saving cost and reducing power consumption.

Example 3:

As shown in FIG. 6 , it is a schematic structural diagram of an apparatus for refreshing a data packet transmission port according to an embodiment of the present invention. The device for refreshing the data packet transmission port in this embodiment includes one or more processors 21 and a memory 22 . Among them, one processor 21 is taken as an example in FIG. 6 .

The processor 21 and the memory 22 may be connected through a bus or in other ways, and the connection through a bus is taken as an example in FIG. 6 .

As a non-volatile computer-readable storage medium, the memory 22 can be used to store non-volatile software programs and non-volatile computer-executable programs, such as the method for refreshing a data packet transmission port in Embodiment 1. The processor 21 executes the refreshing method of the data packet transmission port by running the non-volatile software programs and instructions stored in the memory 22 .

Memory 22 may include high speed random access memory, and may also include nonvolatile memory, such as at least one magnetic disk storage device, flash memory device, or other nonvolatile solid state storage device. In some embodiments, memory 22 may optionally include memory located remotely from processor 21, which may be connected to processor 21 via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The program instructions/modules are stored in the memory 22, and when executed by the one or more processors 21, execute the method for refreshing the data packet transmission port in the above-mentioned Embodiment 1, for example, execute the above-described diagram. 1 for each step shown.

It is worth noting that the information exchange and execution process between the modules and units in the above-mentioned device and the system are based on the same concept as the processing method embodiments of the present invention. For details, please refer to the descriptions in the method embodiments of the present invention. , and will not be repeated here.

Those skilled in the art can understand that all or part of the steps in the various methods of the embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium can include: Read memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or CD, etc.

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention shall be included in the protection of the present invention. within the range.

Claims

A method for refreshing a data packet transmission port, characterized in that, after the obtained data packet obtains a corresponding hash value according to a specified hash calculation rule, it is designated to a target output port according to a preset mapping relationship to output the data packet, and the method include:

Obtain the data packet-output port relationship table in multiple adjacent time periods; wherein, the data packet-output port relationship table records the mapping relationship between the number of data packets in each time period and each target output port ;

adding a weighting method for the continuous appearance of the output port to a designated output port, and/or adjusting the data packet-output port relationship table by adding an offset method when the output port is polled;

After each round of adjustment, the total amount of data packets corresponding to each output port is calculated recursively; after one or more rounds of adjustment and calculation, the average degree of the total amount of data packets corresponding to each output port reaches a preset level The weight value and/or offset of the condition is used as a parameter for adjusting the mapping relationship between the data packet and the target output port, so that the newly received data packet is transmitted according to the adjusted mapping relationship.
The method for refreshing a data packet transmission port according to claim 1, wherein the default weight value corresponding to each port is 1, and each corresponding port cooperates to complete the corresponding hash value according to a preset polling sequence. Transmission of individual data packets;

The weight is a natural number. After setting the weight value to n for the first target port, when polling the mapping relationship between the corresponding hash value and the target port, the first target port is located at the first target port and the first target port after the polling sequence of the original first target port. The first target port under n-1 times of polling is inserted between the two target ports, so that the data packets originally polled and output from the second port, the third port, ..., and the n-1th port are all polled by the The first target port is output; further, when the n+1th polling is performed, it will be restored to the second target port for data packet output;

Among them, after completing the polling of the owned target port, when periodically returning to the first target port to output the data packet, the above-mentioned repeating n-1 times of the first target port polling is still performed according to the weight value of n. After that, poll the remaining target ports in sequence;

Wherein, the corresponding default weight value in the original polling sequence is 1, the polling interval between the second target port and the first target port is the default weight value 1; m target ports with the original default weight value of 1 m polling operations for m target ports are replaced by m+n-1 polling operations for m target ports.
The method for refreshing a data packet transmission port according to claim 1, wherein the default offset between each output port is 1, and the increase of the offset during the polling of the output port is used for every time the output port is polled. After performing the polling for all output ports, at the beginning of the polling of the next cycle, the first target port in the polling of the previous cycle is offset by the corresponding offset to get the current cycle of the The new first target port in the query; in which, all the output ports owned form an offset closed loop, specifically, if the port number exceeds the maximum port number after the offset is increased, the corresponding jump to the minimum port is completed to complete the corresponding Increment the offset to find the operation to the first target port of the current cycle.
The method for refreshing a data packet transmission port according to claim 1, wherein the data packet-output port relationship table includes a column of index numbers, a column of data packet groups corresponding to each index number, and a column for outputting corresponding data packets. The output port of each data packet contained in, specifically:

Described index number is the ID value of table item bit width, and its value completes the index establishment of table item from small to large, and as the index number of corresponding data packet grouping;

The data packet grouping is the total amount of data packets output from the same port, which is counted according to the preset time length or the preset total amount of data flow;

The output port for outputting each data packet included in the corresponding data packet is represented in the default original table as each output port number is periodically associated with each index number according to the polling sequence.
The method for refreshing a data packet transmission port according to claim 4, characterized in that, after each round of adjustment is calculated recursively, the total amount of data packets corresponding to each output port specifically includes:

Group the data packets in each data packet-output port relationship table and arrange them in descending order;

Extract one or more data packet groups and corresponding index numbers whose amount of data packets contained in each data packet group is greater than the preset threshold from the arranged data packet groups; The amount of included data packets is used as a discrete relationship curve of the ordinate; thus a corresponding discrete relationship curve is obtained for each data packet-output port relationship table;

Wherein, the increase of the weight value for the specified output port is represented by the compression of the abscissa and the extension of the ordinate corresponding to the specified area of the corresponding discrete relationship curve; wherein, the extension of the ordinate is the ordinate value corresponding to the compressed abscissa. to accumulate;

Among them, the increase of the offset is manifested by inserting a vacancy whose ordinate is 0 in the specified area of the discrete relationship curve;

After adjusting the discrete relationship curve corresponding to each data packet-output port relationship table according to the weight value and/or offset, the discrete relationship corresponding to each analyzed data packet-output port relationship table The curve is fitted, and if the smoothness of the fitted curve reaches a preset condition, the corresponding weight value and/or offset adjustment parameter is used as the weight value and/or offset that meets the preset condition.
The method for refreshing a data packet transmission port according to claim 4, wherein the data packet-output port relationship table is based on the unit time or the total number of packets as a derivation basis; When the total number of packets is reached, a data packet-output port relationship table is derived; or, when the recorded packet grouping time reaches the unit time, a data packet-output port relationship table is derived .
The method for refreshing a data packet transmission port according to any one of claims 1-6, wherein the specified hash calculation rule includes a mac calculation method, a sip+dip binary calculation method, and a sip+dip+sport+ One or more of the dport quaternary calculation methods.
The refresh method of the data packet transmission port according to any one of claims 1-6, wherein the timing of the refresh method triggering comprises:

When receiving network management alarms and/or syslog logs, determine whether the weight value and/or offset need to be adjusted according to the bandwidth usage of the output port;

If adjustment is required, use the refresh method to configure the weight value and/or offset of the output port according to the generated optimal cache entry.
A device for refreshing a data packet transmission port, characterized in that the device comprises:

at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the processor for The method for refreshing the data packet transmission port according to any one of claims 1-8 is performed.