CN116846836A

CN116846836A - Link load adjusting method and device, storage medium and electronic device

Info

Publication number: CN116846836A
Application number: CN202310806355.9A
Authority: CN
Inventors: 余学山; 杨利进; 吴仲阳; 杨飘飘
Original assignee: Industrial and Commercial Bank of China Ltd ICBC
Current assignee: Industrial and Commercial Bank of China Ltd ICBC
Priority date: 2023-06-30
Filing date: 2023-06-30
Publication date: 2023-10-03

Abstract

The application discloses a method and a device for adjusting link load, a storage medium and an electronic device. Relates to the field of financial science and technology, and the method comprises the following steps: detecting a target network attribute of an initial network, wherein the target network attribute is used for indicating the possibility of load imbalance of a physical link in the initial network, and the initial network is a network which relies on an equivalent multipath routing algorithm to perform load balancing; under the condition that the attribute of the target network falls into the target threshold range, detecting the load information of each physical link in N physical links included in the initial network to obtain a load information set; and regulating the load of the physical link in the initial network to an equilibrium state according to the load information set to obtain a target network. The application solves the problems of poor link load adjusting effect and the like in the related technology.

Description

Link load adjusting method and device, storage medium and electronic device

Technical Field

The present application relates to the field of financial science and technology, and in particular, to a method and apparatus for adjusting link load, a storage medium, and an electronic apparatus.

Background

In a network which relies on an ECMP (Equal-cost-Cost Multipath Routing) algorithm to perform load balancing, although network redundancy and reliability can be improved to a certain extent, because the ECMP algorithm records flow statistics information based on a flow table in the process of performing load balancing and dynamically adjusts link load balancing according to the flow statistics information, a certain scale is required to be reached for the flow number transmitted by a link in the network to achieve a better network balancing effect, when the flow number is small, a hash collision problem may exist, so that the network balancing effect is poor, the situation that partial links are unbalanced in load and partial links are overloaded is caused even though the network is subjected to load balancing by the ECMP algorithm, the phenomenon that the resource utilization rate of the links in the network is low and the risk of link congestion is increased is caused even if the data is lost due to the link congestion is caused.

Aiming at the problems of poor link load adjusting effect and the like in the related technology, no effective solution is proposed at present.

Disclosure of Invention

The application mainly aims to provide a method and a device for regulating link load, a storage medium and an electronic device, so as to solve the problems of poor regulating effect of the link load and the like in the related technology.

In order to achieve the above object, according to one aspect of the present application, there is provided a method of adjusting a link load.

The method comprises the following steps:

detecting a target network attribute of an initial network, wherein the target network attribute is used for indicating the possibility of load imbalance of a physical link in the initial network, and the initial network is a network which relies on an equivalent multipath routing algorithm to perform load balancing;

under the condition that the target network attribute falls into a target threshold range, detecting load information of each physical link in N physical links included in the initial network to obtain a load information set, wherein the load information is used for indicating the load of the corresponding physical link, and the network attribute falls into the target threshold range and is used for indicating that the possibility of the occurrence of the load imbalance condition of the physical link in the network is greater than the target possibility;

and regulating the load of the physical link in the initial network to an equilibrium state according to the load information set to obtain a target network.

Optionally, the detecting the target network attribute of the initial network includes:

detecting the number of data streams which meet a target condition and are transmitted in the initial network to obtain a target number, wherein the data streams which meet the target condition are data streams with the transmitted data amount being larger than the target data amount;

And determining the target network attribute of the initial network according to the target quantity.

Optionally, the detecting the number of data flows transmitted in the initial network and meeting the target condition, to obtain the target number includes:

detecting quintuple information of each data stream in the initial network, wherein the quintuple information is used for indicating the service type of the corresponding data stream;

determining a service with a service type being a target service type as a target service according to the quintuple information, wherein the data volume on which the service belonging to the target service type depends in the execution process is larger than the target data volume;

and extracting the number of the data streams corresponding to the target service to obtain the target number.

Optionally, the determining the target network attribute of the initial network according to the target number includes:

determining the target network attribute as a first attribute value under the condition that the target number is smaller than a target number threshold, wherein the first attribute value falls into the target threshold range;

and under the condition that the target number is greater than or equal to a target number threshold, determining the target network attribute as a second attribute value, wherein the second attribute value does not fall into the target threshold range.

Optionally, the adjusting the load of the physical link in the initial network to an equilibrium state according to the load information set to obtain a target network includes:

determining the link type of each physical link in the N physical links according to the load information set, wherein the link type of the physical link comprises: a high load type and a low load type, wherein the load of a physical link belonging to the high load type is larger than a first load threshold value, and the load of a physical link belonging to the low load type is smaller than a second load threshold value;

and migrating the data flow in the physical link belonging to the high load type to the physical link belonging to the low load type to obtain the target network.

Optionally, the determining, according to the load information set, a link type of each of the N physical links includes:

generating a current load average value of the initial network according to the load information set, wherein the load average value is an average value of loads indicated by all the load information in the load information set;

determining a link type of a physical link for which the load information is used for indicating that the load exceeds the first load threshold as the high load type, wherein the first load threshold is larger than the load average value;

And determining the link type of the physical link with the load smaller than the load average value, which is used for indicating the load information, as the low load type.

Optionally, the migrating the data flow in the physical link belonging to the high load type to the physical link belonging to the low load type to obtain the target network includes:

extracting a migration data stream from the physical links belonging to the high load type, wherein after the physical links belonging to the high load type migrate out of the migration data stream, the load of the physical links belonging to the high load type is smaller than the load average value;

and migrating the migration data flow from the physical link belonging to the high load type to the corresponding physical link belonging to the low load type until the load of the physical link in the initial network is in an equilibrium state, so as to obtain the target network.

Optionally, the migrating the migration data flow from the physical link belonging to the high load type to the corresponding physical link belonging to the low load type includes:

sequentially sequencing the N physical links according to the load size to obtain a target load sequence, wherein the load sizes of the N physical links recorded in the target load sequence are sequentially reduced;

And migrating the migration data flow in the physical link belonging to the high load type in the positive sequence i to the physical link belonging to the low load type in the reverse sequence i, wherein i is a positive integer which is greater than or equal to 1 and less than or equal to N.

In order to achieve the above object, according to another aspect of the present application, there is provided an adjusting apparatus of a link load.

The device comprises:

the first detection module is used for detecting target network attributes of an initial network, wherein the target network attributes are used for indicating the possibility of load imbalance of a physical link in the initial network, and the initial network is a network which relies on an equivalent multipath routing algorithm to perform load balancing;

the second detection module is used for detecting the load information of each physical link in the N physical links included in the initial network to obtain a load information set under the condition that the target network attribute falls into a target threshold range, wherein the load information is used for indicating the load of the corresponding physical link, and the network attribute falls into the target threshold range and is used for indicating that the possibility of the occurrence of the load imbalance condition of the physical link in the network is greater than the target possibility;

And the adjusting module is used for adjusting the load of the physical link in the initial network to an equilibrium state according to the load information set to obtain a target network.

According to the application, the following steps are adopted: detecting a target network attribute of an initial network which relies on an equivalent multipath routing algorithm to perform load balancing, wherein the target network attribute is used for indicating the possibility of the occurrence of load imbalance of physical links in the initial network, when the target network attribute falls into a target threshold range, namely, when the possibility of the occurrence of load imbalance of the physical links in the network is greater than the target possibility, detecting the load information of each physical link in N physical links included in the initial network to obtain a load information set, wherein the load information is used for indicating the load of the corresponding physical link, finally adjusting the load of the physical links in the initial network to an equilibrium state according to the load information set to obtain a target network, and judging whether the possibility of the occurrence of load imbalance of the physical links in the network is greater than the target possibility according to the target network attribute. Thereby achieving the technical effect of improving the link load adjusting effect.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

fig. 1 is a flowchart of a method for adjusting link load according to an embodiment of the present application;

FIG. 2 is a schematic diagram of a high load type and a low load type provided in accordance with an embodiment of the present application;

FIG. 3 is a schematic diagram of a first load threshold and a second load threshold provided in accordance with an embodiment of the present application;

FIG. 4 is a schematic diagram of a migration data flow provided in accordance with an embodiment of the present application;

FIG. 5 is a schematic diagram of migration of a migration data stream provided according to an embodiment of the present application;

fig. 6 is a schematic diagram of a link load adjustment procedure provided according to an embodiment of the present application;

FIG. 7 is a schematic diagram of a link load adjustment system provided in accordance with an embodiment of the present application;

FIG. 8 is a schematic diagram of a link load adjustment device according to an embodiment of the application;

fig. 9 is a schematic diagram of an electronic device according to an embodiment of the application.

Detailed Description

It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.

In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the application herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

First, partial terms or terminology appearing in the course of describing the embodiments of the application are applicable to the following explanation:

ECMP algorithm: the Equal-cost multipath routing algorithm records flow statistics based on the flow table, and dynamically adjusts link load balancing according to the flow statistics, namely, a plurality of paths with Equal cost reaching the same destination address exist. When the device supports equivalent routing, three layers of forwarding traffic addressed to the destination IP (Internet Protocol ) or destination network segment can be shared by different paths, so as to realize load balancing of network links, and realize fast switching when the links fail. The load balancing algorithm is the most commonly used load balancing algorithm at present, is suitable for more scenes with streaming links, has the advantages of no disorder, has the disadvantage of small number of streaming, has the HASH collision problem in AI (Artificial Intelligence ) training scenes, and has poor network balancing effect.

DCI: data Center Interconnect, the data center interconnection is a network solution for realizing network interconnection and interworking across the data centers, has the characteristics of flexible interconnection, high efficiency and safety, operation and maintenance simplification and the like, and meets the scene requirements of high-efficiency data exchange, disaster recovery and the like between the data centers. The data center interconnection links can be classified into:

1) Short-range interconnect link: within 5 km, the interconnection of the data centers of the parks is generally realized by adopting comprehensive wiring;

2) Medium interconnect link: within 80 km, it is generally meant that optical modules are used to achieve interconnection in adjacent cities or intermediate geographic locations;

3) Long-range interconnect link: thousands of kilometers, generally refers to the use of optical transmission equipment to implement long distance data center interconnections, such as undersea optical fiber cable networks.

It should be noted that, the link load adjusting method provided by the present application is applicable to a wide range of scenarios, and may include, but not limited to: in the following embodiments, the method for adjusting the link load may be described by taking the DCI long-distance interlink scenario as an example, but the usage scenario is not limited, and any network that uses the ECMP algorithm to perform load balancing may be applied to the method for adjusting the link load provided by the present application.

It should be noted that, related information (including, but not limited to, user equipment information, user personal information, etc.) and data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present disclosure are information and data authorized by a user or sufficiently authorized by each party. For example, an interface is provided between the system and the relevant user or institution, before acquiring the relevant information, the system needs to send an acquisition request to the user or institution through the interface, and acquire the relevant information after receiving the consent information fed back by the user or institution.

The present application will be described with reference to preferred implementation steps, and fig. 1 is a flowchart of a method for adjusting link load according to an embodiment of the present application, as shown in fig. 1, where the method includes the following steps:

step S102, detecting a target network attribute of an initial network, wherein the target network attribute is used for indicating the possibility of load imbalance of a physical link in the initial network, and the initial network is a network which relies on an equivalent multipath routing algorithm to perform load balancing;

step S104, under the condition that the target network attribute falls into a target threshold range, detecting the load information of each physical link in N physical links included in the initial network to obtain a load information set, wherein the load information is used for indicating the load of the corresponding physical link, and the network attribute falls into the target threshold range and is used for indicating that the possibility of the occurrence of the load imbalance condition of the physical link in the network is greater than the target possibility;

and step S106, regulating the load of the physical link in the initial network to be in an equilibrium state according to the load information set to obtain a target network.

Through the steps, the target network attribute of the initial network which depends on the equal-cost multipath routing algorithm for load balancing is detected, wherein the target network attribute is used for indicating the possibility of the occurrence of the load imbalance condition of the physical links in the initial network, when the target network attribute falls into a target threshold range, namely, when the possibility of the occurrence of the load imbalance condition of the physical links in the network is larger than the target possibility, the load information of each physical link in N physical links included in the initial network is detected, a load information set is obtained, wherein the load information is used for indicating the load of the corresponding physical link, finally, the load of the physical links in the initial network is adjusted to an equilibrium state according to the load information set, so that the target network is obtained, the target network attribute of the initial network can be actively detected, whether the possibility of the occurrence of the load imbalance condition of the physical links in the network is larger than the target possibility is judged according to the target network attribute, and when the possibility of the occurrence of the load imbalance condition of the physical links in the network is larger than the target possibility is judged, the load information of the physical links in the initial network is adjusted to the state according to the load information set, and therefore the load balancing effect of the physical links in the initial network is poorer than the relevant network is obtained, and the load balancing effect in the network is solved. Thereby achieving the technical effect of improving the link load adjusting effect.

In the technical solution provided in the step S102, the initial network is a network that relies on the equal cost multipath routing algorithm to perform load balancing, that is, the initial network uses the ECMP algorithm to perform load balancing, after the initial network performs load balancing through the ECMP algorithm, the load balancing effect may be poor due to the smaller scale of the traffic in the initial network, and by the method for adjusting the link load provided by the present application, the condition of uneven load can be greatly avoided, so that the load of the link in the network is in an balanced state, and the network resource of the link is utilized to the greatest extent.

Optionally, in this embodiment, the target network attribute is used to indicate the possibility that the load imbalance occurs in the physical link in the initial network, that is, according to the target network attribute of the initial network, the probability that the load imbalance occurs in the link in the initial network at present can be known, when the target network attribute indicates that the probability that the load imbalance occurs is relatively high, the load of the link in the initial network can be actively adjusted, so that the load in the network is ensured to be in an equilibrium state, and readjustment when the load is severely unbalanced is avoided, and at this time, transmission loss of link data may occur.

In one exemplary embodiment, the target network attribute of the initial network may be, but is not limited to, detected by: detecting the number of data streams which meet a target condition and are transmitted in the initial network to obtain a target number, wherein the data streams which meet the target condition are data streams with the transmitted data amount being larger than the target data amount; and determining the target network attribute of the initial network according to the target quantity.

Optionally, in this embodiment, the above-mentioned method may know the probability of the occurrence of the load imbalance condition of the link in the initial network according to the target network attribute of the initial network, and when the target network attribute indicates that the probability of the occurrence of the load imbalance condition is greater, may actively adjust the load of the link in the initial network, which is described in the following detection manner for the target network attribute: in the foregoing process of carrying out load balancing by using the ECMP algorithm, traffic statistics information is recorded based on a flow table, and link load balancing is dynamically adjusted according to the traffic statistics information, so that a good network balancing effect can be achieved only by needing to achieve a certain scale of the number of flows transmitted by links in a network, when the number of flows is small, there may be a problem of hash collision, so that the network balancing effect is poor, especially when the number of large flows in the links of an initial network is small, the load unevenness will be obvious, wherein the large flows refer to the large flows of the data volume, therefore, in the link load adjusting method provided by the present application, the number of data flows which meet the target condition and are transmitted in the initial network can be detected to obtain the target number, and it is worth noting that the number of data flows which meet the target condition is larger than the target data amount is detected, namely, the target number is obtained by detecting the number of large flows transmitted in the initial network, the possibility that the initial network has the load imbalance condition can be indicated to a certain extent, and when the target number is small, namely the initial network has the large number of links and the load unevenness is small.

In one exemplary embodiment, the number of data flows transmitted in the initial network that meet a target condition may be, but is not limited to, detected by: detecting quintuple information of each data stream in the initial network, wherein the quintuple information is used for indicating the service type of the corresponding data stream; determining a service with a service type being a target service type as a target service according to the quintuple information, wherein the data volume on which the service belonging to the target service type depends in the execution process is larger than the target data volume; and extracting the number of the data streams corresponding to the target service to obtain the target number.

Optionally, in this embodiment, the above-mentioned detection manner of the target network attribute is implemented by detecting the number of data flows that meet the target condition and are transmitted in the initial network, to obtain a target number, where the data flows that meet the target condition, that is, the data flows whose data amount is greater than the target data amount, and the detection manner of the target number is to directly detect the data amounts of all the traffic currently transmitted in the link of the initial network, determine the data flows whose data amounts are greater than the target data amount as the data flows that meet the target condition, and then count the number of the data flows that meet the target condition to obtain the target number;

Another way is to detect five-tuple information of each data flow in the initial network, where the five-tuple information may indicate a service type to which the corresponding data flow belongs; because various services exist in the initial network, the traffic corresponding to each service generally has certain characteristics, for example, the data volume of the data stream corresponding to the services such as AI training, bulk storage, data backup and the like is larger, therefore, the data stream corresponding to the services such as AI training, bulk storage, data backup and the like can be regarded as the data stream meeting the target condition, and the target quantity can be obtained by counting the quantity of the data corresponding to the services such as AI training, bulk storage, data backup and the like, that is, the service with the service type being the target service type is determined as the target service according to five-tuple information, wherein the data volume depending on the service belonging to the target service type in the execution process is larger than the target data volume; and extracting the number of data streams corresponding to the target service to obtain the target number, wherein the target service types comprise AI training, massive storage, data backup and other service types.

In one exemplary embodiment, the target network attribute of the initial network may be, but is not limited to, determined from the target number by: determining the target network attribute as a first attribute value under the condition that the target number is smaller than a target number threshold, wherein the first attribute value falls into the target threshold range; and under the condition that the target number is greater than or equal to a target number threshold, determining the target network attribute as a second attribute value, wherein the second attribute value does not fall into the target threshold range.

Optionally, in this embodiment, in a case where the target number is smaller than the target number threshold, the target network attribute is determined to be a first attribute value, where the first attribute value falls within the target threshold range, that is, in a case where the target number is smaller than the target number threshold, it indicates that the number of large traffic in the links of the initial network is smaller, and the possibility that load unevenness occurs is greater, so in order to avoid link load imbalance in the network, load in the network needs to be adjusted.

Optionally, in this embodiment, when the target number is greater than or equal to the target number threshold, the target network attribute is determined to be a second attribute value, where the first attribute value does not fall within the target threshold range, that is, when the target number is greater than or equal to the target number threshold, it indicates that the number of large traffic in the links of the initial network is greater, and the possibility of occurrence of load unevenness is less, and no adjustment is required for the load in the network.

In the technical solution provided in step S104, N load information is recorded in the load information set, where each load information is used to indicate a load of a corresponding physical link, where the load may be, but is not limited to, presented in a form of a load rate, and when a network attribute falls within the target threshold range, the possibility of indicating that a load imbalance condition occurs in a physical link in a network is greater than a target possibility, and at this time, the load of each of N physical links included in an initial network is adjusted according to the N load information recorded in the load information set, so as to implement load balancing.

In the technical solution provided in step S106, the load of the physical link in the initial network is adjusted to an equilibrium state according to the load information set to obtain the target network, where the adjustment process may be single-round adjustment or multi-round adjustment, that is, the initial network may still be in an unbalanced state after one round of adjustment, at this time, the same adjustment step may be repeatedly performed to perform the next round of adjustment, and finally the target network with balanced load is obtained.

In an exemplary embodiment, the load of the physical link in the initial network may be adjusted to an equilibrium state according to the load information set, and the target network may be obtained by the following ways: determining the link type of each physical link in the N physical links according to the load information set, wherein the link type of the physical link comprises: a high load type and a low load type, wherein the load of a physical link belonging to the high load type is larger than a first load threshold value, and the load of a physical link belonging to the low load type is smaller than a second load threshold value; and migrating the data flow in the physical link belonging to the high load type to the physical link belonging to the low load type to obtain the target network.

Optionally, in this embodiment, fig. 2 is a schematic diagram of a high load type and a low load type provided according to an embodiment of the present application, as shown in fig. 2, taking an N value of 6 as an example, 6 physical links (link 1 to link 6) exist in an initial network, where a load information set describes loads (load rates) of the 6 physical links, and the loads are respectively: link 1 (45%), link 2 (65%), link 3 (40%), link 4 (20%), link 5 (70%), and link 6 (60%), the first and second load thresholds are shown, wherein the physical links with loads greater than the first load threshold include: links 2, 5 and 6, i.e. links 2, 5 and 6 are of the high load type, the physical links with loads less than the second load threshold include: links 1, 3 and 4, i.e., links 1, 3 and 4, are of a low load type.

Alternatively, in the present embodiment, the first load threshold may be, but is not limited to being, greater than or equal to the second load threshold.

In one exemplary embodiment, the link type of each of the N physical links may be determined from the set of load information by, but not limited to: generating a current load average value of the initial network according to the load information set, wherein the load average value is an average value of loads indicated by all the load information in the load information set; determining a link type of a physical link for which the load information is used for indicating that the load exceeds the first load threshold as the high load type, wherein the first load threshold is larger than the load average value; and determining the link type of the physical link with the load smaller than the load average value, which is used for indicating the load information, as the low load type.

Optionally, in this embodiment, fig. 3 is a schematic diagram of a first load threshold and a second load threshold provided according to an embodiment of the present application, as shown in fig. 3, taking an N value of 6 as an example, 6 physical links (link 1 to link 6) exist in an initial network, where a load information set describes loads (load rates) of the 6 physical links, and the loads are respectively: the purpose of this is that after a high load type of link is subjected to load adjustment, the load rate of the physical link is reduced to about the average value of the load, for example, the original load rate of the link 2 is 65%, and after load adjustment, the link 2 reaches 52% to be considered as successful, because if the first load threshold and the second load threshold are both consistent with 50%, then a small amount of traffic will still need to be migrated out of a part, possibly involving packet migration, with a risk of data loss or error, and therefore a certain margin is reserved above the load average value, because the average load value of the load indicated by all load information in the load information set is 50%, the second load threshold may be 50%, and the first load threshold is greater than the average load value, and the second load threshold may be 58%.

In one exemplary embodiment, the target network may be obtained, but is not limited to, by migrating data flows in physical links belonging to the high load type to physical links belonging to the low load type by: extracting a migration data stream from the physical links belonging to the high load type, wherein after the physical links belonging to the high load type migrate out of the migration data stream, the load of the physical links belonging to the high load type is smaller than the load average value; and migrating the migration data flow from the physical link belonging to the high load type to the corresponding physical link belonging to the low load type until the load of the physical link in the initial network is in an equilibrium state, so as to obtain the target network.

Optionally, in this embodiment, fig. 4 is a schematic diagram of migration data flow provided according to an embodiment of the present application, as shown in fig. 4, there are 6 physical links (link 1 to link 6) in the initial network, where the load information set describes loads (load rates) of the 6 physical links, and the loads are respectively: link 1 (45%), link 2 (65%), link 3 (40%), link 4 (20%), link 5 (70%) and link 6 (60%), the second load threshold is 50%, the first load threshold is 58%, link 2, link 5 and link 6 belong to a high load type, and the migration data stream is extracted from the physical links belonging to the high load type, wherein after the physical links belonging to the high load type migrate out of the migration data stream, the load of the physical links belonging to the high load type is less than the load average value, for example, after the link 2 (65%) migrates out of the migration data stream, the load is less than 50% of the load average value.

Optionally, in this embodiment, a migration data stream is extracted from a physical link belonging to the high load type, where after the physical link belonging to the high load type migrates out of the migration data stream, the load of the physical link belonging to the high load type is smaller than a first load threshold.

In one exemplary embodiment, the migration data flow may be migrated from a physical link belonging to the high load type to a corresponding physical link belonging to the low load type by, but not limited to: sequentially sequencing the N physical links according to the load size to obtain a target load sequence, wherein the load sizes of the N physical links recorded in the target load sequence are sequentially reduced; and migrating the migration data flow in the physical link belonging to the high load type in the positive sequence i to the physical link belonging to the low load type in the reverse sequence i, wherein i is a positive integer which is greater than or equal to 1 and less than or equal to N.

Optionally, in this embodiment, fig. 5 is a schematic diagram of migration data flow provided in this embodiment of the present application, as shown in fig. 5, the load of a physical link in the initial network is adjusted to an equilibrium state according to the load information set, so as to obtain a target network, where the adjustment process may be single-round adjustment or multi-round adjustment, a migration process of a certain round of which is shown in the drawing, and 6 physical links (link 1 to link 6) exist in the initial network, where the load information set records loads (load rates) of the 6 physical links, and the load information set includes: link 1 (45%), link 2 (65%), link 3 (40%), link 4 (20%), link 5 (70%) and link 6 (60%), the second load threshold is 50%, the first load threshold is 58%, link 2, link 5 and link 6 are of high load type, firstly, N physical links are ordered in sequence according to the load size to obtain a target load sequence, which is in sequence: link 5 (70%), link 2 (65%), link 6 (60%), link 1 (45%), link 3 (40%), link 4 (20%). Then, the migration data stream in the physical link belonging to the high load type in the positive sequence i is migrated to the physical link belonging to the low load type in the reverse sequence i, for example, the migration data stream in the physical link belonging to the high load type in the positive sequence 1 (i.e. link 5) is migrated to the physical link belonging to the low load type in the reverse sequence 1 (link 4), the migration data stream in the physical link belonging to the high load type in the positive sequence 2 (i.e. link 2) is migrated to the physical link belonging to the low load type in the reverse sequence 2 (link 3), and so on, the migration of the migration data stream of the round is completed. At this time, if a physical link with a high load type still exists after migration, the migration of the migration data stream of the next round may be performed in the same step until the target network is obtained.

It should be noted that, the method for adjusting the link load provided by the present application can overcome the shortages of the existing DCI interconnection long-distance link load balancing technology, and implement DCI long-distance link high-quality transmission by using a low-overhead large-flow (which can be understood as a large-flow) switching strategy of a data flow system calculation path, and fig. 6 is a schematic diagram of a link load adjusting flow provided according to an embodiment of the present application, as shown in fig. 6, including the following steps:

step 601: the network monitors and periodically counts the traffic load condition of the DCI remote link, and collects and analyzes the traffic flow of the link load.

Step 602: according to the service flow record condition in step 601, the network monitoring system analyzes quintuple information of the data flow, defines flows corresponding to service IP addresses or network segments such as AI training, bulk storage, and data backup as large flows (which can be understood as large flows), defaults other flows as small flows (which can be understood as small flows), and counts the number of large flow entries (which can be understood as the target number of the data flows which are detected to meet the target condition).

Step 602a: if the number of large flow entries exceeds 100 times of the number of DCI remote load links, it is considered that load non-uniformity is not likely to occur, and the conventional ECMP Ha Xixuan is continuously used, for example, 6 DCI remote load links exist in the network, and if the number of large flow entries is greater than 600, it is considered that load non-uniformity is not likely to occur.

Step 602b: if the number of large flow entries does not exceed 100 times the number of DCI long-range load links, it is considered that load non-uniformity is likely to occur and that part of the large flow of the high load port (i.e., the port of the high load type physical link) needs to be rerouted to the low load port (i.e., the port of the low load type physical link).

Step 603: and calculating the flow load average value of all the link ports, taking over a large flow from a high-load port (which can be understood as migrating the migration data flow in the high-load type physical link to the low-load type physical link), selecting a low-load link port according to the specified path of the converter table, reducing the high load to be close to the link average load value, and expanding the load of the low-load port to be close to the average load value.

Step 604: and taking all streams of which the high load ports exceed the average load value as the regular stream (can be understood as migration data streams), sequencing from large to small, sequencing all the low load port load values from small to large, enabling the maximum stream to correspond to the lowest load port according to a greedy algorithm, enabling the next-greatest regular stream to correspond to the next-lowest load port, always matching the mapping relation between all the regular streams and the low load ports, and forming a large stream conversion flow table according to the corresponding relation.

Step 605: and carrying out large-flow migration according to the large-flow path switching table, if the migration flow is more than the low-load ports, after completing the sequential migration (which can be understood as migration) to all the low-load ports, reordering the migration according to step 604 until the migration flow is completely migrated.

Step 606: the network monitoring system monitors and analyzes the remote link port load condition again, and if unbalance is found to still exist, the steps 604 and 605 are repeated until the remote link port load degree converges.

Step 607: and (5) finishing the load balancing optimization of the DCI remote link.

Fig. 7 is a schematic diagram of a link load adjustment system according to an embodiment of the present application, as shown in fig. 7, where the system includes: the network monitoring module 100, the large flow entry statistics module 200, the large flow conversion table module 300 and the DCI remote link load optimization module 400, and the network monitoring system 100 is responsible for counting the remote link port load condition of the data center. The large flow entry statistical system 200 is responsible for collecting and analyzing quintuple information of data flows and classifying the large and small flows by combining corresponding service types. The large-flow converter table system 300 is responsible for forming a mapping relation table between the sloshing flow and the low-load port according to the greedy algorithm principle. The DCI remote link load optimization system 400 is responsible for the optimized promotion of the utilization rate of the bandwidth resources of the DCI remote link.

According to the link load adjusting method provided by the embodiment of the application, the target network attribute of the initial network which depends on the equal cost multi-path routing algorithm for carrying out load balancing is detected, wherein the target network attribute is used for indicating the possibility of the load imbalance condition of the physical links in the initial network, under the condition that the target network attribute falls into a target threshold range, namely, under the condition that the possibility of the load imbalance condition of the physical links in the network is larger than the target possibility, the load information of each physical link in N physical links included in the initial network is detected, a load information set is obtained, wherein the load information is used for indicating the load of the corresponding physical link, finally, the load of the physical links in the initial network is adjusted to an equilibrium state according to the load information set, so that the target network is obtained, the target network attribute of the initial network can be actively detected, whether the possibility of the load imbalance condition of the physical links in the network is larger than the target possibility of the physical links in the network is judged, under the condition that the possibility of the load imbalance condition of the physical links in the network is larger than the target possibility of the physical links is judged, the load information set is adjusted to the load of the physical links in the initial network according to the load information set, and the relative load balancing effect of the links in the initial network is poor network is solved, and the relevant technical effects of the load adjusting is further solved. Thereby achieving the technical effect of improving the link load adjusting effect.

It should be noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.

The embodiment of the application also provides a device for adjusting the link load, and the device for adjusting the link load can be used for executing the method for adjusting the link load provided by the embodiment of the application. The following describes a device for adjusting link load provided by an embodiment of the present application.

Fig. 8 is a schematic diagram of a link load adjustment device according to an embodiment of the present application. As shown in fig. 8, the apparatus includes:

a first detection module 802, configured to detect a target network attribute of an initial network, where the target network attribute is used to indicate a possibility that a load imbalance condition occurs in a physical link in the initial network, and the initial network is a network that relies on an equal cost multipath routing algorithm to perform load balancing;

a second detection module 804, configured to detect load information of each of N physical links included in the initial network to obtain a load information set when the target network attribute falls within a target threshold range, where the load information is used to indicate a load of a corresponding physical link, and the probability that the network attribute falls within the target threshold range is used to indicate that a load imbalance condition occurs in the physical link in the network is greater than a target probability;

And the adjusting module 806 is configured to adjust the load of the physical link in the initial network to an equilibrium state according to the load information set, so as to obtain a target network.

The device for adjusting the link load provided by the embodiment of the application detects the target network attribute of the initial network which depends on the equal cost multipath routing algorithm to perform load balancing, wherein the target network attribute is used for indicating the possibility of the load imbalance condition of the physical links in the initial network, under the condition that the target network attribute falls into a target threshold range, namely, under the condition that the possibility of the load imbalance condition of the physical links in the network is greater than the target possibility, the load information of each physical link in N physical links included in the initial network is detected, a load information set is obtained, wherein the load information is used for indicating the load of the corresponding physical link, finally, the load of the physical links in the initial network is adjusted to an equilibrium state according to the load information set, so that the target network is obtained. Thereby achieving the technical effect of improving the link load adjusting effect.

Optionally, in the adjusting device for link load provided in the embodiment of the present application, the first detection module includes:

a detection unit, configured to detect the number of data flows that meet a target condition and are transmitted in the initial network, so as to obtain a target number, where the data flows that meet the target condition are data flows that have a transmitted data volume that is greater than a target data volume;

and the first determining unit is used for determining the target network attribute of the initial network according to the target quantity.

Optionally, in the device for adjusting a link load provided in the embodiment of the present application, the detection unit is further configured to:

Optionally, in the adjusting device for link load provided in the embodiment of the present application, the first determining unit is further configured to:

Optionally, in the adjusting device for link load provided in the embodiment of the present application, the adjusting module includes:

a second determining unit, configured to determine a link type of each of the N physical links according to the load information set, where the link type of the physical link includes: a high load type and a low load type, wherein the load of a physical link belonging to the high load type is larger than a first load threshold value, and the load of a physical link belonging to the low load type is smaller than a second load threshold value;

and the migration unit is used for migrating the data flow in the physical link belonging to the high load type to the physical link belonging to the low load type to obtain the target network.

Optionally, in the link load adjusting device provided in the embodiment of the present application, the second determining unit is further configured to:

Optionally, in the device for adjusting link load provided in the embodiment of the present application, the migration unit is further configured to:

The link load adjusting device comprises a processor and a memory, wherein the modules/units and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

The processor includes a kernel, and the kernel fetches the corresponding program unit from the memory. The kernel can be provided with one or more than one, and the problems of poor link load adjusting effect and the like in the related technology are solved by adjusting kernel parameters.

The memory may include volatile memory, random Access Memory (RAM), and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM), among other forms in computer readable media, the memory including at least one memory chip.

The embodiment of the application provides a computer readable storage medium, on which a program is stored, which when executed by a processor, implements the method for regulating link load.

The embodiment of the application provides a processor which is used for running a program, wherein the program runs to execute the link load adjusting method.

Fig. 9 is a schematic diagram of an electronic device according to an embodiment of the present application, and as shown in fig. 9, an embodiment of the present application provides an electronic device, where the device includes a processor, a memory, and a program stored on the memory and executable on the processor, and the processor implements the following steps when executing the program:

Optionally, the following steps may be implemented when the above processor executes a program:

The device herein may be a server, PC, PAD, cell phone, etc.

The application also provides a computer program product adapted to perform, when executed on a data processing device, a program initialized with the method steps of:

Optionally, the computer program product mentioned above, when executed on a data processing device, is adapted to perform a program initialized with the method steps of:

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, etc., such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.

Claims

1. A method for regulating link load, comprising:

2. The method of claim 1, wherein detecting the target network attribute of the initial network comprises:

3. The method according to claim 2, wherein detecting the number of data flows transmitted in the initial network that meet a target condition, to obtain a target number, comprises:

4. The method of claim 2, wherein said determining the target network attribute of the initial network from the target number comprises:

5. The method according to claim 1, wherein the adjusting the load of the physical link in the initial network to the balanced state according to the load information set, to obtain the target network, includes:

6. The method of claim 5, wherein determining the link type of each of the N physical links based on the set of load information comprises:

7. The method according to claim 6, wherein said migrating the data flow in the physical link belonging to the high load type to the physical link belonging to the low load type to obtain the target network comprises:

8. The method of claim 7, wherein said migrating the migrated data stream from the physical link belonging to the high load type to the corresponding physical link belonging to the low load type comprises:

9. A link load adjustment device, comprising:

10. A computer-readable storage medium, characterized in that the computer-readable storage medium comprises a stored program, wherein the program when run performs the method of any one of claims 1 to 8.

11. An electronic device comprising a memory and a processor, wherein the memory has stored therein a computer program, the processor being arranged to perform the method of any of claims 1 to 8 by means of the computer program.