[go: up one dir, main page]

WO2022262798A1 - 链路处理方法、装置、网络设备和存储介质 - Google Patents

链路处理方法、装置、网络设备和存储介质 Download PDF

Info

Publication number
WO2022262798A1
WO2022262798A1 PCT/CN2022/099093 CN2022099093W WO2022262798A1 WO 2022262798 A1 WO2022262798 A1 WO 2022262798A1 CN 2022099093 W CN2022099093 W CN 2022099093W WO 2022262798 A1 WO2022262798 A1 WO 2022262798A1
Authority
WO
WIPO (PCT)
Prior art keywords
risk
link
duration
state
calculation
Prior art date
Application number
PCT/CN2022/099093
Other languages
English (en)
French (fr)
Inventor
高勇
钱少正
陆钱春
齐进
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Priority to EP22824271.5A priority Critical patent/EP4354826A4/en
Priority to BR112023026511A priority patent/BR112023026511A2/pt
Priority to US18/563,551 priority patent/US20240250893A1/en
Publication of WO2022262798A1 publication Critical patent/WO2022262798A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/22—Alternate routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/02—Topology update or discovery
    • H04L45/036—Updating the topology between route computation elements, e.g. between OpenFlow controllers
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H04L45/123—Evaluation of link metrics
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
    • H04L43/0811—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking connectivity
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/28—Routing or path finding of packets in data switching networks using route fault recovery
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12—Discovery or management of network topologies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/70—Routing based on monitoring results

Definitions

  • the present application relates to the technical field of communications, and in particular to a link processing method, device, network device and storage medium.
  • link flapping occurs in the network, it is usually accompanied by a large number of rerouting and switchback operations for services in the network, resulting in frequent service changes and increasing the risk of data packet loss on the forwarding plane.
  • the present application provides a link processing method, device, network equipment and storage medium.
  • An embodiment of the present application provides a link processing method, which is applied to a link processing device, and the method includes: obtaining the number of state changes and the duration of the first state of the link within a predetermined period of time; based on the number of state changes and the first Calculate the risk information of the link according to the duration of the state; perform path calculation based on the risk information; and perform risk suppression on path switching according to the result of the path calculation.
  • An embodiment of the present application provides a link processing device, including: an information acquisition module configured to acquire the number of state changes of the link within a predetermined time period and the duration of the first state; a risk calculation module configured to and the duration of the first state, calculate the risk information of the link; the path calculation module is configured to perform path calculation based on the risk information; and the risk suppression module is configured to perform risk on path switching according to the result of the path calculation inhibition.
  • An embodiment of the present application provides a network device, including: one or more processors; memory, on which one or more programs are stored, and when one or more programs are executed by one or more processors, one or more Each processor implements any link processing method in the embodiments of the present application.
  • the embodiment of the present application further provides a storage medium, the storage medium stores a computer program, and when the computer program is executed by a processor, any link processing method in the embodiment of the present application is implemented.
  • FIG. 1 shows a schematic diagram of a network management system architecture according to an embodiment of the present application.
  • FIG. 2 shows a schematic flowchart of a link processing method according to an embodiment of the present application.
  • Fig. 3 shows a detailed flowchart of a link processing method according to an exemplary embodiment of the present application.
  • Fig. 4 shows a schematic diagram of a system architecture of another embodiment of the present application.
  • FIG. 5 shows a schematic structural diagram of a link processing device provided by an embodiment of the present application.
  • Fig. 6 is a structural diagram showing an exemplary hardware architecture of a computing device capable of implementing the link processing method and apparatus according to the embodiments of the present invention.
  • FIG. 1 shows a schematic diagram of a network management system architecture according to an embodiment of the present application.
  • the architecture of the network management system shown in FIG. 1 includes: a network controller 10 and a network 20 .
  • the network controller 10 may include: a southbound interface module 11, a message management module 12 and a path computation element (Path Computation Element, PCE) module 13.
  • PCE Path Computation Element
  • the southbound interface module 11 can be configured to receive a link change message reported by a device in the network 20 , and deliver the path calculation result of the network controller 10 to the device in the network 20 .
  • the link change message may be included in an alarm message reported by a network device in the network 20 .
  • the southbound interface module 11 is a module for the network controller 10 to exchange data with switching devices in the network 20 .
  • the southbound interface can include: the interface based on the network configuration netconf protocol, the interface based on the Simple Network Management Protocol Agent (SNMP) and the interface based on the open protocol (openflow protocol), so as to realize the corresponding protocol interface analysis.
  • SNMP Simple Network Management Protocol Agent
  • the message management module 12 can be configured to analyze the messages reported by the devices in the network 20 through the southbound interface, and distribute the received messages.
  • the link change message reported by the device in the network 20 through the southbound interface may include: a link disconnection message or a link connection message.
  • the message management module 12 can distribute the link change message to the PCE module 13 .
  • the PCE module 13 can be configured to use existing network topology information in the network 20 to calculate an end-to-end path that satisfies a predetermined constraint condition and a predetermined calculation policy.
  • the PCE module 13 can be configured to identify risky links, and obtain a feasible path from the source node to the sink node based on path calculation. Appropriate detours.
  • the type of network 20 may be, for example, any one of the following different types of networks: microwave bearer network, IP bearer network, optical transport network-based bearer network, and flexible Ethernet-based bearer network.
  • the devices in the network 20 may be, for example, switching devices or terminal devices; wherein, the switching devices may include, for example, switches and routers, such as router 21 , router 22 , router 23 and router 24 shown in FIG. 1 .
  • the network controller 10 in FIG. 1 may also include more other modules, and the number of network function modules of different types of networks and the number of southbound interfaces are only illustrative. According to actual application needs, it can be flexibly adjusted. Specifically, it can be flexibly configured according to requirements, and there is no limit to the content in this aspect.
  • link (or link interface) oscillation may be caused by various factors, and different factors are superimposed on each other, making the link oscillation sporadic, that is, in a certain period of time, the link will oscillate, After a period of time, the link returns to normal.
  • the usual network management and control is to reduce the invalid rerouting and invalid switchback related to the link; for the link that recovers from the vibration to normal, restore the normal rerouting and switchback.
  • some factors that cause link oscillations are periodic and repeatable, for example, the temperature is high during the day and low at night, so there is a high probability that the link that has been vibrated before will reoccur in the subsequent period.
  • a link that vibrates again has certain risks, and is called a risk link in the present invention.
  • the above-mentioned network management and control lacks the management of risky links, and treats risky links as normal links, does not take into account the impact of risky links on business, and lacks self-adaptive means for oscillating links, often using delay Switching back or accumulating multiple times of rerouting suppresses oscillations and lacks adaptability.
  • FIG. 2 shows a schematic flowchart of a link processing method according to an embodiment of the present application.
  • the link processing method in the embodiment of the present application may include the following steps S210-S240.
  • the risk information of the link is calculated according to the received number of state changes of the link within a predetermined time period and the duration of the first state, and the business that needs path switching is performed according to the risk information.
  • Path calculation, and path switching risk suppression based on path calculation results so as to realize risk link identification and adaptive suppression based on historical statistical analysis, effectively identify link risk information, and carry out corresponding risk suppression from the management and control level, In this way, frequent service changes can be avoided, and the risk of packet loss on the forwarding plane can be reduced.
  • the link processing method before step S210, further includes: determining that the current time reaches the time to regularly update the risk information of the link; or, determining that the state change information of the link is received.
  • the link risk information can be calculated periodically or in response to the received link state change information, that is, the data update of the link risk information can include two types: There are two methods, one is regular update, and the other is update triggered by link information to calculate and identify link risk information.
  • the risk information includes a risk level; the number of predetermined durations is at least one.
  • S210 may specifically include the following steps S11-S13.
  • the shortest duration is the minimum value among at least one predetermined duration, and the shortest duration Less than or equal to the predetermined duration threshold.
  • the risk level of the link can be determined according to the number of state changes and the corresponding limiting conditions for the number of times, as well as the duration of the first state and the corresponding limiting conditions for the duration, so as to effectively identify different influences.
  • Risky links allow users to perform network management on links with different risk levels to suppress risks.
  • the number of state changes of the link refers to the number of changes from the first state to the second state.
  • the first state is a link connection state
  • the second state is a link disconnection state
  • the first state is a link disconnection state
  • the second state is a link connection state.
  • the limiting condition for the number of times in step S12 is: the number of times of state changes within the obtained non-shortest duration is greater than or equal to the predetermined number threshold corresponding to the obtained non-shortest duration.
  • the duration of the first state is the working time
  • the duration limitation condition is: the proportion of the working time of the link within the non-shortest duration, It is less than the working hours proportion threshold corresponding to the acquired non-shortest hours.
  • the duration of the first state is the duration of the outage
  • the duration limit condition is: the length of the outage of the link within the obtained non-shortest duration accounts for ratio, which is greater than or equal to the percentage threshold of the non-shortest duration corresponding to the obtained non-shortest duration.
  • the assessment of the risk level of the link can be mainly based on two variables: the number of link state changes and the duration of the first state.
  • the two variables may be, for example, the number of state changes of the link from connected to disconnected and the proportion of working hours, or may be the number of times of state changes of the link from disconnected to connected and the proportion of stop working time.
  • the first state may be set as a connected state (or working state) of the link; the first state may also be set as a disconnected state (or non-working state) of the link.
  • the total number of predetermined durations is N, and N is an integer greater than or equal to 1.
  • the predetermined duration is a preset shortest duration and the shortest duration is less than or equal to the preset duration threshold; for example, the preset shortest duration may be a minute-level duration.
  • the predetermined duration may include the preset shortest duration and at least one non-shortest duration, and the non-shortest duration is greater than the shortest duration; for example, the non-shortest duration may be an hour-level duration.
  • the predetermined duration is the preset shortest duration, for example, can be set to 10 minutes, and the number of times threshold corresponding to the shortest duration can be set to 5 times.
  • the risk level of the link is determined to be high risk if the number of state changes of the link within 10 minutes (from connected state to disconnected state, or from disconnected state to connected state) is greater than or equal to 5 times.
  • the predetermined durations include a preset shortest duration and at least one non-shortest duration.
  • the predetermined duration may include two durations, three durations or other number of durations.
  • the method for determining the risk level of a link is described below by taking the predetermined duration as two durations.
  • the predetermined duration is two durations, for example, 10 minutes and 1 hour. Within the two durations, the number of "disconnections" and the proportion of working hours are counted respectively.
  • the link from the "connected” state to the "disconnected” state, or the link from the "disconnected” state to the “connected” state is recorded as a state change, and it is determined that the link is The total number of state changes that occurred within a certain period of time; the proportion of working hours within a certain period of time refers to the ratio of the time that the link is in the "connected” state to the period; The ratio of the duration of the "disconnected” state to this duration.
  • the number of times threshold corresponding to 10 minutes can be set to 5 times
  • the number of times threshold corresponding to 1 hour can be set to 7 times
  • the proportion of working hours corresponding to 1 hour can be set to 0.8
  • the proportion of working hours corresponding to 1 hour can be set to Can be set to 0.2.
  • the link is directly determined to be at high risk. If within 1 hour, the number of state changes of the link is greater than or equal to 7 times, or the proportion of working hours is less than 0.8, or the proportion of stop working hours is greater than or equal to 0.2, the link is considered to be at medium risk.
  • the method for determining the risk level of a link is described below by taking the predetermined duration as three durations.
  • the three durations are referred to as the shortest duration, the medium duration, and the longest duration according to the duration, for example, 10 minutes, 1 hour, and 24 hours respectively, corresponding to the thresholds of the times of the above three durations It can be set to 5 times, 7 times, and 10 times, and the proportion of working hours corresponding to the medium duration and the longest duration is set to 0.8 and 0.8.
  • the link can be directly determined to be at high risk.
  • the link Within 1 hour of medium duration, if the number of interruptions exceeds the medium duration threshold 7 times, or the proportion of working hours is less than the medium duration threshold 0.8, the link is considered to be at medium risk; within the longest duration of 24 hours, the number of interruptions exceeds the long duration threshold If the number of times is 10, or the proportion of working hours is less than the long-time threshold of 0.8, the link is considered to be at medium risk.
  • the number and specific duration of the predetermined duration may be set according to actual conditions.
  • the shortest duration can be minute-level (less than 1 hour)
  • the longest duration can be level-level duration (greater than or equal to 24 hours)
  • medium durations other than the shortest and longest durations can be hour-level Duration (greater than or equal to 1 hour and less than 24 hours)
  • the number and value of the above-mentioned durations can be set according to actual conditions, and are not specifically limited in this embodiment of the present application.
  • risk information includes risk levels and risk rates.
  • the above step S220 may specifically include the following sub-steps S21-S24.
  • the time period used when the risk level is calculated is used as the time period for participating in the calculation.
  • the normalization process in step S23 may include, for example, the following sub-steps S31-S33.
  • the first normalization process may be performed through the following expression (1-1), to obtain the risk rate of state changes within the participating calculation time.
  • f downNum represents the risk rate generated by the number of state changes of the link from the connected state to the disconnected state
  • downnum is the number of times the current link changes from the connected state to the disconnected state within the calculation period.
  • the number of state changes, threshold1 is the threshold for the number of state changes corresponding to the calculation duration.
  • k1 0.4
  • k2 0.4. It should be understood that the values of k1 and k2 can be customized according to actual conditions, and are not specifically limited in this embodiment of the present application.
  • the first normalization process may be performed through the following expression (1-2) to obtain the risk rate of the state change within the participating calculation time.
  • f upNum represents the risk rate generated by the number of state changes of the link from the disconnected state to the connected state
  • upNum is the risk rate of the current link from the disconnected state to the connected state within the calculation period.
  • the number of state changes, the meaning of threshold1 and the values of k1 and k2 are the same as the above expression (1-1), and will not be repeated in the embodiment of this application.
  • the second normalization process can be performed through the following expression (2-1) to obtain the risk rate generated by the proportion of working hours within the participating calculation hours.
  • the following expression (2-2) can be used to perform the second normalization process to obtain the proportion of the disconnected duration of the first state participating in the calculation and the corresponding A state duration ratio threshold.
  • the risk rate caused by the state change can be expressed as the following expression (3):
  • the proportion of the duration of the first state is the proportion of working hours
  • the risk rate generated by the proportion of working hours can be expressed as the following expression (4 ):
  • the risk rate caused by the state change can be expressed as the following expression (5):
  • the risk rate generated by the first state change can be expressed as the following expression (6):
  • the above-mentioned first normalization method and the second normalization method can also be based on a zero-mean (z-score) normalization method or based on a hyperbolic tangent function (Hyperbolic Tangent Function, Tanth) the normalization method.
  • z-score zero-mean
  • Tanth hyperbolic Tangent Function
  • the z-score normalization method can be expressed as the following expression (7):
  • the normalization method based on Tanth can be expressed as the following expression (8):
  • f2(x) is the calculated state change corresponding to the participating computing duration Generated risk rate; if the value of x is the proportion of the duration of the first state in the calculation period, then f2(x) is the calculated risk rate corresponding to the duration of the calculation in the first state.
  • the expression of the first normalization processing and the expression of the second normalization processing above may also be other normalization processing methods, which can be set according to actual application scenarios, and are not specifically limited in the embodiments of the present disclosure. .
  • risk information includes risk levels and risk rates.
  • the above step S220 may specifically include the following steps S41-S42.
  • the risk information of each link can be added to the graph resources required for path calculation.
  • the risk suppression in step S240 can use different risk levels to avoid risky links as much as possible to achieve the purpose of suppressing risky links, thereby performing path calculation based on the risk information of the links, and setting high-risk links as disconnected Links, low-risk links are regarded as links that do not affect rerouting calculations, and the corresponding risk rate strategy is added to the path calculation for medium-risk links, so that the risk information and risk rate of the link can be referred to during the path calculation process , so as to carry out risk avoidance and risk suppression according to the corresponding path calculation results.
  • risk mitigation includes rerouting mitigation.
  • the step of calculating the path based on the risk rate of the medium-risk link in the above step S42 may specifically include the following steps S42-01 to S42-03.
  • the risk rate policy is a calculation policy set according to the risk rate of the link with medium risk.
  • multiple path calculation strategies can be used in the path calculation process using the path calculation strategy.
  • the multiple path calculation strategies include the original rerouting calculation strategy and the newly added risk rate strategy.
  • the path with the smallest cumulative risk rate is selected as the optimal switching path for rerouting, so as to avoid risks in the path switching of service rerouting.
  • using the preset path calculation strategy to perform path calculation on the scheduled service includes: performing path calculation on the scheduled service according to the priority of the strategy included in the preset path calculation strategy; wherein, in the preset path calculation strategy Among the included policies, the priority of the risk rate policy is lower than that of the highest priority policy, and the highest priority policy is used to complete the rerouting of scheduled services first.
  • the original rerouting calculation strategy may include, for example, strategies required by rerouting itself such as minimum hops, and the risk strategy is the lowest priority strategy, that is, on the basis of satisfying other strategies, try to select the strategy with the lowest cumulative risk rate.
  • the risk strategy may also adopt a second-lower priority strategy, that is, on the basis of satisfying the priority rerouting strategy, the link with the lowest cumulative risk rate is preferentially selected.
  • risk suppression includes cutback suppression.
  • the step of calculating the path based on the risk rate of the medium-risk link in the above step S42 may specifically include the following steps S42-04 to S42-05.
  • the cumulative risk rate of each link in the current route and the cumulative risk rate of the switchback route are calculated.
  • the cumulative risk rate of the switchback route is smaller, the link is switched back; Otherwise, it will not switch; realize risk avoidance for path switching of service rerouting.
  • the link in the embodiment of the present application is a link composed of underlying network devices connected by the network controller through the southbound interface, and the network controller is the network controller to which the link processing device in the embodiment of the present application belongs ;
  • the state of the link is the link state information obtained in advance from the received link change message of the link, wherein the link change message is a message reported by the underlying network device through the southbound interface.
  • the link processing method in the embodiment of the present application further includes: sending the path calculation result to the underlying network device through the southbound interface of the network controller, so that the underlying network device performs path switching according to the path calculation result.
  • the network controller sends the path result to the device through the southbound interface to complete path switching.
  • the link processing method further includes: in the case of performing path calculation again, setting a penalty weight for the high risk link, so as to reduce deployment according to the penalty weight The traffic volume of the link with the highest risk.
  • the penalty weight of the link can be increased in subsequent path calculations, thereby reducing the number of services on the link, and finally adaptively suppressing frequent switching of services from the management and control level, effectively suppressing oscillating links Invalid path switching caused by the path.
  • the risk information includes a risk rate; the link control method further includes: determining a link with a risk rate greater than a predetermined risk rate threshold in the network as a link that may cause oscillation; link to generate corresponding risk warning information.
  • a link with a risk rate greater than a predetermined risk rate threshold in the network is regarded as a link that may vibrate, and corresponding risk prompt information is generated to give a risk to a link that may vibrate in the future hint.
  • the link processing method of the embodiment of the application can effectively identify risky links with different influences, and adopt a cross-level response plan to effectively suppress the ongoing oscillation Invalid rerouting and invalid switchback caused by links; and give risk warnings to links that may fluctuate in the future, increase the penalty weight of the link in subsequent path calculations, thereby reducing the number of services on the link, and finally from the management and control Layer self-adaption suppresses frequent switching of services.
  • Fig. 3 shows a detailed flowchart of a link processing method according to an exemplary embodiment of the present application.
  • the link processing method may include the following steps S301-S311.
  • link data updates are divided into two types, one is timer-triggered data update for link status changes of all links;
  • the subsequent corresponding processes of the two update methods are the same.
  • the threshold for the number of times of link disconnection and the threshold for the proportion of working hours can be set for each preset duration.
  • the different predetermined durations include the shortest duration, medium duration, and longest duration described in the above embodiments.
  • step S306 judging whether the number of disconnections of the non-shortest duration exceeds the corresponding disconnection times threshold corresponding to the non-shortest duration, or whether the proportion of working hours exceeds the threshold of the proportion of working hours corresponding to the non-shortest duration. If yes, execute step S307 to determine that the link is of medium risk; if not, execute step S308 to determine that the link is of low risk.
  • risk information is added to the graph resources required for path calculation.
  • Set high-risk links as disconnected links continue path calculation for low-risk links, and increase the risk rate strategy for medium-risk links.
  • S310 in the rerouting phase of risk suppression, use a multi-path calculation strategy to perform path calculation.
  • the multipath calculation strategy includes the original rerouting calculation strategy and the newly added risk rate strategy.
  • the risk link identification process is based on the real historical data of the link, and each link in the network is analyzed separately.
  • the algorithm has a clear idea and does not belong to the black box algorithm.
  • the recognition rate can be as high as more than 90%, and the risk recognition rate is high.
  • statistics are performed based on historical link data, so as to realize prediction of future link flapping conditions. Since the time consumption required by the risk identification algorithm is very small, for controlled products, the time consumption of the algorithm is negligible, so that rapid risk identification can be carried out.
  • the risk identification algorithm and the path calculation of rerouting/switching can be combined, and the risk rate strategy can be embedded in the path algorithm, which can effectively adaptively suppress invalid rerouting and switching caused by oscillation.
  • the path algorithm can effectively adaptively suppress invalid rerouting and switching caused by oscillation.
  • For high-frequency oscillation links it can be 100% suppressed; for low-frequency oscillation links, under the premise of ensuring the smoothness of the actual network, it can be effectively suppressed, and the suppression rate of risky links is high.
  • Fig. 4 shows a schematic diagram of a system architecture of another embodiment of the present application.
  • the same reference numerals are used for the same or equivalent structures in FIG. 4 as those in FIG. 1 .
  • the network controller 10 may include: a risk identification module 101 and a risk suppression module 102 .
  • the risk identification module 101 can be configured to calculate the risk rate of the link; the risk suppression module 102 can be configured to perform path calculation based on the risk information of the link, so as to suppress the risk of path switching according to the path calculation result.
  • this module can react in time to the status change information of the link uploaded by each network device in the network 20, and can determine and output the risk information of the link in real time.
  • the risk information in the embodiment of the present application may include: risk level and risk rate.
  • the risk level (that is, the risk level) may include, for example, three levels of low risk, medium risk, and high risk. Each risk level has a corresponding risk rate.
  • the value range of rate is 0-1. Specifically, high risk (the calculated risk rate can be greater than or equal to 0.8, for example), medium risk (the calculated risk rate can be greater than or equal to 0.6 and less than 0.8), low risk (the calculated risk rate is less than 0.6) .
  • the link is suppressed, that is, no rerouting or switchback is calculated until the link is no longer high risk.
  • no rerouting or switchback is calculated until the link is no longer high risk.
  • For medium-risk links you can continue to calculate rerouting or switchback, but you need to consider the impact of the risk rate.
  • For low-risk links recalculation routing or switchback is not affected.
  • the network controller 10 can trigger risk link identification according to the received link oscillation message (such as the link change message reported by the network device through the southbound interface module 11), and the risk identification module 101 starts to calculate the current trigger The risk level and risk rate of the link; the risk suppression module 102 starts rerouting or switching back, during which the path calculation gives priority to the risk rate; after the risk suppression module 102 completes the path calculation, it returns the result to the network controller 10 .
  • the network controller 10 sends the path result to the network device through the southbound interface module 11 to complete path switching.
  • FIG. 5 shows a schematic structural diagram of a link processing device provided by an embodiment of the present application. As shown in Fig. 5, the link processing device may include the following modules.
  • the information obtaining module 510 is configured to obtain the number of state changes of the link and the duration of the first state within a predetermined time period.
  • the risk calculation module 520 is configured to calculate the risk information of the link based on the number of state changes and the duration of the first state.
  • the path calculation module 530 is configured to perform path calculation based on the risk information.
  • the risk suppression module 540 is configured to perform risk suppression on path switching according to the path calculation result.
  • the risk information includes a risk level; the number of predetermined durations is at least one; the risk calculation module 520 includes: a high risk determination unit configured to determine that the number of state changes within the shortest duration is greater than or equal to the minimum duration Corresponding number of thresholds, then determine that the risk level of the link is a high risk; wherein, the shortest duration is the minimum value in at least one predetermined duration, and the shortest duration is less than or equal to the predetermined duration threshold; the medium risk determination unit is configured to be at least one In the predetermined duration, if it is determined that the number of state changes obtained within a non-shortest duration satisfies the corresponding limit condition for the number of times, or if it is determined that the duration of the first state within the acquired non-shortest duration satisfies the corresponding duration limit condition, then determine the chain The risk level of the road in the obtained non-shortest duration is medium risk; the low-risk determination unit is configured to determine that the number of state changes obtained within a non-shortest duration does not meet
  • the condition for limiting the number of times is: the number of state changes obtained within the non-shortest duration is greater than or equal to the predetermined threshold value corresponding to the obtained non-shortest duration; if the first state is a link connection state, the second state If the link is disconnected, the duration of the first state is the working time, and the duration limitation condition is: the working time proportion of the link in the non-shortest time period is less than the working time proportion threshold corresponding to the obtained non-shortest time length.
  • the duration of the first state is the duration of the stop working
  • the duration limitation condition is: the link is not the shortest duration of acquisition
  • the proportion of non-working duration within the period is greater than or equal to the non-minimum duration corresponding to the obtained non-minimum duration corresponding to the threshold of the proportion of non-working duration.
  • the risk information includes a risk level and a risk rate
  • the risk calculation module 520 includes: a duration determination unit configured to use the duration used to calculate the risk level as the duration of the calculation; the duration ratio calculation unit, It is configured to obtain the proportion of the duration of the first state in each participating calculation period according to the ratio of the first state duration in each participating calculation period to the corresponding participating calculation period; the risk rate calculation unit is configured to be based on each The number of state changes within the participating calculation time and the corresponding state change threshold, as well as the proportion of the first state duration in each participating calculation time and the corresponding threshold of the first state duration are normalized to obtain the chain The risk rate of the link in each participating calculation time length; the maximum risk rate determination unit is configured to use the maximum value of the risk rate of the link in each participating calculation time length as the risk rate of the link.
  • the risk rate calculation unit when the risk rate calculation unit performs normalization processing, it is further configured to: perform the first normalization processing on the number of state changes and the corresponding threshold of the number of state changes within the participating calculation time to obtain the link The risk rate caused by state changes within the participating calculation time; based on the proportion of the first state duration and the corresponding threshold of the first state duration within the participating calculation time, the second normalization process is performed to obtain the link The risk rate generated by the proportion of the duration of the first state within the participating calculation time; from the risk rate generated by the state change and the risk rate generated by the proportion of the duration of the first state, the maximum value of the risk rate is obtained as the chain The risk rate corresponding to the risk level of the road.
  • the link processing device further includes: a link processing triggering module, configured to determine that the current moment reaches the timing of updating the link before calculating the risk information of the link based on the number of state changes and the duration of the first state. The moment of the risk information; or, determining that the state change information of the link is received.
  • a link processing triggering module configured to determine that the current moment reaches the timing of updating the link before calculating the risk information of the link based on the number of state changes and the duration of the first state. The moment of the risk information; or, determining that the state change information of the link is received.
  • risk information includes risk levels and risk rates.
  • the path calculation module 530 may include: a graph information updating unit configured to add a risk level of the link to the topology graph information of the network to which the link belongs to obtain updated topology graph information; a triggering unit configured to, according to the updated topology graph information, Trigger path calculation for scheduled services; the path calculation module 530 is also configured to: during the path calculation process: for links with high risk levels, set high-risk links as disconnected links; for low-risk links with risk levels For links whose risk level is medium risk, path calculation is performed based on the risk rate of the link with medium risk.
  • the risk suppression includes rerouting suppression; when the path calculation module 530 is configured to perform path calculation based on the risk rate of a medium-risk link, it may include the following units: a strategy path calculation unit configured to adopt a preset path The calculation strategy calculates the path of the scheduled business to obtain the rerouting candidate path of the scheduled business.
  • the preset path calculation strategy includes: the original rerouting calculation strategy and the newly added risk rate strategy.
  • the risk rate strategy is based on the medium risk
  • the calculation strategy of the risk rate setting of the link configured to calculate the sum of the risk rates of each link in each rerouting candidate path, and obtain the cumulative risk rate of each rerouting candidate path; switching path
  • the determining unit is configured to select the rerouting candidate path corresponding to the minimum value of the cumulative risk rate as the optimal switching path for path calculation.
  • the strategy path calculation unit when the strategy path calculation unit uses the preset path calculation strategy to perform path calculation on the predetermined service, it may be specifically configured to: perform path calculation on the predetermined service according to the priority of the strategy included in the preset path calculation strategy; Among the policies included in the preset path calculation policy, the priority of the risk rate policy is lower than the priority of the highest priority policy, and the highest priority policy is used to complete the rerouting of the predetermined service with priority.
  • risk suppression includes switchback suppression; when the path calculation module 530 is configured to perform path calculation based on the risk rate of a medium-risk link, it may include the following unit: a current path risk rate calculation unit configured to calculate a predetermined The sum of the risk rates of each link in the current business path of the business is taken as the current path risk rate; the original path risk rate calculation unit is configured to calculate the sum of the risk rates of each link in the original business path of the scheduled business as the original path risk rate; the switchback path calculation unit is configured to use the original path risk rate as the path calculation result of the business switchback when the risk rate of the original path is less than the risk rate of the current path.
  • a current path risk rate calculation unit configured to calculate a predetermined The sum of the risk rates of each link in the current business path of the business is taken as the current path risk rate
  • the original path risk rate calculation unit is configured to calculate the sum of the risk rates of each link in the original business path of the scheduled business as the original path risk rate
  • the switchback path calculation unit is configured to use the
  • the link is a link composed of underlying network devices connected by a network controller through a southbound interface, and the network controller is the network controller to which the link processing device belongs; the state of the link is received in advance from The link status information is obtained from the link change message of the link.
  • the link change message is the message reported by the underlying network device through the southbound interface.
  • the link processing device further includes: a calculation result sending module configured to send the path calculation result to the underlying network device through the southbound interface, so that the underlying network device performs path switching according to the path calculation result.
  • the link processing device further includes: a weight setting module, configured to set a penalty weight for the high risk link in the case of performing path calculation again, according to Penalty weights reduce the amount of traffic deployed to high-risk links.
  • the risk information includes a risk rate; the link processing device further includes: an oscillating link determination module configured to determine a link whose risk rate is greater than a predetermined risk rate threshold as a link that may be oscillated; risk warning A module configured to generate corresponding risk prompt information for links determined to be likely to vibrate.
  • the link processing device of the embodiment of the present application aiming at the problem of link oscillation, it can effectively identify the risk links with different influences, and adopt a cross-level solution to effectively suppress the invalidity caused by the link oscillation. Rerouting and invalid switchback; and give risk warnings to links that may fluctuate in the future, increase the penalty weight of the link in subsequent path calculations, thereby reducing the number of services on the link, and finally adaptively suppress services from the management and control level Switch frequently.
  • Fig. 6 is a structural diagram showing an exemplary hardware architecture of a computing device capable of implementing the link processing method and apparatus according to the embodiments of the present invention.
  • the computing device 600 includes an input device 601 , an input interface 602 , a central processing unit 603 , a memory 604 , an output interface 605 , and an output device 606 .
  • the input interface 602, the central processing unit 603, the memory 604, and the output interface 605 are connected to each other through the bus 610, and the input device 601 and the output device 606 are respectively connected to the bus 610 through the input interface 602 and the output interface 605, and then connected to the computing device 600 other component connections.
  • the input device 601 receives input information from the outside, and transmits the input information to the central processing unit 603 through the input interface 602; the central processing unit 603 processes the input information based on computer-executable instructions stored in the memory 604 to generate output information, temporarily or permanently store the output information in the memory 604, and then transmit the output information to the output device 606 through the output interface 605; the output device 606 outputs the output information to the outside of the computing device 600 for the user to use.
  • the computing device shown in FIG. 6 may be implemented as a network device, and the network device may include: a memory configured to store a program; a processor configured to run the program stored in the memory to Execute the link processing method described in the foregoing embodiments.
  • the various embodiments of the present application can be implemented in hardware or special purpose circuits, software, logic or any combination thereof.
  • some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software, which may be executed by a controller, microprocessor or other computing device, although the application is not limited thereto.
  • Computer program instructions may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code written in any combination of one or more programming languages or object code.
  • ISA instruction set architecture
  • Any logic flow block diagrams in the drawings of the present application may represent program steps, or may represent interconnected logic circuits, modules and functions, or may represent a combination of program steps and logic circuits, modules and functions.
  • Computer programs can be stored on memory.
  • the memory may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as, but not limited to, read-only memory (ROM), random-access memory (RAM), optical memory devices and systems (digital versatile disc DVD or CD), etc.
  • Computer readable media may include non-transitory storage media.
  • the data processor can be of any type suitable for the local technical environment, such as but not limited to general purpose computer, special purpose computer, microprocessor, digital signal processor (DSP), application specific integrated circuit (ASIC), programmable logic device (FGPA) and processors based on multi-core processor architectures.
  • DSP digital signal processor
  • ASIC application specific integrated circuit
  • FGPA programmable logic device

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Theoretical Computer Science (AREA)
  • Computing Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Algebra (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Pure & Applied Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Environmental & Geological Engineering (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本申请提出一种链路处理方法、装置、网络设备和存储介质，该方法包括：获取链路在预定时长内的状态变化次数和第一状态持续时长；基于状态变化次数和第一状态持续时长，计算链路的风险信息；基于风险信息进行路径计算；根据路径计算的结果对路径切换进行风险抑制。

Description

链路处理方法、装置、网络设备和存储介质 技术领域
本申请要求2021年6月18日提交的中国专利申请No.202110678522.7的优先权，该中国专利申请的全部内容通过引用合并于此。
技术领域
本申请涉及通信技术领域，具体涉及一种链路处理方法、装置、网络设备和存储介质。
背景技术
在实际网络中，由于接触不良、温度、湿度、运行时间等多个原因会导致部分链路在某个时间段频繁出现“断开-连接-断开-连接”的现象，这种现象称之为链路震荡。
当网络中出现链路震荡时，通常会伴随大量的针对网络中业务进行的重路由与回切操作，造成业务频繁变动，增大了转发面数据丢包风险。
发明内容
本申请提供一种链路处理方法、装置、网络设备和存储介质。
本申请实施例提供一种链路处理方法，应用于链路处理装置，该方法包括：获取链路在预定时长内的状态变化次数和第一状态持续时长；基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息；基于所述风险信息进行路径计算；以及根据所述路径计算的结果对路径切换进行风险抑制。
本申请实施例提供一种链路处理装置，包括：信息获取模块，配置为获取链路在预定时长内的状态变化次数和第一状态持续时长；风险计算模块，配置为基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息；路径计算模块，配置为基于所述风险信息进行路径计算；以及风险抑制模块，配置为根据所述路径计算的结果对路径切换进行风险抑制。
本申请实施例提供一种网络设备，包括：一个或多个处理器；存储器，其上存储有一个或多个程序，当一个或多个程序被一个或多个处理器执行，使得一个或多个处理器实现本申请实施例中的任意一种链路处理方法。
本申请实施例还提供了一种存储介质，存储介质存储有计算机程序，计算机程序被处理器执行时实现本申请实施例中的任意一种链路处理方法。
关于本申请的以上实施例和其他方面以及其实现方式，在附图说明、具体实施方式和权利要求中提供更多说明。
附图说明
图1示出本申请一实施例的网络管理系统架构示意图。
图2示出本申请实施例的链路处理方法的流程示意图。
图3示出根据本申请示例性实施例的链路处理方法的详细流程图。
图4示出本申请另一实施例的系统架构的示意图。
图5示出本申请实施例提供的链路处理装置的结构示意图。
图6是示出能够实现根据本发明实施例的链路处理方法和装置的计算设备的示例性硬件架构的结构图。
具体实施方式
为使本申请的目的、技术方案和优点更加清楚明白，下文中将结合附图对本申请的实施例进行详细说明。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互任意组合。
图1示出本申请一实施例的网络管理系统架构示意图。图1所示的网络管理系统的架构包括：网络控制器10和网络20。其中，网络控制器10可以包括：南向接口模块11、消息管理模块12和路径计算单元(Path Computation Element，PCE)模块13。
在一些实施例中，南向接口模块11可配置为接收网络20中的设备上报的链路变更消息，以及将网络控制器10的路径计算结果下发至网络20中的设备。示例性地，链路变更消息可以包含在网络20中的网络设备上报的告警消息中。
在一些实施例中，南向接口模块11是网络控制器10与网络20中的交换设备进行数据交互的模块。南向接口根据预定的网络协议可以包括：基于网络配置netconf协议的接口，基于简单网络管理协议(Simple Network Management Protocol Agent，SNMP)的接口和基于开放协议(openflow协议)的接口，从而实现对应协议的接口解析。
在一些实施例中，消息管理模块12可配置为解析网络20中的设备通过南向接口上报的消息，并对接收到的消息进行消息分发。在本申请实施例中，网络20中的设备通过南向接口上报的链路变更消息可以包括：链路断开消息或链路连接消息。消息管理模块12可以将该链路变更消息分发至PCE模块13。
在一些实施例中，PCE模块13可配置为利用网络20中已有的网络拓扑信息计算出一条满足预定约束条件和预定计算策略的端到端路径。在本申请实施例中，PCE模块13可配置为进行风险链路识别，并基于路径计算得到一条源节点到宿节点的可行路径，在路径计算过程中，结合链路风险信息对风险链路进行适当的绕路。
在一些实施例中，网络20的类型例如可以是如下不同类型网络中的任一种：微波承载网络、IP承载网络、基于光传送网的承载网络、以及基于灵活以太网的承载网络。网络20中的设备例如可以是交换设备或终端设备；其中，交换设备例如可以包括交换机和路由器，例如图1中示出的路由器21、路由器22、路由器23和路由器24。
应该理解，图1中网络控制器10还可以包含更多的其他模块，不同类型网络的网络功能模块的数目以及南向接口的数目仅仅是示意性的。根据实际应用需要，可以进行灵活调整。具体可以根据需求灵活配置，此方面内容不做限制。
在一些应用场景中，当出现链路断开时，为保证网络的畅通，需要通过网络管控将已部署的业务从断开的链路上迁移到其它正常连接的链路上，并可以称该处理为重路由；当链路从断开变为连接时，为保证原先业务的最优路径，同时保持网络流量分配的最优性和稳定性，需要将之前迁移走的业务还原到原先所在的链路，并可以称该处理为回切。当某个链路频繁出现“断开-连接-断开-连接”时，伴随着大量的重路由与回切操作，进而造成业务频繁变动， 增大了转发面数据丢包风险；同时，在大规模网络下，频繁的重路由和回切操作对控制面的计算能力、处理能力都带来极大的挑战。
在一些应用场景中，由于多种因素都可能导致链路(或链路接口)震荡，不同的因素相互叠加，使得链路的震荡具有偶发性，即在某一个时段，链路会发生震荡，持续一段时间后，链路恢复正常。对于正在发生震荡的链路，通常进行的网络管控是：减少该链路相关的无效重路由与无效回切；对于从震荡恢复到正常的链路，恢复正常的重路由和回切。由于导致链路震荡的部分因素具有一定的周期性、可重复性，例如温度白天高、晚上低，所以之前震荡过的链路较大的概率会在后续时间内再发生震荡，这种未来可能再次发生震荡的链路就存在一定的风险，在本发明中称为风险链路。
上述网络管控缺少对风险链路的管理，且将风险链路与正常链路同等对待，没有考虑到风险链路对于业务的影响，且对于震荡的链路缺乏自适应的手段，往往采用延时回切或累计多次再重路由的方式抑制震荡，缺乏自适应性。
图2示出本申请实施例的链路处理方法的流程示意图。如图2所示，本申请实施例中的链路处理方法可以包括以下步骤S210-S240。
S210，获取链路在预定时长内的状态变化次数和第一状态持续时长。
S220，基于状态变化次数和第一状态持续时长，计算链路的风险信息。
S230，基于风险信息进行路径计算。
S240，根据路径计算的结果对路径切换进行风险抑制。
根据本申请实施例的链路处理方法，根据接收到的链路在预定时长内状态变化次数和第一状态持续时长，计算链路的风险信息，根据该风险信息对需要进行路径切换的业务进行路径计算，并根据路径计算结果进行路径切换的风险抑制，从而实现基于历史统计分析进行的风险链路识别和自适应抑制，有效识别链路的风险信息，并从管控层面进行相应的风险抑制，从而可以避免业务频繁变动，并降低转发面数据的丢包风险。
在一些实施例中，在步骤S210之前，该链路处理方法还包括：确定当前时刻达到定时更新链路的风险信息的时刻；或者，确定接收到链路的状态变化信息。
根据本申请实施例的链路管理方法，可以定时或者响应于接收到的链路的状态变化信息，计算得到链路的风险信息，也就是说，链路的风险信息的数据更新可以包括两种方式，一种是定时更新，一种是链路信息触发引起的更新，进行链路风险信息的计算和识别。
在一些实施例中，风险信息包括风险等级；预定时长的数量为至少一个。在该实施例中，S210具体可以包括如下步骤S11-S13。
S11，若最短时长内的状态变化次数，大于或等于最短时长所对应的次数阈值，则确定链路的风险等级为高风险；其中，最短时长为至少一个预定时长中的最小值，且最短时长小于或等于预定时长阈值。
S12，在至少一个预定时长中，若获取的一个非最短时长内的状态变化次数满足对应的次数限定条件，或者，所获取非最短时长内的第一状态持续时长满足对应的时长限定条件，则确定链路在所获取非最短时长内的风险等级为中风险。
S13，若获取的一个非最短时长内的状态变化次数不满足对应的次数限定条件，并且，所获取非最短时长内的第一状态持续时长不满足对应的时长限定条件，则确定链路在所获取非最短时长内的风险等级为低风险。
根据本申请实施例的链路管理方法，可以根据状态变化次数和对应的次数限定条件，以及第一状态持续时长和对应的时长限定条件，判定链路的风险等级，从而有效甄别出不同影响的风险链路，以用户后续根据不同风险等级的链路进行网络管理，抑制风险发生。
在本申请实施例中，链路的状态变化次数是指从第一状态到第二状态的变化次数。其中，在第一状态为链路连接状态的情况下，第二状态为链路断开状态；在第一状态为链路断开状态，第二状态为链路连接状态。
在一些实施例中，步骤S12中的次数限定条件为：获取的非最短时长内的状态变化次数，大于或等于获取的非最短时长所对应的预定次数阈值。
其中，若第一状态为链路连接状态，第二状态为链路断开状态，则第一状态持续时长为工作时长，时长限定条件为：链路在非最短时长内的工作时长占比，小于获取的非最短时长所对应的工作时长占比阈值。
若第一状态为链路断开状态，第二状态为链路连接状态，则第一状态持续时长为停止工作时长，时长限定条件为：链路在获取的非最短时长内的停止工作时长占比，大于或等于获取的非最短时长所对应的停止工作时长占比阈值。
根据本申请实施例的链路管理方法，对链路的风险等级的评估主要可以根据两个变量：链路状态变化次数和第一状态持续时长。该两个变量例如可以是链路从连接到断开的状态变化次数和工作时长占比，或者，可以是链路从断开到连接的状态变化次数和停止工作时长占比。换言之，第一状态可被设定为链路的连接状态(或工作状态)；第一状态也可以被设定为链路的断开状态(或非工作状态)。
在本申请实施例中，预定时长的总数量为N，N为大于或等于1的整数。N＝1时，预定时长为预设的一个最短时长且该最短时长小于或等于预定时长阈值；示例性地，该预设的最短时长可以是分钟级别的时长。N大于1时，预定时长可以包括该预设的最短时长和至少一个非最短时长，非最短时长大于最短时长；示例性地，非最短时长可以是小时级别的时长。
作为示例，若预定时长的总数量为1，则该预定时长为预设的最短时长，例如可以设置为10分钟，最短时长所对应的次数阈值可以设置为5次。在该示例中，若链路在10分钟内的状态变化次数(从连接状态到断开状态，或者从断开状态到连接状态)大于或等于5次，则确定链路的风险等级为高风险。
作为示例，若预定时长的总数量大于1，则该预定时长包括预设的最短时长和至少一个非最短时长。例如，预定时长可以包括两个时长，三个时长或其他数量的时长。
下面以预定时长为两个时长，说明链路的风险等级的确定方法。
在该示例中，预定时长为两个时长时，例如可以是10分钟和1小时。在该两个时长内，分别统计“断开”次数和工作时长占比。
在本申请实施例中，将链路从“连接”状态到“断开”状态，或将链路从“断开”状态到“连接”状态记为一次状态变化，确定链路在每个预定时长内总共发生的状态变化次数；某个时长内的工作时长占比，指链路处于“连接”状态的时长与该时长的比值；某个时长内的停止工作时长占比，指链路处于“断开”状态的时长与该时长的比值。
示例性地，10分钟对应的次数阈值可以设置为5次，1小时对应的次数阈值可以设置为7次，1小时对应的工作时长占比可以设置为0.8，1小时对应的停止工作时长占比可以设置为0.2。
在该示例中，若10分钟内网络中某链路发生状态变化的次数发生6次，由于超过了该时长对应的次数阈值5次，则直接认定该链路处于高风险。若在1小时内，该链路发生状态变化的次数大于或等于7次，或者，工作时长占比小于0.8，或者，停止工作时长占比大于或等于0.2，则认定链路处于中风险。
下面以预定时长为三个时长，说明链路的风险等级的确定方法。为了描述方便，将该三个时长按照时长从短到长，称为是最短时长、中时长和最长时长，例如分别是10分钟、1小时和24小时，对应于上述三个时长的次数阈值可以设置为5次、7次、10次，对应于中时长、最长时长的工作时长占比设为0.8、0.8。
在最短时长10分钟内，中断次数超过短时长阈值次数5次，例如10分钟内发生6次，则可以直接认定链路处于高风险。在中时长1小时内，中断次数超过中时长阈值次数7次，或者工作时长占比小于中时长阈值0.8，则认定链路处于中风险；在最长时长24小时内，中断次数超过长时长阈值次数10次，或者工作时长占比小于长时长阈值0.8，则认定链路处于中风险。
应理解，在实际应用场景中，预定时长的数量和具体时长可以根据实际情况进行设置。例如，最短时长可以是分钟级别(小于1小时)的时长，最长时长可以是级别的时长(大于或等于24小时)，除最短时长和最长时长之外的中时长，可以是小时级别的时长(大于或等于1小时，且小于24小时)，上述时长的数量和时长值可以根据实际情况进行设置，本申请实施例不做具体限定。
在一些实施例中，风险信息包括风险等级和风险率。上述步骤S220具体可以包括如下子步骤S21-S24。
S21，将计算得到风险等级时所使用的时长，作为参与计算时长。
S22，根据每个参与计算时长内的第一状态持续时长与对应的参与计算时长的比值，得到每个参与计算时长内的第一状态持续时长占比。
S23，基于每个参与计算时长内的状态变化次数和对应的状态变化次数阈值，以及每个参与计算时长内的第一状态持续时长占比和对应的第一状态持续时长占比阈值进行归一化处理，得到链路在每个参与计算时长内的风险率。
S24，将链路在每个参与计算时长内的风险率的最大值，作为链路的风险率。
通过上述步骤S21-S24，可以计算得到网络中每条链路的风险率。
在一些实施例中，步骤S23中的归一化处理，例如可以包括如下子步骤S31-S33。
S31，对参与计算时长内的状态变化次数和对应的状态变化次数阈值，进行第一归一化处理，得到链路在参与计算时长内的由状态变化产生的风险率。
作为示例，在第一状态为连接状态的情况下，可以通过下述表达式(1-1)进行第一归一化处理，得到参与计算时长内的状态变化产生的风险率。
Figure PCTCN2022099093-appb-000001
在上述表达式(1-1)中，f downNum表示链路从连接状态到断开状态的状态变化次数产生的风险率，downnum为当前链路在参与计算时长内从连接状态到断开状态的状态变化次数，threshold1为参与计算时长对应的状态变化次数阈值。示例性地，k1＝0.4，k2＝0.4，应理解，k1和k2的取值可以根据实际情况进行自定义设置，本申请实施例不做具体限定。
作为示例，在第一状态为断开状态的情况下，可以通过下述表达式(1-2)进行第一归一化处理，得到参与计算时长内的状态变化产生的风险率。
Figure PCTCN2022099093-appb-000002
在上述表达式(1-2)中，f upNum表示链路从断开状态到连接状态的状态变化次数产生的风险率，upNum为当前链路在参与计算时长内从断开状态到连接状态的状态变化次数，threshold1的含义以及k1、k2的取值与上述表达式(1‐1)相同，本申请实施例不再赘述。
S32，基于参与计算时长内的第一状态持续时长占比和对应的第一状态持续时长占比阈值，进行第二归一化处理，得到链路在参与计算时长内的由第一状态持续时长占比产生的风险率。
作为示例，在第一状态为连接状态的情况下，可以通过下述表达式(2-1)进行第二归一化处理，得到参与计算时长内的工作时长占比产生的风险率。
Figure PCTCN2022099093-appb-000003
在上述表达式(2-1)中，
Figure PCTCN2022099093-appb-000004
是工作时长占比产生的风险率，uptime是参与计算时长内的工作时长占比，threshold2是参与计算时长对应的工作时长占比阈值。示例性地，k3＝0.8，k4＝0.2，应理解，k3和k4的取值可以根据实际情况进行自定义设置，本申请实施例不做具体限定。
作为示例，在第一状态为断开状态的情况下，可以通过下述表达式(2-2)进行第二归一化处理，得到参与计算的第一状态断开时长占比和对应的第一状态持续时长占比阈值。
Figure PCTCN2022099093-appb-000005
在上述表达式(2-2)中，
Figure PCTCN2022099093-appb-000006
是停止工作时长占比产生的风险率，downTime是参与计算时长内的停止工作时长占比，threshold2的含义以及k3、k4的取值与上述表达式(2‐1)相同，本申请实施例不再赘述。
S33，从由状态变化产生的风险率和由第一状态持续时长占比产生的风险率中，获取风险率的最大值作为链路的风险等级对应的风险率。
作为示例，若将上述中时长作为参与计算时长，则针对该中时长，由状态变化产生的风险率可以表示为下述表达式(3)：
Figure PCTCN2022099093-appb-000007
在上述表达式(3)中，
Figure PCTCN2022099093-appb-000008
为参与计算的中时长对应的中断次数产生的风险率，
Figure PCTCN2022099093-appb-000009
为参与计算的中时长对应的中断次数，
Figure PCTCN2022099093-appb-000010
是参与计算的中时长对应的中断次数阈值。
在该示例中，针对该中时长，若第一状态为连接状态，则第一状态持续时长占比为工作时长占比，由工作时长占比产生的风险率可以表示为下述表达式(4)：
Figure PCTCN2022099093-appb-000011
在上述表达式(4)中，
Figure PCTCN2022099093-appb-000012
为参与计算的中时长对应的工作时长占比产生的风险率，
Figure PCTCN2022099093-appb-000013
为参与计算的中时长对应的工作时长占比，
Figure PCTCN2022099093-appb-000014
为参与计算的中时长对应的工作时长占比阈值。
作为示例，若将上述最长时长作为参与计算时长，则针对该最长时长，由状态变化产生的风险率可以表示为下述表达式(5)：
Figure PCTCN2022099093-appb-000015
在上述表达式(5)中，
Figure PCTCN2022099093-appb-000016
为参与计算的最长时长对应的中断次数产生的风险率，
Figure PCTCN2022099093-appb-000017
为参与计算的最长时长对应的中断次数，
Figure PCTCN2022099093-appb-000018
是参与计算的最长时长对应的中断次数阈值。
在该示例中，针对该最长时长，若第一状态为连接状态，则由第一状态变化产生的风险率可以表示为下述表达式(6)：
Figure PCTCN2022099093-appb-000019
在上述表达式(6)中，
Figure PCTCN2022099093-appb-000020
为参与计算的最长时长对应的工作时长占比产生的风险率，
Figure PCTCN2022099093-appb-000021
为参与计算的最长时长对应的工作时长占比，
Figure PCTCN2022099093-appb-000022
为参与计算的最长时长对应的工作时长占比阈值。
在一些实施例中，上述第一归一化方式和第二归一化方式还可以是基于零-均值(z-score)的归一化方式或基于双曲正切函数(Hyperbolic Tangent Function，Tanth)的归一化方式。
作为示例，z-score归一化方式可以表示为下面的表达式(7)：
Figure PCTCN2022099093-appb-000023
在上述表达式(7)中，若x取值为参与计算时长对应的从第一状态到第二状态的状态变化次数，则μ为预定N个时长内从第一状态到第二状态的状态变化次数的均值，σ为预定N个时长内从第一状态到第二状态的状态变化次数的方差，f1(x)为计算得到的该参与计算时长对应的由状态变化产生的风险率；若x取值为参与计算时长内的第一状态持续时长占比，则μ为预定N个时长内的第一状态持续时长的均值，σ为预定N个时长内的第一状态持续时长的方差，f1(x)为计算得到的该参与计算时长对应的由第一状态持续时长占比产生的风险率。
作为示例，基于Tanth的归一化方式可以表示为下面的表达式(8)：
Figure PCTCN2022099093-appb-000024
在上述表达式(8)中，若x取值为参与计算时长对应的从第一状态到第二状态的状态变化次数，则f2(x)为计算得到的该参与计算时长对应的由状态变化产生的风险率；若x取值为参与计算时长的第一状态持续时长占比，则f2(x)为计算得到的该参与计算时长对应的由第一状态持续时长占比产生的风险率。
应理解，上述第一归一化处理的表达式和第二归一化处理的表达式也可以是其他归一化处理方式，具体可以根据实际应用场景进行设置，本公开实施例不做具体限定。
在一些实施例中，风险信息包括风险等级和风险率。上述步骤S220具体可以包括如下步骤S41-S42。
S41，在链路所属网络的拓扑图信息中增加链路的风险等级，得到更新的拓扑图信息，并根据更新的拓扑图信息，触发对预定业务进行的路径计算。
S42，在路径计算过程中：对于风险等级为高风险的链路，设置高风险的链路为断开链路；对于风险等级为低风险的链路，继续进行路径计算；对于风险等级为中风险的链路，基于中风险的链路的风险率进行路径计算。
通过上述步骤，在对预定业务进行重路由时，可以在路径计算所需要的图资源上，增加每条链路的风险信息。这样，步骤S240中的风险抑制可以为利用不同的风险等级尽力规避风险链路，起到抑制风险链路的目的，从而基于链路的风险信息进行路径计算，将高风险链路设为断开链路，低风险链路视为不影响重路由计算的链路，对中风险链路在路径计算中增加相应的风险率策略，从而在路径计算的过程中参考链路的风险信息和风险率，从而根据相应 的路径计算结果进行风险规避和风险抑制。
在一些实施例中，风险抑制包括重路由抑制。上述步骤S42中的基于中风险的链路的风险率进行路径计算的步骤，具体可以包括如下步骤S42-01至S42-03。
S42-01，采用预设路径计算策略对预定业务进行路径计算，得到预定业务的重路由候选路径；其中，预设路径计算策略包括：原有的重路由计算策略和新增的风险率策略，风险率策略是根据中风险的链路的风险率设置的计算策略。
S42-02，计算每条重路由候选路径中各链路的风险率之和，得到每条重路由候选路径的累计风险率。
S42-03，选择累计风险率中的最小值对应的重路由候选路径，作为路径计算的最优切换路径。
通过上述步骤，采用路径计算策略进行路径计算的过程，可以采用多个路径计算策略，多个路径计算策略包含原有的重路由计算策略和新增的风险率策略，在结合新增风险率策略的路径计算结果中，选择累计风险率最小的路径作为重路由的最优切换路径，从而对业务重路由的路径切换进行风险规避。
在一些实施例中，采用预设路径计算策略对预定业务进行路径计算，包括：按照预设路径计算策略所包含策略的优先级，对预定业务进行路径计算；其中，在预设路径计算策略所包含策略中，风险率策略的优先级低于最高优先级策略的优先级，最高优先级策略用于优先完成预定业务的重路由。
在该实施例中，原有的重路由计算策略例如可以包括最小跳等重路由本身需要的策略，风险策略是最低优先级策略，即在满足其它策略的基础上，尽量选择累计风险率最低的链路；考虑到需要优先完成重路由，风险策略例如也可以采用次低优先级策略，即在满足优先完成重路由策略的基础上，优先选择累计风险率最低的链路。
在一些实施例中，风险抑制包括回切抑制。上述步骤S42中的基于中风险的链路的风险率进行路径计算的步骤，具体可以包括如下步骤S42-04至S42-05。
S42-04，计算预定业务的当前业务路径中各链路的风险率之和，作为当前路径风险率，以及计算预定业务在原业务路径中各链路的风险率之和，作为原路径风险率；
S42-05，在原路径风险率小于当前路径风险率的情况下，将原路径风险率作为业务回切的路径计算结果。
通过上述步骤，在业务的路径需要进行回切时，计算当前路由中各链路的累计风险率与回切路由的累计风险率，当回切路由的累计风险率更小时，链路回切；否则不会切；实现对业务重路由的路径切换进行风险规避。
在一些实施例中，本申请实施例中的链路为网络控制器通过南向接口连接的底层网络设备组成的链路，网络控制器为本申请实施例的链路处理装置所属的网络控制器；链路的状态是预先从接收到的链路的链路变更消息中获取的链路状态信息，其中，链路变更消息是底层网络设备通过南向接口上报的消息。
本申请实施例的链路处理方法还包括：通过网络控制器的该南向接口将路径计算结果发送至底层网络设备，以使底层网络设备根据路径计算结果进行路径切换。
在该实施例中，网络控制器通过南向接口将路径结果下发至设备，完成路径的切换。
在一些实施例中，对于风险等级为高风险的链路，链路处理方法还包括：在再次进行路 径计算的情况下，为高风险的链路设置惩罚权重，以用于根据惩罚权重减少部署至高风险的链路的业务数量。
在该实施例中，对于高风险链路，可以在后续路径计算时增加该链路的惩罚权重，进而减少该链路的业务数量，最终从管控层面自适应抑制业务频繁切换，有效抑制震荡链路导致的无效路径切换。
在一些实施例中，风险信息包括风险率；链路控制方法还包括：将网络中存在的风险率大于预定风险率阈值的链路，确定为可能发生震荡的链路；对确定为可能发生震荡的链路，生成对应的风险提示信息。
在该实施例中，将网络中存在的风险率大于预定风险率阈值的链路，视为可能发生震荡的链路，并生成对应的风险提示信息，以对未来可能震荡的链路给出风险提示。
在本申请实施例中，针对链路震荡存在的问题，本申请实施例的链路处理方法可以有效的甄别出不同影响的风险链路，并采用跨层次的应对方案，有效抑制了正发生震荡链路导致的无效重路由和无效回切；并对未来可能震荡的链路给出风险提示，在后续路径计算时增加该链路的惩罚权重，进而减少该链路的业务数量，最终从管控层面自适应抑制业务频繁切换。
图3示出根据本申请示例性实施例的链路处理方法的详细流程图。如图3所示，链路处理方法可以包括如下步骤S301-S311。
S301，链路数据更新。
在该步骤中，链路的数据更新分为两种，一种是定时器触发的针对所有链路的链路状态变化的数据更新；一种是响应于本控制器的南向接口接收到的任一链路上报的链路状态变化信息而引起的该链路的链路状态变化的数据更新。两种更新方式后续对应的流程一致。
S302，参数初始化。
在该步骤中，可以对预设的每个时长的链路断开次数阈值和工作时长占比阈值进行设置。
S303，统计不同的预定时长内的链路断开次数和工作时长占比。
示例性地，不同的预定时长包括上述实施例中描述的最短时长、中时长和最长时长。
S304，判断最短时长内链路的断开次数是否超过链路断开阈值；若是，执行步骤S305，判定该链路的风险等级为高风险；若否，则执行S306。
S306，判断非最短时长的断开次数是否超对应的非最短时长对应的断开次数阈值，或者，工作时长占比是否超过非最短时长对应的工作时长占比阈值。若是，则执行步骤S307，判定该链路为中风险；若否，则执行步骤S308，判定该链路为低风险。
通过上述步骤S301-S308，完成对链路的风险识别。
S309，更新图资源。
在该步骤中，在算路所需要的图资源上，增加风险信息。将高风险链路设为断开链路，对低风险链路继续进行路径计算，对中风险链路增加风险率策略。
S310，如图3中“多策略算路”所示，在风险抑制的重路由阶段，采用多路径计算策略进行路径计算。多路径计算策略中包括原有的重路由计算策略和新增的风险率策略。
S311，在路径计算结果中，选择累计风险率最低的链路，以完成重路由。
S312，如图3中判断“当前路由风险率小于回切路由”所示，在风险抑制的回切阶段，判断计算得到的当前路由的累计风险率是否小于计算得到的回切路由的累计风险率；若否，表示回切路由的累计风险率更小，则执行S313，判定为执行链路回切；否则执行S314，判定 为不执行链路回切。
通过上述步骤S306-S311，实现对链路在重路由和回切过程中的风险抑制。
根据本申请实施例的链路处理方法，风险链路识别过程是基于链路的真实历史数据，对网络中的每条链路单独分析，算法思路清晰，不属于黑盒算法，对于风险链路的识别率可高达90％以上，风险识别率高。
并且，本申请实施例采用基于链路历史数据进行统计，实现对于未来链路震荡情况的预测。由于风险识别算法所需的耗时非常小，对于管控产品来说，算法耗时可忽略不计，从而可以进行快速的风险识别。
在本申请实施例中，可以结合风险识别算法和重路由/回切的路径计算，在路径算法中嵌入风险率的策略，可以有效自适应抑制震荡产生的无效重路由和回切。对于高频震荡链路，可以百分百抑制；对于低频震荡链路，在保证实际网络通畅的前提下，高效的进行抑制，风险链路抑制率高。
图4示出本申请另一实施例的系统架构的示意图。图4与图1中相同或等同的结构使用相同的标号。如图4所示，与图1中系统架构的不同之处在于，图4中的系统架构中，网络控制器10可以包括：风险识别模块101和风险抑制模块102。
其中，风险识别模块101，可配置为计算链路的风险率；风险抑制模块102，可配置为基于链路的风险信息进行路径计算，以根据路径计算结果对路径切换进行风险抑制。
在风险识别模块101中，该模块可以对网络20中各网络设备上传的链路的状态变化信息进行及时的反应，并可以实时确定并输出该链路的风险信息。
本申请实施例中的风险信息可以包括：风险等级和风险率，风险等级(即风险程度)例如可以包括低风险、中风险、高风险三个等级，每个风险等级具有对应的风险率，风险率的取值范围为0-1。具体对应，高风险(计算得到的风险率例如可以大于或等于0.8)、中风险(计算得到的风险率例如可以大于或等于0.6，且小于0.8)、低风险(计算得到的风险率小于0.6)。
在风险抑制模块102中，对于识别为高风险的链路，无论链路处于何种状态，都将该链路抑制，即不再计算重路由或回切，直至其不为高风险。对于中风险链路，可以继续计算重路由或回切，但是需要考虑风险率的影响。对于低风险链路，不影响算重路由或回切。
在该系统架构中，网络控制器10可以根据接收到的链路震荡消息(例如网络设备通过南向接口模块11上报的链路变更消息)触发风险链路识别，风险识别模块101开始计算当前触发链路的风险等级和风险率；风险抑制模块102开始重路由或回切，期间路径计算优先考虑风险率；风险抑制模块102完成路径计算后，将结果返回网络控制器10。网络控制器10通过南向接口模块11将路径结果下发至网络设备，完成路径的切换。
下面结合附图，详细介绍根据本发明实施例的链路处理装置。图5示出本申请实施例提供的链路处理装置的结构示意图。如图5所示，链路处理装置可以包括如下模块。
信息获取模块510，配置为获取链路在预定时长内的状态变化次数和第一状态持续时长。
风险计算模块520，配置为基于状态变化次数和第一状态持续时长，计算链路的风险信息。
路径计算模块530，配置为基于风险信息进行路径计算。
风险抑制模块540，配置为根据路径计算的结果对路径切换进行风险抑制。
在一些实施例中，风险信息包括风险等级；预定时长的数量为至少一个；风险计算模块 520，包括：高风险判定单元，配置为若判定最短时长内的状态变化次数，大于或等于最短时长所对应的次数阈值，则确定链路的风险等级为高风险；其中，最短时长为至少一个预定时长中的最小值，且最短时长小于或等于预定时长阈值；中风险判定单元，配置为在至少一个预定时长中，若判定获取的一个非最短时长内的状态变化次数满足对应的次数限定条件，或者，若判定所获取非最短时长内的第一状态持续时长满足对应的时长限定条件，则确定链路在所获取非最短时长内的风险等级为中风险；低风险判定单元，配置为若判定获取的一个非最短时长内的状态变化次数不满足对应的次数限定条件，并且，所获取非最短时长内的第一状态持续时长不满足对应的时长限定条件，则确定链路在所获取非最短时长内的风险等级为低风险。
在一些实施例中，次数限定条件为：获取的非最短时长内的状态变化次数，大于或等于获取的非最短时长所对应的预定次数阈值；若第一状态为链路连接状态，第二状态为链路断开状态，则第一状态持续时长为工作时长，时长限定条件为：链路在非最短时长内的工作时长占比，小于获取的非最短时长所对应的工作时长占比阈值。
在一些实施例中，若第一状态为链路断开状态，第二状态为链路连接状态，则第一状态持续时长为停止工作时长，时长限定条件为：链路在获取的非最短时长内的停止工作时长占比，大于或等于获取的非最短时长所对应的停止工作时长占比阈值。
在一些实施例中，风险信息包括风险等级和风险率；风险计算模块520，包括：时长确定单元，配置为将计算得到风险等级时所使用的时长，作为参与计算时长；时长占比计算单元，配置为根据每个参与计算时长内的第一状态持续时长与对应的参与计算时长的比值，得到每个参与计算时长内的第一状态持续时长占比；风险率计算单元，配置为基于每个参与计算时长内的状态变化次数和对应的状态变化次数阈值，以及每个参与计算时长内的第一状态持续时长占比和对应的第一状态持续时长占比阈值进行归一化处理，得到链路在每个参与计算时长内的风险率；最大风险率确定单元，配置为将链路在每个参与计算时长内的风险率的最大值，作为链路的风险率。
在一些实施例中，风险率计算单元在进行归一化处理时，进一步配置为：对参与计算时长内的状态变化次数和对应的状态变化次数阈值，进行第一归一化处理，得到链路在参与计算时长内的由状态变化产生的风险率；基于参与计算时长内的第一状态持续时长占比和对应的第一状态持续时长占比阈值，进行第二归一化处理，得到链路在参与计算时长内的由第一状态持续时长占比产生的风险率；从由状态变化产生的风险率和由第一状态持续时长占比产生的风险率中，获取风险率的最大值作为链路的风险等级对应的风险率。
在一些实施例中，链路处理装置还包括：链路处理触发模块，配置为在基于状态变化次数和第一状态持续时长，计算链路的风险信息之前，确定当前时刻达到定时更新链路的风险信息的时刻；或者，确定接收到链路的状态变化信息。
在一些实施例中，风险信息包括风险等级和风险率。路径计算模块530可以包括：图信息更新单元，配置为在链路所属网络的拓扑图信息中增加链路的风险等级，得到更新的拓扑图信息；触发单元，配置为根据更新的拓扑图信息，触发对预定业务进行的路径计算；路径计算模块530还配置为在路径计算过程中：对于风险等级为高风险的链路，设置高风险的链路为断开链路；对于风险等级为低风险的链路，继续进行路径计算；对于风险等级为中风险的链路，基于中风险的链路的风险率进行路径计算。
在一些实施例中，风险抑制包括重路由抑制；路径计算模块530在配置为基于中风险的链路的风险率进行路径计算时，可以包括以下单元：策略路径计算单元，配置为采用预设路径计算策略对预定业务进行路径计算，得到预定业务的重路由候选路径，其中，预设路径计算策略包括：原有的重路由计算策略和新增的风险率策略，风险率策略是根据中风险的链路的风险率设置的计算策略；风险率之和计算单元，配置为计算每条重路由候选路径中各链路的风险率之和，得到每条重路由候选路径的累计风险率；切换路径确定单元，配置为选择累计风险率中的最小值对应的重路由候选路径，作为路径计算的最优切换路径。
在一些实施例中，策略路径计算单元在采用预设路径计算策略对预定业务进行路径计算时，具体可以配置为：按照预设路径计算策略所包含策略的优先级，对预定业务进行路径计算；其中，在预设路径计算策略所包含策略中，风险率策略的优先级低于最高优先级策略的优先级，最高优先级策略用于优先完成预定业务的重路由。
在一些实施例中，风险抑制包括回切抑制；路径计算模块530在配置为基于中风险的链路的风险率进行路径计算时，可以包括以下单元：当前路径风险率计算单元，配置为计算预定业务的当前业务路径中各链路的风险率之和，作为当前路径风险率；原路径风险率计算单元，配置为计算预定业务在原业务路径中各链路的风险率之和，作为原路径风险率；回切路径计算单元，配置为在原路径风险率小于当前路径风险率的情况下，将原路径风险率作为业务回切的路径计算结果。
在一些实施例中，链路为网络控制器通过南向接口连接的底层网络设备组成的链路，网络控制器为链路处理装置所属的网络控制器；链路的状态是预先从接收到的链路的链路变更消息中获取的链路状态信息，链路变更消息是底层网络设备通过南向接口上报的消息。
在该实施例中，链路处理装置还包括：计算结果下发模块，配置为通过南向接口将路径计算结果发送至底层网络设备，以使底层网络设备根据路径计算结果进行路径切换。
在一些实施例中，对于风险等级为高风险的链路，链路处理装置还包括：权重设置模块，配置为在再次进行路径计算的情况下，为高风险的链路设置惩罚权重，以根据惩罚权重减少部署至高风险的链路的业务数量。
在一些实施例中，风险信息包括风险率；链路处理装置还包括：震荡链路确定模块，配置为将风险率大于预定风险率阈值的链路，确定为可能发生震荡的链路；风险提示模块，配置为对确定为可能发生震荡的链路，生成对应的风险提示信息。
根据本申请实施例的链路处理装置，针对链路震荡存在的问题，可以有效的甄别出不同影响的风险链路，并采用跨层次的应对方案，有效抑制了正发生震荡链路导致的无效重路由和无效回切；并对未来可能震荡的链路给出风险提示，在后续路径计算时增加该链路的惩罚权重，进而减少该链路的业务数量，最终从管控层面自适应抑制业务频繁切换。
需要明确的是，本发明并不局限于上文实施例中所描述并在图中示出的特定配置和处理。为了描述的方便和简洁，这里省略了对已知方法的详细描述，并且上述描述的系统、模块和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
图6是示出能够实现根据本发明实施例的链路处理方法和装置的计算设备的示例性硬件架构的结构图。
如图6所示，计算设备600包括输入设备601、输入接口602、中央处理器603、存储器604、输出接口605、以及输出设备606。其中，输入接口602、中央处理器603、存储器604、 以及输出接口605通过总线610相互连接，输入设备601和输出设备606分别通过输入接口602和输出接口605与总线610连接，进而与计算设备600的其他组件连接。
具体地，输入设备601接收来自外部的输入信息，并通过输入接口602将输入信息传送到中央处理器603；中央处理器603基于存储器604中存储的计算机可执行指令对输入信息进行处理以生成输出信息，将输出信息临时或者永久地存储在存储器604中，然后通过输出接口605将输出信息传送到输出设备606；输出设备606将输出信息输出到计算设备600的外部供用户使用。
在一个实施例中，图6所示的计算设备可以被实现为一种网络设备，该网络设备可以包括：存储器，被配置为存储程序；处理器，被配置为运行存储器中存储的程序，以执行上述实施例描述的链路处理方法。
以上，仅为本申请的示例性实施例而已，并非用于限定本申请的保护范围。一般来说，本申请的多种实施例可以在硬件或专用电路、软件、逻辑或其任何组合中实现。例如，一些方面可以被实现在硬件中，而其它方面可以被实现在可以被控制器、微处理器或其它计算装置执行的固件或软件中，尽管本申请不限于此。
本申请的实施例可以通过移动装置的数据处理器执行计算机程序指令来实现，例如在处理器实体中，或者通过硬件，或者通过软件和硬件的组合。计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码。
本申请附图中的任何逻辑流程的框图可以表示程序步骤，或者可以表示相互连接的逻辑电路、模块和功能，或者可以表示程序步骤与逻辑电路、模块和功能的组合。计算机程序可以存储在存储器上。存储器可以具有任何适合于本地技术环境的类型并且可以使用任何适合的数据存储技术实现，例如但不限于只读存储器(ROM)、随机访问存储器(RAM)、光存储器装置和系统(数码多功能光碟DVD或CD光盘)等。计算机可读介质可以包括非瞬时性存储介质。数据处理器可以是任何适合于本地技术环境的类型，例如但不限于通用计算机、专用计算机、微处理器、数字信号处理器(DSP)、专用集成电路(ASIC)、可编程逻辑器件(FGPA)以及基于多核处理器架构的处理器。
通过示范性和非限制性的示例，上文已提供了对本申请的示范实施例的详细描述。但结合附图和权利要求来考虑，对以上实施例的多种修改和调整对本领域技术人员来说是显而易见的，但不偏离本发明的范围。因此，本发明的恰当范围将根据权利要求确定。

Claims (16)

  1. 一种链路处理方法，应用于链路处理装置，所述方法包括：
    获取链路在预定时长内的状态变化次数和第一状态持续时长；
    基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息；
    基于所述风险信息进行路径计算；以及
    根据所述路径计算的结果对路径切换进行风险抑制。
  2. 根据权利要求1所述的方法，其中，所述风险信息包括风险等级；所述预定时长的数量为至少一个；并且所述基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息，包括：
    若最短时长内的所述状态变化次数，大于或等于所述最短时长所对应的次数阈值，则确定所述链路的风险等级为高风险；其中，所述最短时长为所述至少一个预定时长中的最小值，且所述最短时长小于或等于预定时长阈值；
    在所述至少一个预定时长中，若获取的一个非最短时长内的所述状态变化次数满足对应的次数限定条件，或者，所获取非最短时长内的所述第一状态持续时长满足对应的时长限定条件，则确定所述链路在所获取非最短时长内的风险等级为中风险；
    若获取的一个非最短时长内的所述状态变化次数不满足对应的次数限定条件，并且，所获取非最短时长内的所述第一状态持续时长不满足对应的时长限定条件，则确定所述链路在所获取非最短时长内的风险等级为低风险。
  3. 根据权利要求2所述的方法，其中，
    所述次数限定条件为：获取的所述非最短时长内的所述状态变化次数，大于或等于获取的所述非最短时长所对应的预定次数阈值；
    所述状态变化次数为第一状态到第二状态的变化次数；并且其中，
    所述第一状态为链路连接状态，第二状态为链路断开状态，所述第一状态持续时长为工作时长，所述时长限定条件为：所述链路在所述非最短时长内的工作时长占比，小于获取的所述非最短时长所对应的工作时长占比阈值；或者
    所述第一状态为链路断开状态，第二状态为链路连接状态，所述第一状态持续时长为停止工作时长，所述时长限定条件为：所述链路在获取的所述非最短时长内的停止工作时长占比，大于或等于获取的所述非最短时长所对应的停止工作时长占比阈值。
  4. 根据权利要求1所述的方法，其中，所述风险信息包括风险等级和风险率；所述基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息，包括：
    将计算得到所述风险等级时所使用的时长，作为参与计算时长；
    根据每个参与计算时长内的所述第一状态持续时长与对应的参与计算时长的比值，得到所述每个参与计算时长内的第一状态持续时长占比；
    基于所述每个参与计算时长内的所述状态变化次数和对应的状态变化次数阈值，以及所述每个参与计算时长内的第一状态持续时长占比和对应的第一状态持续时长占比阈值进行归 一化处理，得到所述链路在所述每个参与计算时长内的风险率；以及
    将所述链路在所述每个参与计算时长内的风险率的最大值，作为所述链路的风险率。
  5. 根据权利要求4所述的方法，其中，所述归一化处理，包括：
    对所述参与计算时长内的所述状态变化次数和对应的状态变化次数阈值，进行第一归一化处理，得到所述链路在所述参与计算时长内的由状态变化产生的风险率；
    基于所述参与计算时长内的第一状态持续时长占比和对应的第一状态持续时长占比阈值，进行第二归一化处理，得到所述链路在所述参与计算时长内的由第一状态持续时长占比产生的风险率；以及
    从所述由状态变化产生的风险率和所述由第一状态持续时长占比产生的风险率中，获取风险率的最大值作为所述链路的风险等级对应的风险率。
  6. 根据权利要求1所述的方法，其中，在所述基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息之前，所述方法还包括：
    确定当前时刻达到定时更新所述链路的风险信息的时刻；或者，
    确定接收到所述链路的状态变化信息。
  7. 根据权利要求1所述的方法，其中，所述风险信息包括风险等级和风险率；并且所述基于所述风险信息进行路径计算，包括：
    在所述链路所属网络的拓扑图信息中增加所述链路的风险等级，得到更新的拓扑图信息；
    根据所述更新的拓扑图信息，触发对预定业务进行的所述路径计算；
    其中，在所述路径计算过程中：
    对于所述风险等级为高风险的链路，设置所述高风险的链路为断开链路；
    对于所述风险等级为低风险的链路，继续进行所述路径计算；并且
    对于所述风险等级为中风险的链路，基于所述中风险的链路的风险率进行所述路径计算。
  8. 根据权利要求7所述的方法，其中，所述风险抑制包括重路由抑制；并且所述基于所述中风险的链路的风险率进行所述路径计算，包括：
    采用预设路径计算策略对所述预定业务进行所述路径计算，得到所述预定业务的重路由候选路径；其中，所述预设路径计算策略包括：原有的重路由计算策略和新增的风险率策略，所述风险率策略是根据所述中风险的链路的风险率设置的计算策略；
    计算每条所述重路由候选路径中各链路的风险率之和，得到每条所述重路由候选路径的累计风险率；以及
    选择所述累计风险率中的最小值对应的重路由候选路径，作为所述路径计算的最优切换路径。
  9. 根据权利要求8所述的方法，其中，所述采用预设路径计算策略对所述预定业务进行所述路径计算，包括：
    按照所述预设路径计算策略所包含策略的优先级，对所述预定业务进行所述路径计算；
    其中，在所述预设路径计算策略所包含策略中，所述风险率策略的优先级低于最高优先级策略的优先级，所述最高优先级策略用于优先完成所述预定业务的重路由。
  10. 根据权利要求7所述的方法，其中，所述风险抑制包括回切抑制；并且所述基于所述中风险的链路的风险率进行所述路径计算，包括：
    计算所述预定业务的当前业务路径中各链路的风险率之和，作为当前路径风险率，以及计算所述预定业务在原业务路径中各链路的风险率之和，作为原路径风险率；以及
    在所述原路径风险率小于所述当前路径风险率的情况下，将所述原路径风险率作为所述业务回切的路径计算结果。
  11. 根据权利要求1所述的方法，其中，
    所述链路为网络控制器通过南向接口连接的底层网络设备组成的链路，所述网络控制器为所述链路处理装置所属的网络控制器；
    所述链路的状态是预先从接收到的所述链路的链路变更消息中获取的链路状态信息，所述链路变更消息是所述底层网络设备通过所述南向接口上报的消息；并且
    所述方法还包括：通过所述南向接口将所述路径计算结果发送至所述底层网络设备，以使所述底层网络设备根据所述路径计算结果进行路径切换。
  12. 根据权利要求1所述的方法，其中，所述风险信息包括风险等级，并且对于所述风险等级为高风险的链路，所述方法还包括：
    在再次进行所述路径计算的情况下，为所述高风险的链路设置惩罚权重，以根据所述惩罚权重减少部署至所述高风险的链路的业务数量。
  13. 根据权利要求1所述的方法，其中，所述风险信息包括风险率；并且所述方法还包括：
    将风险率大于预定风险率阈值的所述链路，确定为可能发生震荡的链路；以及
    对确定为可能发生震荡的链路，生成对应的风险提示信息。
  14. 一种链路处理装置，其中，所述装置包括：
    信息获取模块，配置为获取链路在预定时长内的状态变化次数和第一状态持续时长；
    风险计算模块，配置为基于所述状态变化次数和第一状态持续时长，计算所述链路的风险信息；
    路径计算模块，配置为基于所述风险信息进行路径计算；以及
    风险抑制模块，配置为根据所述路径计算的结果对路径切换进行风险抑制。
  15. 一种网络设备，包括：
    一个或多个处理器；
    存储器，其上存储有一个或多个程序，当所述一个或多个程序被所述一个或多个处理器 执行，使得所述一个或多个处理器实现权利要求1-13中任一项所述的方法。
  16. 一种存储介质，所述存储介质存储有计算机程序，所述计算机程序被处理器执行时实现权利要求1-13任一项所述的方法。
PCT/CN2022/099093 2021-06-18 2022-06-16 链路处理方法、装置、网络设备和存储介质 WO2022262798A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP22824271.5A EP4354826A4 (en) 2021-06-18 2022-06-16 METHOD AND DEVICE FOR CONNECTION PROCESSING, NETWORK DEVICE AND STORAGE MEDIUM
BR112023026511A BR112023026511A2 (pt) 2021-06-18 2022-06-16 Método e aparelho de processamento de link, dispositivo de rede e meio de armazenamento
US18/563,551 US20240250893A1 (en) 2021-06-18 2022-06-16 Link processing method and apparatus, network device, and storage medium

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110678522.7 2021-06-18
CN202110678522.7A CN115499364A (zh) 2021-06-18 2021-06-18 链路处理方法、装置、网络设备和存储介质

Publications (1)

Publication Number Publication Date
WO2022262798A1 true WO2022262798A1 (zh) 2022-12-22

Family

ID=84464158

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/099093 WO2022262798A1 (zh) 2021-06-18 2022-06-16 链路处理方法、装置、网络设备和存储介质

Country Status (5)

Country Link
US (1) US20240250893A1 (zh)
EP (1) EP4354826A4 (zh)
CN (1) CN115499364A (zh)
BR (1) BR112023026511A2 (zh)
WO (1) WO2022262798A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116996964A (zh) * 2023-07-28 2023-11-03 中国电信股份有限公司技术创新中心 终端数据传输方法、装置、电子设备及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1764154A (zh) * 2004-10-20 2006-04-26 日立通讯技术株式会社 标记交换通路的路径控制
US20160182344A1 (en) * 2014-12-17 2016-06-23 Ciena Corporation Systems and methods to detect, diagnose, and mitigate issues in multi-layer networks
CN107769951A (zh) * 2016-08-22 2018-03-06 南京中兴软件有限责任公司 链路处理方法及装置
CN112422442A (zh) * 2020-10-21 2021-02-26 鹏城实验室 处理双向转发检测协议的会话中断消息的方法及存储介质
CN112532518A (zh) * 2020-11-26 2021-03-19 新华三技术有限公司 一种段路由策略的路径选择方法及装置

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008128572A1 (en) * 2007-04-20 2008-10-30 Telefonaktiebolaget Lm Ericsson (Publ) A method and apparatus for quality of service (qos) planning for an ethernet based network
CN101572674A (zh) * 2009-06-12 2009-11-04 杭州华三通信技术有限公司 一种路由计算方法及装置
US10108803B2 (en) * 2016-03-31 2018-10-23 International Business Machines Corporation Automatic generation of data-centric attack graphs

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1764154A (zh) * 2004-10-20 2006-04-26 日立通讯技术株式会社 标记交换通路的路径控制
US20160182344A1 (en) * 2014-12-17 2016-06-23 Ciena Corporation Systems and methods to detect, diagnose, and mitigate issues in multi-layer networks
CN107769951A (zh) * 2016-08-22 2018-03-06 南京中兴软件有限责任公司 链路处理方法及装置
CN112422442A (zh) * 2020-10-21 2021-02-26 鹏城实验室 处理双向转发检测协议的会话中断消息的方法及存储介质
CN112532518A (zh) * 2020-11-26 2021-03-19 新华三技术有限公司 一种段路由策略的路径选择方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4354826A4 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116996964A (zh) * 2023-07-28 2023-11-03 中国电信股份有限公司技术创新中心 终端数据传输方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
EP4354826A1 (en) 2024-04-17
US20240250893A1 (en) 2024-07-25
CN115499364A (zh) 2022-12-20
BR112023026511A2 (pt) 2024-03-05
EP4354826A4 (en) 2024-10-30

Similar Documents

Publication Publication Date Title
US11240153B1 (en) Scoring policies for predictive routing suggestions
US11044191B2 (en) Coupling reactive routing with predictive routing in a network
US10977574B2 (en) Prediction of network device control plane instabilities
US9838317B1 (en) Policy-based selective traffic reroute based on predicted traffic loss
US11398958B2 (en) Reverting routing decisions made based on incorrect network predictions
US10313153B2 (en) Adaptive MAC grouping and timeout in EVPN environments using machine learning
US9225616B2 (en) Feedback-based tuning of control plane traffic by proactive user traffic observation
US11063861B2 (en) Ensuring backup path performance for predictive routing in SD-WANs
WO2012166886A1 (en) Management of misbehaving nodes in a computer network
US9563440B2 (en) Fast learning to train learning machines using smart-triggered reboot
US11616712B2 (en) Machine learning approach for dynamic adjustment of BFD timers in SD-WAN networks
US20220294738A1 (en) Globally avoiding simultaneous reroutes in a network
US20230128567A1 (en) Elastic allocation of resources for optimizing efficiency of predictive routing systems
US20230136635A1 (en) Selective formation and maintenance of tunnels within a mesh topology
US20210351989A1 (en) Self-managed networks and services with artificial intelligence and machine learning
WO2022262798A1 (zh) 链路处理方法、装置、网络设备和存储介质
CN104639557A (zh) 一种建立pcep会话的方法、系统及设备
US20190372852A1 (en) Event-aware dynamic network control
WO2023065756A1 (zh) 确定性路由的计算方法、装置、存储介质和电子设备
US11711291B2 (en) Progressive automation with predictive application network analytics
US20190273633A1 (en) Link state packet transmission method and routing node
US20220377005A1 (en) Safety net engine for machine learning-based network automation
US20240089204A1 (en) Communication Method, Device, and System
WO2016160007A1 (en) Method and apparatus for flow control
US11483234B2 (en) Predictive routing-based policy adjustments in software-defined networks

Legal Events

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

Ref document number: 22824271

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 18563551

Country of ref document: US

WWE Wipo information: entry into national phase

Ref document number: 2022824271

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2022824271

Country of ref document: EP

Effective date: 20231218

REG Reference to national code

Ref country code: BR

Ref legal event code: B01A

Ref document number: 112023026511

Country of ref document: BR

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 112023026511

Country of ref document: BR

Kind code of ref document: A2

Effective date: 20231215