WO2022022558A1 - 用于聚合链路的流量分配方法、装置、光线路终端及存储介质 - Google Patents

用于聚合链路的流量分配方法、装置、光线路终端及存储介质 Download PDF

Info

Publication number
WO2022022558A1
WO2022022558A1 PCT/CN2021/108872 CN2021108872W WO2022022558A1 WO 2022022558 A1 WO2022022558 A1 WO 2022022558A1 CN 2021108872 W CN2021108872 W CN 2021108872W WO 2022022558 A1 WO2022022558 A1 WO 2022022558A1
Authority
WO
WIPO (PCT)
Prior art keywords
port
adjusted
traffic
rate
predicted
Prior art date
Application number
PCT/CN2021/108872
Other languages
English (en)
French (fr)
Inventor
廖志成
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
Publication of WO2022022558A1 publication Critical patent/WO2022022558A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04Q—SELECTING
    • H04Q11/00—Selecting arrangements for multiplex systems
    • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
    • H04Q11/0062—Network aspects
    • H04Q11/0067—Provisions for optical access or distribution networks, e.g. Gigabit Ethernet Passive Optical Network (GE-PON), ATM-based Passive Optical Network (A-PON), PON-Ring
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/005—Control of transmission; Equalising
    • 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
    • H04L45/24—Multipath
    • 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
    • H04L45/24—Multipath
    • H04L45/245—Link aggregation, e.g. trunking
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/80—Actions related to the user profile or the type of traffic
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04Q—SELECTING
    • H04Q11/00—Selecting arrangements for multiplex systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04Q—SELECTING
    • H04Q11/00—Selecting arrangements for multiplex systems
    • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
    • H04Q11/0062—Network aspects
    • H04Q2011/0086—Network resource allocation, dimensioning or optimisation

Definitions

  • the embodiments of the present disclosure relate to the technical field of optical line terminals, and in particular, to a traffic distribution method, apparatus, optical line terminal, and storage medium for aggregated links.
  • Link aggregation technology provides a solution.
  • Link aggregation technology aggregates multiple physical links together to form a complete logical link.
  • the aggregated link bandwidth will be the sum of the bandwidth of each physical port, so the bandwidth of the data forwarding channel can be expanded, solving the bandwidth bottleneck; in addition, aggregating multiple physical links together to form a logical channel means that even if a certain Due to various reasons, one of the links is damaged, and the data will also be adjusted to use other links for transmission and then completely sent to thousands of households, which undoubtedly improves the reliability of the network and solves the chain caused by link congestion. type risk.
  • IEEE802.3ad defines the frame structure and rules of link aggregation, it does not define which method to implement. After link aggregation is performed on multiple physical ports, the following scenarios may occur: A port with high transmission load may encounter high burst traffic, causing the port traffic to exceed the limit. Some physical ports are heavily loaded to forward data, while other physical ports are in a state of low traffic load for a long time and bandwidth is not fully utilized, resulting in a waste of resources. Therefore, how to dynamically allocate the traffic of each physical port of the aggregated link and reduce the load difference of each physical port as much as possible has become an urgent problem to be solved.
  • the purpose of the embodiments of the present disclosure is to provide a traffic distribution method, device, optical line terminal OLT and storage medium for an aggregated link, which can dynamically allocate the traffic of each physical port of the aggregated link and reduce the load difference of each physical port , reduce waste of resources.
  • the present disclosure proposes a traffic allocation method for an aggregated link, the method comprising: predicting a predicted average rate of a port in a current monitoring period and rates of all ports based at least on the port traffic in one monitoring period on the port
  • the average value of the rate is the average value of the predicted average rate
  • the port traffic in the current monitoring period of the port to be adjusted is adjusted to the adjusted port according to the load balancing element, where the load balancing element includes preset reports of different rates in the port traffic.
  • the length range of the text, the port to be adjusted is the port whose predicted traffic rate is greater than the average rate, and the port to be adjusted is the port whose predicted traffic rate is less than the average rate.
  • the present disclosure provides a traffic distribution device for an aggregated link
  • the device includes: a rate prediction module for predicting the predicted average of the port in the current monitoring period based on at least the port traffic in the previous monitoring period of the port The rate and the average rate of all ports, the average rate is the average of the predicted average rates; the adjustment module is used to adjust the port traffic in the current monitoring period of the port to be adjusted to the adjusted port according to the load balancing element, among which the load balancing The element includes preset packet length ranges of different rates in port traffic, the port to be adjusted is the port whose predicted traffic rate is greater than the average rate, and the adjusted port is the port whose predicted traffic rate is less than the average rate.
  • the present disclosure proposes an optical line terminal including the traffic distribution apparatus for an aggregated link as described above.
  • the present disclosure provides a computer-readable storage medium, the storage medium stores one or more programs, and when the one or more programs can be executed by one or more processors, realizes the above-mentioned use The steps of the traffic distribution method for aggregated links.
  • FIG. 1 is a schematic diagram of steps of a traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 2 is a schematic diagram of steps provided by another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 3 is a schematic diagram of steps of another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 4 is a schematic diagram of steps of yet another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 5 is a schematic diagram of steps of yet another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 6 is a schematic diagram of steps of another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 7 is a schematic diagram of steps of another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 8 is a schematic diagram of steps of another traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 9 is a schematic structural diagram of a traffic distribution apparatus for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 10 is a schematic structural diagram of an optical line terminal provided by an embodiment of the present disclosure.
  • FIG. 11 is a schematic diagram of traffic prediction in a traffic allocation method for an aggregated link provided by an embodiment of the present disclosure.
  • FIG. 12 is a schematic diagram of traffic adjustment in a traffic distribution method for an aggregated link provided by an embodiment of the present disclosure.
  • a traffic distribution method for an aggregated link provided by an embodiment of the present disclosure is suitable for an aggregated link, and can dynamically adjust the port traffic among the physical ports of the aggregated link, realizes the dynamic distribution of the aggregated link traffic, reduces the The load difference of each physical port is small.
  • the traffic distribution method for an aggregated link and each step thereof provided in this specification will be described in detail below.
  • the traffic distribution method for an aggregated link provided by the embodiment of the present disclosure, under the condition that the total port traffic of each physical port of the aggregated link remains unchanged in the current monitoring period, the traffic between the various ports remains unchanged. Dynamically adjust the port traffic of some physical ports. Both the port to be adjusted and the port to be adjusted are ports on the aggregated link.
  • FIG. 1 it is a schematic diagram of steps of a traffic distribution method for an aggregated link provided by an embodiment of the present disclosure. It can be understood that the traffic distribution method for an aggregated link provided by the embodiment of the present disclosure is applicable to an optical line terminal OLT having an aggregated link.
  • the traffic distribution method for aggregated links includes the following steps S10 to S20.
  • step S10 the predicted average rate of the port in the current monitoring period and the average rate of all ports are predicted based on the port traffic in the previous monitoring period of the port at least, and the average rate is the average of the predicted average rates.
  • the predicted average rate of the port in the current monitoring period can be predicted according to the historical port traffic, so as to meet the traffic demand of a complex traffic field.
  • the port traffic of each port can be directly obtained without introducing additional calculation overhead, which saves time for the rapid allocation of port traffic.
  • the second is to divide the load sharing element according to the packet length, and use the load sharing element as the adjustment granularity to distribute the port traffic of each port.
  • the current network monitoring module can directly obtain the statistics of the packet length, which also saves the rapid distribution of port traffic. time.
  • the current network monitoring module can periodically obtain the traffic statistics of each port in the aggregated link and the statistics of the packet length in the traffic of each port.
  • the predicted average rate of a port in the current monitoring period is the pass rate of the port traffic in the current monitoring period, that is, the sum of the port traffic in the current monitoring period divided by the current monitoring period.
  • the average rate of port traffic of each port in the previous monitoring period After reading the traffic statistics of each port from the current network monitoring module, calculate the average rate of port traffic of each port in the previous monitoring period, and regard the average rate of port traffic of one port in one monitoring period as the average rate of port traffic of one port.
  • the predicted average rate of ports in the current monitoring period can be predicted based on the historical average rate of port traffic in the previous monitoring period, and the average rate of all ports can be obtained based on the predicted average rate of each port, which is Next, prepare for the distinction between the port to be adjusted and the port to be adjusted.
  • step S20 the port traffic in the current monitoring period of the port to be adjusted is adjusted to the adjusted port according to the load balancing element, wherein the load balancing element includes preset packet length intervals of different rates in the port traffic, and the port to be adjusted is the predicted port in the port. For ports whose traffic rate is greater than the average rate, the adjusted port is the port whose predicted traffic rate is less than the average rate.
  • the current network monitoring module periodically obtains the traffic statistics information of each port and the statistics information of the packet length in the traffic of each port.
  • the rate of each preset packet length interval in the current monitoring period can be calculated based on the above-mentioned statistical information of the packet length in the traffic of each port.
  • the preset packet length interval can be set according to the traffic distribution requirements, for example, it needs to be adjusted every time The larger the granularity of the load sharing element, the larger the packet length range set in the preset packet length interval, and vice versa.
  • the rate in the preset packet length interval is the pass rate of packets of different lengths in the preset packet length interval in the current monitoring period, that is, the sum of the traffic of packets of different lengths in the current monitoring period divided by the current monitoring period.
  • the predicted average rate of the port to be adjusted in the current monitoring period is greater than the average rate. Therefore, the port traffic of the port to be adjusted in the current monitoring period needs to be adjusted, and the port traffic of the port to be adjusted in the current monitoring period needs to be adjusted from the port to be adjusted. Comes over, that is, it needs to share the traffic with the port to be adjusted.
  • the preset packet length interval For the definition of the preset packet length interval, please refer to Table 1.
  • Table 1 different preset packet length intervals are encoded with IDs to facilitate subsequent calls.
  • the preset packet length intervals of different rates in Table 1 become traffic allocation. Minimum granularity that can be adjusted.
  • Preset message length range Preset message length interval ID rate (0, 64] 1 rate_1 (64, 128] 2 rate_2 (128, 256] 3 rate_3 (256, 512] 4 rate_4 (512, 1024] 5 rate_5 (1024, 1500] 6 rate_6
  • the predicted average rate of the port in the current monitoring period is predicted based on at least the port traffic in the previous monitoring period of the port, and the The following steps S100 to S110 are included.
  • step S100 N sampling samples of the port are obtained based on the port traffic, and the N sampling samples include the average rate of port traffic in N monitoring periods including the previous monitoring period.
  • the average rate of each port in the last monitoring period can be calculated according to the traffic statistics of each port, that is, the sum of the traffic in the last monitoring period divided by the one monitoring period.
  • the average rate in the last monitoring period is regarded as a sampling sample, and several monitoring periods including the last monitoring period are regarded as a sampling time window.
  • N data samples are obtained, a least squares mathematical model is established for the N data samples, and the least squares mathematical model is used to predict the predicted average rate of each port in the current monitoring period.
  • step S110 the predicted average rate of the port in the current monitoring period is predicted based on the N sampling samples.
  • the predicted average rate of port traffic in the current monitoring period can be predicted based on the least squares mathematical model or other mathematical algorithms, so as to meet the traffic requirements of the complex traffic field.
  • the monitoring period is configured as T seconds, and the default is 30s; the effective sampling time threshold is T sample seconds, and the default is 300s; the sampling time window is N ⁇ T seconds, and N is 10 by default. ;
  • the sequence number dependent variable of the monitoring period is n, which is initialized to 1.
  • the load balancing unit is defined as the smallest unit that can be adjusted when the port performs load distribution.
  • the packet length interval is shown in Table 1.
  • the average rate of the port in this monitoring period is stored to memory as the value of the sampled samples.
  • rate 1(n-1) , ..., rate 1 (nN) as samples, predict the predicted average rate of the first port in n+1 monitoring periods, such as using historical N
  • the predicted average rate of the port in the current monitoring period is obtained by sampling samples, which is recorded as rate 1(n+1) .
  • the traffic distribution method further includes the following steps S120 to: S140.
  • step S120 the predicted average rates of all ports are obtained respectively.
  • the predicted average rate of all ports in the current monitoring period can be predicted by using the least squares mathematical model based on the collected historical N sampling samples, as shown in FIG. 11 .
  • step S130 the predicted average rates of all ports are averaged to obtain an average rate.
  • the average rate is obtained by averaging the predicted average rates of all the ports in the current monitoring period, and the purpose is to distinguish the port to be adjusted and the port to be adjusted based on the predicted average rate and the average rate of all ports.
  • step S140 the port to be adjusted and the port to be adjusted are determined based on the predicted average rate and the rate average.
  • the port to be adjusted is the port whose predicted traffic rate is greater than the average rate among the ports, and the port to be adjusted is the port whose predicted traffic rate is less than the average rate among the ports.
  • the purpose of determining the port to be adjusted and the port to be adjusted is to realize the redistribution of aggregated link traffic.
  • the predicted average rate of the port to be adjusted in the current monitoring period is greater than the average rate. Therefore, the port traffic of the port to be adjusted in the current monitoring period needs to be adjusted, and the port traffic of the port to be adjusted in the current monitoring period needs to be adjusted from the port to be adjusted. Comes over, that is, it needs to share the traffic with the port to be adjusted.
  • the traffic distribution method provided by the embodiment of the present disclosure further includes: The steps S30 to S40 are as follows.
  • step S30 all ports are sorted in descending order according to the predicted average rate to form a port set.
  • All ports are sorted according to the predicted average rate from large to small to form a port set, which is denoted as C1.
  • step S40 after sequentially selecting the ports to be adjusted from the port set, calculate the rate of the preset packet length interval to which the packets of the port to be adjusted in the current monitoring period belong to based on the length of the packets in the port traffic, until all the to-be-adjusted ports are obtained.
  • the rate of the preset packet length range that the packets of the port belong to.
  • the traffic in the preset packet length range to which the packets belong can be divided by the current monitoring period time to obtain the preset packet length range to which the packets belong in the current monitoring period. s speed. Therefore, to obtain the traffic of each load sharing element of a port in the current monitoring period, the method is similar to the statistics of the port traffic in the monitoring period, and then calculate the rate of each load sharing element in turn, and put all the port traffic of each port in all the ports. Load sub-units are sorted in ascending order of rate.
  • rate_elem p1 ⁇ rate p1 , len1, rate p1, len2 , rate p1, len3 , rate p1, len4 ⁇ .
  • the port traffic in the current monitoring period of the port to be adjusted is adjusted to the adjusted port according to the load balancing element, and the following steps S200 to S210 are further included.
  • Step S200 Adjust the port traffic from the port to be adjusted to the port to be adjusted with the preset packet length interval as the granularity in order according to the rate of the preset packet length interval from small to large.
  • the port to be adjusted can only be adjusted according to the single preset packet length interval of the single rate as the adjustment granularity. port traffic is allocated.
  • the port number of the first port is denoted as p balance_1 , because the port set C1 is based on the predicted average rate from large to small.
  • the ports are sorted.
  • the first port has the largest predicted average rate in the current monitoring period. Therefore, the first port has the largest load in the current monitoring period and needs to be shared first.
  • All load subunits from the first port p balance_1 Select the load balancing element for traffic distribution, and note down the number ID of the load balancing element, the port number p balance_1 of the port to be adjusted, and the port number of the corresponding port to be adjusted, until the predicted average rate of the first port is distributed around the rate average. Referring to FIG.
  • the port traffic of the first port is adjusted from the port to be adjusted to the port to be adjusted with the preset packet length interval as the granularity.
  • the direction column on the left represents the preset packet length intervals of different rates with different colored bars. The smaller the area, the smaller the rate.
  • the preset packet lengths are listed in order.
  • the rate of the interval is adjusted from the first port to the adjusted port in order from small to large, and the adjusted port preferentially selects the port with the largest difference between the predicted average rate and the average rate in the current monitoring period.
  • the operation is repeated until the predicted average rate of the first port is assigned near the rate average. Repeat the operation until the predicted average rates of all the adjusted ports are all assigned near the average rate.
  • step S210 the corresponding relationship between the port to be adjusted, the port to be adjusted and the serial number ID of the preset packet length interval is formed.
  • the corresponding relationship between the port to be adjusted, the port to be adjusted and the serial number ID of the preset packet length interval can be used as a uniform load distribution table of the aggregated link, as shown in Table 2.
  • Preset message length interval ID port being adjusted SRC_PORT1 1 ID_N ... ... ... PORT_N 4 DEST_PORT_N
  • the port traffic is adjusted from the port to be adjusted to the port to be adjusted with the preset packet length interval as the granularity in order from small to large according to the rate of the preset packet length interval. Adjusting the port further includes the following step S201.
  • step S201 a greedy algorithm is used to adjust the port traffic from the to-be-adjusted port to the adjusted port with a preset packet length interval as the granularity, until the predicted average rate of the to-be-adjusted port is close to the rate average.
  • the port traffic of the port to be adjusted is adjusted to the port to be adjusted with the preset packet length interval as the granularity, and the serial number ID of the load sharing element, the port to be adjusted and the corresponding adjusted port are recorded. No.
  • the greedy algorithm can ensure that the predicted average rate of the port to be adjusted is as close to the average rate as possible.
  • the step S210 the corresponding relationship between the port of the port to be adjusted, the port of the port to be adjusted and the number ID of the preset packet length interval is formed.
  • the traffic distribution method provided by the embodiment further includes the following steps S220 to S240.
  • step S220 the source port of the packet, the preset packet length interval to which the packet belongs, and the serial number ID of the preset packet length interval are determined based on the parameters carried in the packet in the port traffic.
  • the preset message length interval can be set to six intervals (0, 64], (64, 128], (128, 256], (256, 512], (512, 1024], (1024, 1518), to obtain the The serial number ID of the preset message length interval to which the message belongs. For example, the id of the (0, 64] interval is 1, and the id of the (64, 128] interval is 2, as shown in Table 1.
  • step S230 it is determined whether there is a source port and a serial number ID of a preset message length interval to which the message belongs in the corresponding relationship.
  • step S240 if it exists, the preset destination forwarding port of the packet in the port traffic is replaced with the port of the port to be adjusted in the corresponding relationship and the port of the adjusted port corresponding to the serial number ID of the preset packet length interval.
  • the adjusted port is taken out from the corresponding relationship as the preset destination forwarding port, otherwise, the original preset destination forwarding port dest is kept unchanged, and the packet forwarding is completed.
  • the traffic distribution method provided by the embodiment of the present disclosure further includes: Steps S250 to S260 are as follows.
  • step S250 a CRC operation is performed based on the load sharing mode of the aggregated link.
  • step S260 a preset destination forwarding port for port traffic is obtained according to the operation result.
  • the preset destination forwarding port dest is obtained according to the load sharing mode of the aggregated link.
  • the preset destination forwarding port obtained by different load sharing modes is different, but as long as the corresponding relationship exists with the source port of the packet and the preset destination of the packet. If the serial number ID of the packet length interval is the same as that of the port to be adjusted and the serial number ID of the preset packet length interval, the corresponding adjusted port in the corresponding relationship is used as the preset destination forwarding port.
  • a traffic distribution apparatus 10 for link aggregation provided by an embodiment of the present disclosure, the apparatus 10 includes a rate prediction module 100 and an adjustment module 120 .
  • the rate prediction module 100 is configured to predict the predicted average rate of the port in the current monitoring period and the average rate of all ports based at least on the port traffic in the previous monitoring period of the port, where the average rate is the average of the predicted average rates.
  • step S10 of the traffic distribution method provided by the embodiments of the present disclosure are also applicable to the implementation and functions of the rate prediction module 100 of the traffic distribution apparatus 10, and will not be repeated here.
  • the adjustment module 120 is configured to adjust the port traffic in the current monitoring period of the port to be adjusted to the adjusted port according to the load sharing element, wherein the load sharing element includes preset message length intervals of different rates in the port traffic, and the port to be adjusted is the port in the port to be adjusted. If the predicted traffic rate is greater than the average rate, the adjusted port is the port whose predicted traffic rate is less than the average rate.
  • step S20 of the flow distribution method provided by the embodiments of the present disclosure are also applicable to the implementation and functions of the adjustment module 120 of the flow distribution device 10, and are not repeated here.
  • an optical line terminal 1 is provided in an embodiment of the present disclosure.
  • the optical line terminal 1 includes the traffic distribution apparatus 10 for link aggregation as described above.
  • the traffic distribution apparatus 10 for link aggregation includes a rate prediction module 100 and an adjustment module 120 .
  • the rate prediction module 100 is configured to predict the predicted average rate of the port in the current monitoring period and the average rate of all ports based on the port traffic in the previous monitoring period of the port at least, and the average rate is the average of the predicted average rates .
  • the adjustment module 120 is configured to adjust the port traffic in the current monitoring period of the port to be adjusted to the adjusted port according to the load sharing element, wherein the load sharing element includes preset message length intervals of different rates in the port traffic, and the port to be adjusted is the port in the port to be adjusted. If the predicted traffic rate is greater than the average rate, the adjusted port is the port whose predicted traffic rate is less than the average rate.
  • An embodiment of the present disclosure provides a computer-readable storage medium, where one or more programs are stored in the storage medium, and the one or more programs can be executed by one or more processors, so as to realize the steps shown in FIG. 1 to FIG. 8 .
  • the steps of the traffic distribution method for aggregated links are shown, for example, the following steps S10 to S20 may be performed.
  • step S10 the predicted average rate of the port in the current monitoring period and the average rate of all ports are predicted based at least on the port traffic in the previous monitoring period of the port, and the average rate is the average of the predicted average rates.
  • step S20 the port traffic in the current monitoring period of the port to be adjusted is adjusted to the adjusted port according to the load balancing element, wherein the load balancing element includes preset message length intervals of different rates in the port traffic, and the port to be adjusted is predicted in the port For ports whose traffic rate is greater than the average rate, the adjusted port is the port whose predicted traffic rate is less than the average rate.
  • the load balancing element includes preset message length intervals of different rates in the port traffic
  • the traffic distribution method for an aggregated link uses the port traffic in at least the previous monitoring period of the port to predict the predicted average rate of the port in the current monitoring period and the average rate of all ports. and take the port whose predicted average rate is greater than the average rate as the port to be adjusted, and the port whose predicted average rate is less than the average rate as the port to be adjusted.
  • the port to be adjusted, the load sub-unit can include preset packet length intervals of different rates in port traffic, can be adjusted according to preset packet length intervals of different rates in port traffic, and can dynamically allocate the traffic of each physical port of the aggregated link , reduce the load difference of each physical port and reduce the waste of resources.
  • a typical implementation device is a computer.
  • the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or A combination of any of these devices.
  • Computer-readable storage media includes both persistent and non-permanent, removable and non-removable media, and storage of information can be implemented by any method or technology.
  • Information may be computer readable instructions, data structures, modules of programs, or other data.
  • Examples of computer storage media 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 Disc (DVD) or other optical storage, Magnetic tape cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device.
  • computer-readable media does not include transitory computer-readable media (Transitory Media), such as modulated data signals and carrier waves.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本公开涉及一种用于聚合链路的流量分配方法、装置、光线路终端OLT及存储介质。该用于聚合链路的流量分配方法包括：至少基于端口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和多个端口的速率平均值，速率平均值为预测平均速率的平均值；将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。

Description

用于聚合链路的流量分配方法、装置、光线路终端及存储介质
相关申请的交叉引用
本申请要求享有2020年07月31日提交的名称为“用于聚合链路的流量分配方法、装置、光线路终端及介质”的中国专利申请CN202010755745.4的优先权，其全部内容通过引用并入本申请中。
技术领域
本公开实施例涉及光线路终端OLT技术领域，尤其涉及一种用于聚合链路的流量分配方法、装置、光线路终端及存储介质。
背景技术
随着5G技术的推广，各种高带宽需求的APP如雨后春笋般不断涌现，这对综合接入媒体网关设备等OLT局端设备的数据转发提出了更高的要求。而在数据转发过程中，特别是超大流量的数据转发，即便是超强能力的转发芯片也可能会面临带宽不够、网络传输的可靠性不足和流量阻塞链路等问题。链路聚合技术提供了一种解决方案，链路聚合技术是将多个物理链路聚合到一起形成一条完整的逻辑链路。聚合后的链路带宽将会是各个物理端口带宽的总和，因此数据转发通道的带宽得以扩展，解决了带宽的瓶颈；另外将多条物理链路聚合在一起形成逻辑通道，意味着即便因为某种原因导致其中一条链路受损，数据也会被调整到使用其它链路来传输后完整地送到千家万户，毫无疑问地提高了网络的可靠性，解决了链路阻塞引发的连锁式风险。
但是IEEE802.3ad虽定义了链路聚合的框架结构和规则，却并没有定义使用何种方法来实施。多个物理端口做了链路聚合后可能会出现如下情景：高负荷传输的端口可能会遇到高突发流量，导致该端口流量超限。某些物理端口高负荷转发数据而其他物理端口长期处于低流量负载、带宽没有被充分利用的状态，造成了资源的浪费。因此，如何动态分配聚合链路各个物理端口的流量，尽可能减小各个物理端口的负荷差，成为亟待解决的问题。
发明内容
本公开实施例的目的是提供一种用于聚合链路的流量分配方法、装置、光线路终端OLT及存储介质，可以动态分配聚合链路各个物理端口的流量，减小各个物理端口的负荷差，减少资源浪费。
第一方面，本公开提出了一种用于聚合链路的流量分配方法，该方法包括：至少基于端 口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和所有端口的速率平均值，该速率平均值为预测平均速率的平均值；将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。
第二方面，本公开提出了一种用于聚合链路的流量分配装置，该装置包括：速率预测模块，用于至少基于端口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和所有端口的速率平均值，该速率平均值为预测平均速率的平均值；调整模块，用于将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。
第三方面，本公开提出了一种光线路终端，该光线路终端包括如上文所述用于聚合链路的流量分配装置。
第四方面，本公开提出了一种计算机可读存储介质，该存储介质存储有一个或多个程序，该一个或多个程序可被一个或多个处理器执行时，实现如上文所述用于聚合链路的流量分配方法的步骤。
附图说明
为了更清楚地说明本说明书一个或多个实施例或现有技术中的技术方案，下面将对一个或多个实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本说明书中记载的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动性的前提下，还可以根据这些附图获得其他的附图。
图1是本公开实施例提供一种用于聚合链路的流量分配方法的步骤示意图。
图2是本公开实施例提供另一种用于聚合链路的流量分配方法的步骤示意图。
图3是本公开实施例提供又一种用于聚合链路的流量分配方法的步骤示意图。
图4是本公开实施例提供又一种用于聚合链路的流量分配方法的步骤示意图。
图5是本公开实施例提供又一种用于聚合链路的流量分配方法的步骤示意图。
图6是本公开实施例提供又一种用于聚合链路的流量分配方法的步骤示意图。
图7是本公开实施例提供又一种用于聚合链路的流量分配方法的步骤示意图。
图8是本公开实施例提供又一种用于聚合链路的流量分配方法的步骤示意图。
图9是本公开实施例提供一种用于聚合链路的流量分配装置的结构示意图。
图10是本公开实施例提供一种光线路终端的结构示意图。
图11是本公开实施例提供一种用于聚合链路的流量分配方法中流量预测示意图。
图12是本公开实施例提供一种用于聚合链路的流量分配方法中流量调整示意图。
具体实施方式
为了使本技术领域的人员更好地理解本说明书中的技术方案，下面将结合本说明书一个或多个实施例中的附图，对本说明书一个或多个实施例中的技术方案进行清楚、完整地描述，显然，所描述的一个或多个实施例仅仅是本说明书一部分实施例，而不是全部的实施例。基于本说明书中的一个或多个实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都应当属于本文件的保护范围。
本公开实施例提供的一种用于聚合链路的流量分配方法适用于聚合链路，可以在聚合链路的各个物理端口之间动态调整端口流量，实现了聚合链路的流量动态分配，减小各个物理端口的负荷差。下面将详细地描述本说明书提供的用于聚合链路的流量分配方法及其各个步骤。
需要说明的是，本公开实施例提供的用于聚合链路的流量分配方法中，聚合链路的各个物理端口在当前监控周期下总的端口流量保持不变的情况下，在各个端口之间动态调整部分物理端口的端口流量，待调整端口和被调整端口均是聚合链路上的端口。
实施例一
参照图1所示，为本公开实施例提供的一种用于聚合链路的流量分配方法的步骤示意图。可以理解的是，本公开实施例提供的用于聚合链路的流量分配方法适用于具有聚合链路的光线路终端OLT。该用于聚合链路的流量分配方法包括以下步骤S10至S20。
在步骤S10，至少基于端口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和所有端口的速率平均值，该速率平均值为预测平均速率的平均值。
本公开实施例的新颖之处一是可以根据历史端口流量预测当前监控周期内端口的预测平均速率，满足复杂流量场的流量需求。通过当前网络监控模块可以直接获取各个端口的端口流量，不必引入额外的计算开销，为端口流量的快速分配节省了时间。二是根据报文长度划分负荷分担元，采用负荷分担元作为调整粒度对各个端口的端口流量进行分配，而当前的网络监控模块可以直接获取报文长度的统计，同样为端口流量的快速分配节省时间。
当前网络监控模块可以周期性地获取聚合链路中每一个端口的流量统计信息和每一个 端口流量中报文长度的统计信息。当前监控周期内端口的预测平均速率是该端口在当前监控周期内端口流量的通过率，即当前监控周期内端口流量总和除以当前监控周期。
从当前网络监控模块中读取每一个端口的流量统计信息后，计算在上一个监控周期内每一个端口的端口流量的平均速率，将一个监控周期内对于一个端口的端口流量的平均速率视为一个采样样本，至少可以基于历史出现的上一个监控周期的端口流量的平均速率预测当前监控周期内端口的预测平均速率，并且可以基于每一个端口的预测平均速率得到所有端口的速率平均值，为下面待调整端口和被调整端口的区分做准备。
在步骤S20，将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。
当前的网络监控模块周期性地获取每一个端口的流量统计信息和每一个端口流量中报文长度的统计信息。基于上述每一个端口流量中报文长度的统计信息可以计算出当前监控周期内每一个预设报文长度区间的速率，预设报文长度区间可以根据流量分配要求设定，比如每次需要调整的负荷分担元的粒度大一点可以将预设报文长度区间设定的报文长度范围大一些，反之亦然。预设报文长度区间的速率是当前监控周期内该预设报文长度区间内不同长度报文的通过率，即当前监控周期内不同长度报文的流量总和除以当前监控周期。
待调整端口当前监控周期内的预测平均速率大于速率平均值，因此待调整端口在当前监控周期内的端口流量需要调整，被调整端口在当前监控周期内的端口流量则需要从待调整端口调整一些过来，即需要与待调整端口分担流量。
预设报文长度区间的定义可以参见表1，表1中对不同的预设报文长度区间做了编码ID，以方便后续调用，表1中不同速率的预设报文长度区间成为流量分配可以调整的最小颗粒度。
表1
预设报文长度区间 预设报文长度区间ID 速率
(0，64] 1 rate_1
(64，128] 2 rate_2
(128，256] 3 rate_3
(256，512] 4 rate_4
(512，1024] 5 rate_5
(1024，1500] 6 rate_6
参见图2所示，在一些实施例中，本公开实施例提供的流量分配方法中，上述步骤S10中至少基于端口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率，还包括如下步骤S100至S110。
在步骤S100，基于端口流量获取端口的N个采样样本，N个采样样本包括上一个监控周期在内的N个监控周期内端口流量的平均速率。
可以根据每个端口的流量统计信息计算出在上一个监控周期内每个端口的平均速率，即该上一个监控周期内的流量总和除以该一个监控周期。将上一个监控周期内的平均速率视为一个采样样本，将包括上一个监控周期在内的若干个监控周期视为一个采样时间窗口。当监控时间累计达到有效采样时间阈值时得到N个数据样本，对该N个数据样本建立最小二乘数学模型，利用该最小二乘数学模型预测当前监控周期内每个端口的预测平均速率。
在步骤S110，基于N个采样样本预测当前监控周期内端口的预测平均速率。
获取N个采样样本后，可以基于最小二乘数学模型或者其他数学算法对当前监控周期内端口流量的预测平均速率进行预测，满足复杂流量场的流量需求。
具体的流量分配方法的执行步骤可以参见下面的一个示例。
在初始化网络监控模块的监控参数中，监控周期配置为T秒，缺省为30s；有效采样时间阈值为T sample秒，缺省为300s；采样时间窗口为N×T秒，N缺省为10；监控周期的序号因变量为n，初始化为1。定义负荷分担元为端口进行负荷分配时能够调整的最小单元，这里为报文长度区间，初始化报文长度区间表如表1所示。
开始启动网络监控模块，当前开始时间记为t start，获取端口号，假设为p＝{p 1，...，p n}。
针对第一个端口p 1，获取第一个监控周期内的流量，记为cnt 1n，计算该端口在该监控周期内的平均速率为rate 1n＝cnt 1n/T，采用同样的方法计算其他端口的平均速率，该平均速率作为采样样本的值存储到内存。
判定n是否小于N，如果n小于N，则等待下一个监控周期到来，并令n＝n+1，跳转到上一步骤继续执行，否则执行下一步骤。
按照最小二乘数学模型，以rate 1(n-1)，...，rate 1(n-N)为样本，预测出第一个端口在n+1个监控周期的预测平均速率，比如采用历史N个采样样本得到当前监控周期内端口的预测平均速率，记为rate 1(n+1)。采用同样的方法预测出其他端口在n+1个监控周期比如当前监控周期的预测平均速率，这样得到预测平均速率集合
Figure PCTCN2021108872-appb-000001
参见图3所示，在一些实施例中，基于N个采样样本利用最小二乘数学模型预测当前监控周期内端口的预测平均速率之后，本公开实施例提供的流量分配方法还包括如下步骤 S120至S140。
在步骤S120，分别得到所有端口的预测平均速率。
正如上文所述可以基于采集的历史N个采样样本利用最小二乘数学模型预测所有端口在当前监控周期内的预测平均速率，参见图11所示。
在步骤S130，对所有端口的预测平均速率求平均值得到速率平均值。
对得到的所有端口在当前监控周期内的预测平均速率求平均值得到速率平均值，目的是基于所有端口的预测平均速率和速率平均值区分出待调整端口和被调整端口。
在步骤S140，基于预测平均速率和速率平均值确定待调整端口和被调整端口。
待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。确定出待调整端口和被调整端口的目的是实现聚合链路流量的重新分配。待调整端口当前监控周期内的预测平均速率大于速率平均值，因此待调整端口在当前监控周期内的端口流量需要调整，被调整端口则在当前监控周期内的端口流量需要从待调整端口调整一些过来，即需要与待调整端口分担流量。
参见图4所示，在一些实施例中，在上述步骤S20中将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口之前，本公开实施例提供的流量分配方法还包括如下步骤S30至S40。
在步骤S30，将所有端口按照预测平均速率从大到小排序，形成端口集合。
按照预测平均速率从大到小对所有端口进行排序形成端口集合，该集合记为C1。
在步骤S40，从端口集合中顺次选取待调整端口后，基于端口流量中报文的长度计算当前监控周期内待调整端口的报文所属预设报文长度区间的速率，直至得到所有待调整端口的报文所属预设报文长度区间的速率。
基于当前监控周期内统计端口流量中报文的长度，可以相应地统计出报文所属预设报文长度区间的流量除以当前监控周期时间得到当前监控周期内报文所属预设报文长度区间的速率。因此获取一个端口在当前监控周期的每个负荷分担元的流量，方法类似于监控周期内端口流量的统计，然后依次计算出每个负荷分担元的速率，并将每一个端口的端口流量中所有负荷分单元按照速率从小到大排序。
假设记为：rate_elem p1＝{rate p1，len1，rate p1，len2，rate p1，len3，rate p1，len4}。对其他端口采用同样的操作方法计算出其他端口的端口流量中所有负荷分担元的速率并且按照速率从小到大的顺序排序。采用上面的方法可以得到所有待调整端口的报文所属预设报文长度区间的速率。
参见图5所示，在一些实施例中，上述步骤S20中将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，还包括如下步骤S200至S210。
步骤S200：按照预设报文长度区间的速率由小到大顺次将端口流量以预设报文长度区间为粒度从待调整端口调整至被调整端口。
如果报文长度单一，则只有一个预设报文长度区间，并且该预设报文长度区间只有一个速率，因此只能按照该单一速率的单一预设报文长度区间作为调整粒度将待调整端口的端口流量进行分配。
从所有的端口中取出一个端口，比如可以从端口集合C1中取出第一个端口，该第一个端口的端口号记为p balance_1，由于端口集合C1是按照预测平均速率从大到小对所有端口进行排序，第一个端口在当前监控周期内的预测平均速率最大，因此第一个端口在当前监控周期的负荷是最大的，需要首先被分担。从第一个端口p balance_1的所有负荷分单元
Figure PCTCN2021108872-appb-000002
中选择负荷分担元进行流量分配，并且记下负荷分担元的编号ID、待调整端口的端口号p balance_1和对应的被调整端口的端口号的对应关系，直至第一个端口的预测平均速率被分配到速率平均值附近。参见图12所示，按照预设报文长度区间的速率由小到大顺次将第一端口的端口流量以预设报文长度区间为粒度从待调整端口调整至被调整端口，图12最左边的方向柱以不同颜色的条形块表示不同速率的预设报文长度区间，其中面积越小的条形块表示速率越小，对于第一端口的端口流量，依次从预设报文长度区间的速率由小到大顺次从第一个端口调整至被调整端口，被调整端口优先选择当前监控周期内预测平均速率与速率平均值相差最大的端口。做重复操作，直到第一端口的预测平均速率被分配到速率平均值附近。做重复操作，直到所有的被调整端口的预测平均速率全部被分配到速率平均值附近。
在步骤S210，形成待调整端口、被调整端口和预设报文长度区间的编号ID的对应关系。
待调整端口、被调整端口和预设报文长度区间的编号ID的对应关系可以作为聚合链路的均匀负荷分配表，如表2所示。
表2
待调整端口 预设报文长度区间ID 被调整端口
SRC_PORT1 1 ID_N
... ... ...
PORT_N 4 DEST_PORT_N
在当前监控周期内获取用户数据报文后，可以在统计每一个待调整端口的端口流量中不同速率的报文所属的预设报文长度区间后，查询均匀负荷分配表获得被调整端口，从而可以对聚合链路上所有物理端口进行流量分配。
参见图6所示，在一些实施例中，上述步骤S200中按照预设报文长度区间的速率由小到大顺次将端口流量以预设报文长度区间为粒度从待调整端口调整至被调整端口，还包括如下步骤S201。
在步骤S201，采用贪心算法将端口流量以预设报文长度区间为粒度从待调整端口调整至被调整端口，直至待调整端口的预测平均速率接近速率平均值。
参见图12所示，按照贪心算法将待调整端口的端口流量以预设报文长度区间为粒度调整至被调整端口，并且记下负荷分担元的编号ID、待调整端口和对应的被调整端口号。贪心算法可以保证待调整端口的预测平均速率尽可能地接近速率平均值。
参见图7所示，在一些实施例中，在所述步骤S210中形成待调整端口的端口、被调整端口的端口和预设报文长度区间的编号ID的对应关系，在此之后，本公开实施例提供的流量分配方法还包括如下步骤S220至S240。
在步骤S220，基于端口流量中报文携带的参数确定报文的源端口、报文所属的预设报文长度区间及预设报文长度区间的编号ID。
提取端口流量中报文携带的参数，比如报文头的len字段，获取聚合链路的物理端口个数，该报文的源端口，并且基于len字段判断该报文所属预设报文长度区间。预设报文长度区间可以设定(0，64]，(64，128]，(128，256]，(256，512]，(512，1024]，(1024，1518]六个区间，获得该报文所属预设报文长度区间的编号ID，比如(0，64]区间的id为1，(64，128]区间的id为2，参见表1所示。
在步骤S230，判断对应关系中是否存在源端口和报文所属预设报文长度区间的编号ID。
以该报文的参数获取的源端口和报文所属预设报文长度区间的编号ID作为索引查询对应关系(参见表2)中是否存在相同的待调整端口和预设报文长度区间的编号ID。
在步骤S240，如果存在，则将端口流量中报文的预设目的转发端口替换为对应关系中待调整端口的端口和预设报文长度区间的编号ID对应的被调整端口的端口。
如果存在，则从对应关系中取出被调整端口作为预设目的转发端口，否则保持原有的预设目的转发端口dest不变，完成报文的转发。
参见图8所示，在一些实施例中，上述步骤S230中判断对应关系中是否存在源端口和报文所属预设报文长度区间的编号ID之前，本公开实施例提供的流量分配方法还包括如下 步骤S250至S260。
在步骤S250，基于聚合链路的负荷分担模式进行CRC运算。
原有的目的转发端口dest的获得可以是基于聚合链路的负荷分担模式来确定。如果负荷分担模式基于源端口的mac地址，取源端口的mac地址做CRC32运算，将得到的结果记为mac_crc，然后将其对聚合链路的所有端口的个数取模作为预设目的转发端口dest，即：dest＝mac_crc％port_num。
在步骤S260，根据运算结果得到端口流量的预设目的转发端口。
预设目的转发端口dest是根据聚合链路的负荷分担模式得到的，不同的负荷分担模式得到的预设目的转发端口不同，但是只要对应关系中存在与报文的源端口和报文所属预设报文长度区间的编号ID相同的待调整端口和预设报文长度区间的编号ID，则将对应关系中对应的被调整端口作为预设目的转发端口。
实施例二
参照图9所示，为本公开实施例提供的一种用于聚合链路的流量分配装置10，该装置10包括速率预测模块100和调整模块120。
速率预测模块100用于至少基于端口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和所有端口的速率平均值，该速率平均值为预测平均速率的平均值。
上文针对本公开实施例提供的流量分配方法的步骤S10描述的所有特征、细节和优势同样适用于该流量分配装置10的速率预测模块100的实施和功能，在此不再赘述。
调整模块120用于将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。
上文针对本公开实施例提供的流量分配方法的步骤S20描述的所有特征、细节和优势同样适用于该流量分配装置10的调整模块120的实施和功能，在此不再赘述。
实施例三
参见图10所示，为本公开实施例提供的一种光线路终端1，该光线路终端1包括如上文所述用于聚合链路的流量分配装置10。该用于聚合链路的流量分配装置10包括速率预测模块100和调整模块120。
如上所述，速率预测模块100用于至少基于端口上一个监控周期内的端口流量预测当 前监控周期内端口的预测平均速率和所有端口的速率平均值，该速率平均值为预测平均速率的平均值。
调整模块120用于将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。
上文针对本公开实施例提供的流量分配方法描述的所有特征、细节和优势同样适用于该光线路终端1的实施和功能，在此不再赘述。
实施例四
本公开实施例提供的一种计算机可读存储介质，该存储介质存储有一个或多个程序，该一个或多个程序可被一个或多个处理器执行，以实现如图1至图8所示用于聚合链路的流量分配方法的步骤，例如可以执行以下步骤S10至S20。
在步骤S10，至少基于端口上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和所有端口的速率平均值，速率平均值为预测平均速率的平均值。
在步骤S20，将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中负荷分担元包括端口流量中不同速率的预设报文长度区间，待调整端口为端口中预测流量速率大于速率平均值的端口，被调整端口为端口中预测流量速率小于速率平均值的端口。
上文针对本公开实施例提供的流量分配方法描述的所有特征、细节和优势同样适用于该存储介质的实施和功能，在此不再赘述。
通过以上分析可以看出，本公开实施例提供的用于聚合链路的流量分配方法利用端口的至少上一个监控周期内的端口流量预测当前监控周期内端口的预测平均速率和所有端口的速率平均值，并且将预测平均速率大于速率平均值的端口作为待调整端口，预测平均速率小于速率平均值的端口作为被调整端口，目的是将待调整端口的端口流量按照负荷分单元为调整粒度调整至被调整端口，负荷分单元可以包括端口流量中不同速率的预设报文长度区间，可以按照端口流量中不同速率的预设报文长度区间进行调整，可以动态分配聚合链路各个物理端口的流量，减小各个物理端口的负荷差，减少资源浪费。
总之，以上所述仅为本说明书的较佳实施例而已，并非用于限定本说明书的保护范围。凡在本说明书的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本说明书的保护范围之内。
上述一个或多个实施例阐明的系统、装置、模块或单元，具体可以由计算机芯片或实体实现，或者由具有某种功能的产品来实现。一种典型的实现设备为计算机。具体的，计算机例如可以为个人计算机、膝上型计算机、蜂窝电话、相机电话、智能电话、个人数字助理、媒体播放器、导航设备、电子邮件设备、游戏控制台、平板计算机、可穿戴设备或者这些设备中的任何设备的组合。
计算机可读存储介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(Transitory Media)，如调制的数据信号和载波。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、商品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、商品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、商品或者设备中还存在另外的相同要素。
本说明书中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于系统实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
上述对本说明书特定实施例进行了描述。其它实施例在所附权利要求书的范围内。在一些情况下，在权利要求书中记载的动作或步骤可以按照不同于实施例中的顺序来执行并且仍然可以实现期望的结果。另外，在附图中描绘的过程不一定要求示出的特定顺序或者连续顺序才能实现期望的结果。在某些实施方式中，多任务处理和并行处理也是可以的或者可能是有利的。

Claims (11)

  1. 一种用于聚合链路的流量分配方法，包括：
    至少基于端口上一个监控周期内的端口流量，预测当前监控周期内所述端口的预测平均速率和所有所述端口的速率平均值，所述速率平均值为所述预测平均速率的平均值；
    将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中所述负荷分担元包括所述端口流量中不同速率的预设报文长度区间，所述待调整端口为所述端口中所述预测流量速率大于所述速率平均值的端口，所述被调整端口为所述端口中所述预测流量速率小于所述速率平均值的端口。
  2. 如权利要求1所述的流量分配方法，所述至少基于端口上一个监控周期内的端口流量预测当前监控周期内所述端口的预测平均速率，还包括：
    基于所述端口流量获取所述端口的N个采样样本，所述N个采样样本包括上一个监控周期在内的N个监控周期内所述端口流量的平均速率；
    基于所述N个采样样本预测所述当前监控周期内所述端口的预测平均速率。
  3. 如权利要求2所述的流量分配方法，所述基于所述N个采样样本预测所述当前监控周期内所述端口的预测平均速率之后，所述方法还包括：
    分别得到所有所述端口的预测平均速率；
    对所有所述端口的预测平均速率求平均值得到速率平均值；
    基于所述预测平均速率和所述速率平均值确定所述待调整端口和所述被调整端口。
  4. 如权利要求1或2所述的流量分配方法，所述将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口之前，所述方法还包括：
    将所有所述端口按照预测平均速率从大到小排序，形成端口集合；
    从所述端口集合中选取所述待调整端口后，基于所述端口流量中报文的长度计算当前监控周期内所述待调整端口的所述报文所属预设报文长度区间的速率，直至得到所有所述待调整端口的所述报文所属预设报文长度区间的速率。
  5. 如权利要求4所述的流量分配方法，所述将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，还包括：
    按照所述预设报文长度区间的速率由小到大顺次将所述端口流量以所述预设报文长度区间为粒度从所述待调整端口调整至所述被调整端口；
    形成所述待调整端口、所述被调整端口和所述预设报文长度区间的编号ID的对应关系。
  6. 如权利要求5所述的流量分配方法，所述按照所述预设报文长度区间的速率由小到大顺次将所述端口流量以所述预设报文长度区间为粒度从所述待调整端口调整至所述被调整端口，还包括：
    采用贪心算法将所述端口流量以所述预设报文长度区间为粒度从所述待调整端口调整至所述被调整端口，直至所述待调整端口的预测平均速率接近所述速率平均值。
  7. 如权利要求5所述的流量分配方法，所述形成所述待调整端口的端口、所述被调整端口的端口和所述预设报文长度区间的编号ID的对应关系之后，所述方法还包括：
    基于所述端口流量中报文携带的参数确定所述报文的源端口、所述报文所属的预设报文长度区间及所述预设报文长度区间的编号ID；
    判断所述对应关系中是否存在所述源端口和所述报文所属预设报文长度区间的编号ID；
    如果存在，则将所述端口流量中所述报文的预设目的转发端口替换为所述对应关系中所述待调整端口的端口和所述预设报文长度区间的编号ID对应的所述被调整端口的端口。
  8. 如权利要求7所述的流量分配方法，所述判断所述对应关系中是否存在所述源端口和所述报文所属预设报文长度区间的编号ID之前，所述方法还包括：
    基于所述聚合链路的负荷分担模式进行CRC运算；
    根据运算结果得到所述端口流量的预设目的转发端口。
  9. 一种用于聚合链路的流量分配装置，包括：
    速率预测模块，用于至少基于所述端口上一个监控周期内的端口流量预测当前监控周期内所述端口的预测平均速率和所有所述端口的速率平均值，所述速率平均值为所述预测平均速率的平均值；
    调整模块，用于将待调整端口当前监控周期内的端口流量按照负荷分担元调整至被调整端口，其中所述负荷分担元包括所述端口流量中不同速率的预设报文长度区间，所述待调整端口为所述端口中所述预测流量速率大于所述速率平均值的端口，所述被调整端口为所述端口中所述预测流量速率小于所述速率平均值的端口。
  10. 一种光线路终端，包括如权利要求9所述用于聚合链路的流量分配装置。
  11. 一种计算机可读存储介质，所述存储介质存储有一个或多个程序，所述一个或多个程序可被一个或多个处理器执行时，实现如权利要求1至8中任一项所述用于聚合链路的流量分配方法的步骤。
PCT/CN2021/108872 2020-07-31 2021-07-28 用于聚合链路的流量分配方法、装置、光线路终端及存储介质 WO2022022558A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010755745.4 2020-07-31
CN202010755745.4A CN114095806A (zh) 2020-07-31 2020-07-31 用于聚合链路的流量分配方法、装置、光线路终端及介质

Publications (1)

Publication Number Publication Date
WO2022022558A1 true WO2022022558A1 (zh) 2022-02-03

Family

ID=80037183

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/108872 WO2022022558A1 (zh) 2020-07-31 2021-07-28 用于聚合链路的流量分配方法、装置、光线路终端及存储介质

Country Status (2)

Country Link
CN (1) CN114095806A (zh)
WO (1) WO2022022558A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114979008A (zh) * 2022-05-19 2022-08-30 中国电子科技集团公司第五十八研究所 链路流量控制额度调整方法、装置、片上处理系统
CN116915517A (zh) * 2023-09-14 2023-10-20 厦门快快网络科技有限公司 一种云服务资源风险安全管理方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116149865B (zh) * 2023-04-10 2023-07-21 之江实验室 一种变频执行任务的方法、装置、设备及可读存储介质

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1484413A (zh) * 2002-09-18 2004-03-24 ����ͨѶ�ɷ����޹�˾ 一种实现聚合链路流量均衡的方法
CN103973599A (zh) * 2014-04-25 2014-08-06 中国科学院计算技术研究所 一种基于OpenFlow的信道分配方法及其装置
CN106453111A (zh) * 2015-08-11 2017-02-22 中兴通讯股份有限公司 基于聚合链路的流量管理方法及装置
CN109218216A (zh) * 2017-06-29 2019-01-15 中兴通讯股份有限公司 链路聚合流量分配方法、装置、设备及存储介质
CN110537384A (zh) * 2017-02-06 2019-12-03 奥提欧斯塔网络公司 用于长期演进通信系统的多技术聚合架构
US10579955B1 (en) * 2015-06-30 2020-03-03 Auctane, LLC Methods and systems for providing multi-carrier/multi-channel/multi-national shipping
US10652092B2 (en) * 2016-08-28 2020-05-12 Vmware, Inc. Methods and systems that collect and manage latency data in an automated resource-exchange system
US10715448B2 (en) * 2018-05-25 2020-07-14 Software Ag System and/or method for predictive resource management in file transfer servers

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1484413A (zh) * 2002-09-18 2004-03-24 ����ͨѶ�ɷ����޹�˾ 一种实现聚合链路流量均衡的方法
CN103973599A (zh) * 2014-04-25 2014-08-06 中国科学院计算技术研究所 一种基于OpenFlow的信道分配方法及其装置
US10579955B1 (en) * 2015-06-30 2020-03-03 Auctane, LLC Methods and systems for providing multi-carrier/multi-channel/multi-national shipping
CN106453111A (zh) * 2015-08-11 2017-02-22 中兴通讯股份有限公司 基于聚合链路的流量管理方法及装置
US10652092B2 (en) * 2016-08-28 2020-05-12 Vmware, Inc. Methods and systems that collect and manage latency data in an automated resource-exchange system
CN110537384A (zh) * 2017-02-06 2019-12-03 奥提欧斯塔网络公司 用于长期演进通信系统的多技术聚合架构
CN109218216A (zh) * 2017-06-29 2019-01-15 中兴通讯股份有限公司 链路聚合流量分配方法、装置、设备及存储介质
US10715448B2 (en) * 2018-05-25 2020-07-14 Software Ag System and/or method for predictive resource management in file transfer servers

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114979008A (zh) * 2022-05-19 2022-08-30 中国电子科技集团公司第五十八研究所 链路流量控制额度调整方法、装置、片上处理系统
CN116915517A (zh) * 2023-09-14 2023-10-20 厦门快快网络科技有限公司 一种云服务资源风险安全管理方法
CN116915517B (zh) * 2023-09-14 2023-11-24 厦门快快网络科技有限公司 一种云服务资源风险安全管理方法

Also Published As

Publication number Publication date
CN114095806A (zh) 2022-02-25

Similar Documents

Publication Publication Date Title
WO2022022558A1 (zh) 用于聚合链路的流量分配方法、装置、光线路终端及存储介质
US8780711B2 (en) Coordinated back-off mechanism for path selection in hybrid communication networks
US9391749B2 (en) System and method for distributed data management in wireless networks
US7496367B1 (en) Method of multi-carrier traffic allocation for wireless communication system
CN110248417B (zh) 一种电力物联网中上行通信业务的资源分配方法及系统
US8989037B2 (en) System for performing data cut-through
CN108446179B (zh) 基于负载均衡机制的权重优先Task任务调度方法
US20130021928A1 (en) Evaluating a capacity of a cell of a radio access network
WO2017041492A1 (zh) 一种资源分配方法、装置和计算机可读存储介质
US20200403913A1 (en) Network Resource Scheduling Method, Apparatus, Electronic Device and Storage Medium
CN109617810B (zh) 数据传输方法及装置
CN113225253B (zh) 一种报文转发方法及装置
US10855603B2 (en) Methods and systems for multi-level network capacity allocation
WO2022121079A1 (zh) 一种流量转发设备的链路聚合方法及流量转发设备
CN113328953A (zh) 网络拥塞调整的方法、装置和存储介质
Liebeherr et al. Rate allocation and buffer management for differentiated services
CN115629717B (zh) 一种基于分布式存储的负载均衡方法和存储介质
CN115967687A (zh) 快转发报文方法、网络设备、存储介质及计算机程序产品
CN114500544A (zh) 一种节点间负载均衡方法、系统、设备以及介质
WO2020135648A1 (zh) 资源调度方法、装置、网络设备和可读存储介质
CN114785744B (zh) 数据处理方法、装置、计算机设备和存储介质
CN106572501B (zh) 基于双重阈值判决的内容中心移动自组织网络缓存方法
Katsumoto et al. A bandwidth assignment method for downloading large files with time constraints
CN116847360B (zh) 一种非实时数据传输方法、装置及存储介质
CN117938750B (zh) 调度路由信息的处理方法、装置、设备、存储介质及产品

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: 21848858

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 28/06/2023)

122 Ep: pct application non-entry in european phase

Ref document number: 21848858

Country of ref document: EP

Kind code of ref document: A1