WO2020093502A1 - 一种调整额定带宽的方法和装置 - Google Patents

一种调整额定带宽的方法和装置 Download PDF

Info

Publication number
WO2020093502A1
WO2020093502A1 PCT/CN2018/119597 CN2018119597W WO2020093502A1 WO 2020093502 A1 WO2020093502 A1 WO 2020093502A1 CN 2018119597 W CN2018119597 W CN 2018119597W WO 2020093502 A1 WO2020093502 A1 WO 2020093502A1
Authority
WO
WIPO (PCT)
Prior art keywords
product line
bandwidth
real
adjustment
health value
Prior art date
Application number
PCT/CN2018/119597
Other languages
English (en)
French (fr)
Inventor
赵瑞
Original Assignee
网宿科技股份有限公司
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 网宿科技股份有限公司 filed Critical 网宿科技股份有限公司
Priority to EP18939669.0A priority Critical patent/EP3876485A4/en
Priority to US17/044,608 priority patent/US20210111961A1/en
Publication of WO2020093502A1 publication Critical patent/WO2020093502A1/zh

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08Configuration management of networks or network elements
    • H04L41/0896Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0876Network utilisation, e.g. volume of load or congestion level
    • H04L43/0882Utilisation of link capacity
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0876Network utilisation, e.g. volume of load or congestion level
    • H04L43/0894Packet rate
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/50Testing arrangements
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/22Traffic shaping
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/24Traffic characterised by specific attributes, e.g. priority or QoS
    • H04L47/2416Real-time traffic
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/28Flow control; Congestion control in relation to timing considerations

Definitions

  • the invention relates to the technical field of data transmission, in particular to a method and device for adjusting a rated bandwidth.
  • the rated bandwidth of a network device refers to the total amount of network traffic data allowed to pass through per unit time under normal working conditions of the network device. During the transmission of network traffic in the data transmission network, it is often necessary to determine the current bandwidth load of each network device according to the rated bandwidth of each network device in the data transmission network, and then the specific transmission path of the network traffic can be determined.
  • the rated bandwidth of the network equipment can generally be evaluated by stress testing. Specifically, technicians can first randomly select multiple network equipment, and then Detect the maximum amount of network traffic that each selected network device can transmit within a period of time, and then determine the calculated average or mode of the rated bandwidth of the above multiple network devices as the rated network device bandwidth.
  • the rated bandwidth obtained through sample testing cannot accurately reflect the respective rated bandwidth of each network device; second, after the network device is put into use, it often supports multiple product lines at the same time (it can be divided according to the type of data transmitted) For text transmission product lines, picture transmission product lines, video transmission product lines, etc.), the rated bandwidth of network equipment will also vary with the ratio of the business load of multiple product lines; third, Over time, problems such as hardware aging or faults will cause the rated bandwidth of network devices to decrease. In summary, the rated bandwidth determined by the stress test will not be effectively used in the actual planning and scheduling process of transmission network traffic. Using the above-mentioned rated bandwidth to determine the transmission path will have a greater impact on the transmission quality of the data transmission network.
  • the embodiments of the present invention provide a method and device for adjusting the rated bandwidth.
  • the technical solution is as follows:
  • a method for adjusting a rated bandwidth includes:
  • any product line of any network device in the data transmission network periodically determine the adjustment method of the rated bandwidth of the product line according to the preset reference health value of the product line and the real-time health value within the current period;
  • the adjustment method of the rated bandwidth is upward adjustment, then according to the fluctuation of the real-time health value of the product line in the current period, calculate the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line;
  • the current rated bandwidth corresponding to the product line is adjusted based on the preset downward adjustment ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period.
  • the method further includes:
  • an up-scaling ratio based on the rated bandwidth corresponding to the product line is calculated and stored.
  • the determining the conventional bandwidth value of the network device based on the historical bandwidth data of the network device every preset duration includes:
  • the conventional bandwidth value of each network device under the category is determined in all historical bandwidth data of the network devices under the category.
  • the determining the regular health value of the product line according to the historical health value of the product line every preset time period includes:
  • all product lines are classified by the EM clustering algorithm
  • the general health value of each product line under the category is determined from all historical health values of the product line under the category.
  • calculating and storing an up-scaling ratio corresponding to the rated bandwidth corresponding to the product line includes:
  • a base upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated and stored.
  • the periodically determining the adjustment method of the rated bandwidth of the product line according to the preset reference health value of the product line and the real-time health value in the current period includes:
  • the adjustment method of the rated bandwidth of the product line is upward adjustment, otherwise it is downward adjustment.
  • the calculating the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line according to the fluctuation of the real-time health value of the product line in the current period includes:
  • the median real-time bandwidth of the network device in the current period is greater than the median real-time bandwidth of the network device in the previous week and greater than the current rated bandwidth corresponding to the product line, then according to the current period In the fluctuation of the real-time health value of the product line, calculate the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line, otherwise, set the rated bandwidth adjustment mode to no adjustment.
  • the calculating the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line according to the fluctuation of the real-time health value of the product line in the current period includes:
  • the calculating the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line according to the historical upward adjustment ratio of the rated bandwidth corresponding to the product line in the historical period and the latest smooth gradient includes:
  • the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated.
  • the health value is calculated and generated according to the performance index of the network device and the business parameters of the product line.
  • an apparatus for adjusting a rated bandwidth includes:
  • the determination module is used to periodically determine the rated bandwidth of the product line for any product line of any network device in the data transmission network according to the preset reference health value of the product line and the real-time health value in the current period Adjustment method
  • An upward adjustment module used to calculate the real-time upward adjustment ratio of the nominal bandwidth corresponding to the product line according to the fluctuation of the real-time health value of the product line in the current period if the nominal bandwidth adjustment mode is upward adjustment Adjusting the ratio in real time and adjusting the ratio based on the pre-stored rated bandwidth corresponding to the product line, and adjusting the current rated bandwidth corresponding to the product line;
  • the downward adjustment module is used to adjust the current rated bandwidth corresponding to the product line based on the preset downward adjustment ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period if the rated bandwidth adjustment mode is downward adjustment Adjustment.
  • the device further includes a calculation module for:
  • an up-scaling ratio based on the rated bandwidth corresponding to the product line is calculated and stored.
  • calculation module is specifically used for:
  • the conventional bandwidth value of each network device under the category is determined in all historical bandwidth data of the network devices under the category.
  • calculation module is specifically used for:
  • all product lines are classified by the EM clustering algorithm
  • the general health value of each product line under the category is determined among all historical health values of the product line under the category.
  • calculation module is specifically used for:
  • a base upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated and stored.
  • the determination module is specifically used for:
  • the adjustment method of the rated bandwidth of the product line is upward adjustment, otherwise it is downward adjustment.
  • the upward adjustment module is specifically used for:
  • the median real-time bandwidth of the network device in the current period is greater than the median real-time bandwidth of the network device in the previous week and greater than the current rated bandwidth corresponding to the product line, then according to the current period In the fluctuation of the real-time health value of the product line, calculate the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line, otherwise, set the rated bandwidth adjustment mode to no adjustment.
  • the upward adjustment module is specifically used for:
  • the upward adjustment module is specifically used for:
  • the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated.
  • the health value is calculated and generated according to the performance index of the network device and the business parameters of the product line.
  • a rated bandwidth adjustment device includes a processor and a memory.
  • the memory stores at least one instruction, at least one program, code set, or instruction set.
  • the at least one instruction The at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method for adjusting the rated bandwidth as described in the first aspect.
  • a computer-readable storage medium in which at least one instruction, at least one program, code set or instruction set is stored in the storage medium, the at least one instruction, the at least one program, the code The set or instruction set is loaded and executed by the processor to implement the method for adjusting the rated bandwidth as described in the first aspect.
  • any product line of any network device in the data transmission network periodically determine the adjustment method of the rated bandwidth of the product line according to the preset reference health value of the product line and the real-time health value in the current cycle, Among them, the health value is calculated and generated according to the performance indicators of the network equipment and the business parameters of the product line; if the rated bandwidth adjustment method is up, the corresponding bandwidth of the product line is calculated according to the fluctuation of the real-time health value of the product line in the current cycle Adjust the ratio in real time; adjust the ratio based on the real-time adjustment ratio and the pre-stored rated bandwidth corresponding to the product line, and adjust the current corresponding rated bandwidth of the product line.
  • the basic adjustment ratio is based on the historical bandwidth data of the network equipment and the product line
  • the historical health value is calculated; if the rated bandwidth adjustment method is down, the current rated bandwidth corresponding to the product line is adjusted based on the preset reduction ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period.
  • the adjustment method and the adjustment range of the rated bandwidth of each product line on each network device are dynamically determined in real time according to a preset algorithm, and the adjusted rated bandwidth can be more accurately reflected
  • the actual bandwidth load capacity of various network devices in the data transmission network when performing services of different product lines can be effectively applied to the actual planning and scheduling process of transmission network traffic. Therefore, the dynamically adjusted rated bandwidth is used to determine the transmission path. To a certain extent, the transmission quality of the data transmission network is guaranteed.
  • the adjustment of the rated bandwidth is dynamically adjusted by the rated bandwidth adjustment device based on the real-time data of the network device. There is no need to manually specify specific values, and it has strong adaptability, and the adjustment method does not depend on the initial set bandwidth value, even The initial rated bandwidth is inaccurate, and the adjustment method disclosed in the present invention can also converge the rated bandwidth to a stable and reasonable range with a small number of adjustments. Therefore, the above-mentioned process of adjusting the rated bandwidth has strong convergence and robustness Sex.
  • FIG. 1 is a schematic diagram of a scenario architecture of a data transmission network provided by an embodiment of the present invention
  • FIG. 2 is a flowchart of a method for adjusting a rated bandwidth provided by an embodiment of the present invention
  • FIG. 3 is a structural diagram of a device for adjusting a rated bandwidth provided by an embodiment of the present invention.
  • FIG. 4 is a structural diagram of a device for adjusting a rated bandwidth provided by an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of a rated bandwidth adjustment device according to an embodiment of the present invention.
  • An embodiment of the present invention provides a method for adjusting a rated bandwidth.
  • the method can be applied in a data transmission network, and can be specifically implemented by a rated bandwidth adjustment device in the data transmission network.
  • the rated bandwidth adjustment device may be a device for adjusting the rated bandwidth of different product lines on each network device in the data transmission network, which can analyze the bandwidth data of the network device and the health value of each product line based on the analysis As a result, the adjustment processing of the rated bandwidth is realized in units of network equipment + product lines.
  • each network device in the data transmission network can support services of at least one product line at the same time, and different product lines can be used to transmit network traffic of different data types.
  • the collection and storage of the bandwidth data and the health value of the above network device may be implemented by the rated bandwidth adjustment device, or may be separately completed by the data collection device and the data storage device in the data transmission network.
  • the specific scene architecture can be shown in Figure 1.
  • the above-mentioned rated bandwidth adjusting device may include a processor, a memory, and a transceiver.
  • the processor may be used to perform a process of adjusting the rated bandwidth in the following process, and the memory may be used to store data required and data generated during the following process.
  • the transceiver can be used to receive and send relevant data during the processing described below.
  • the above-mentioned processes of adjusting the rated bandwidth, collecting and storing bandwidth data and health values can all be implemented by a distributed system deployed inside the data transmission network. In this embodiment, "the processing of adjusting the rated bandwidth and collecting and storing data are all performed by a single rated bandwidth adjusting device" is used as an example for description, and other situations are similar and will not be repeated.
  • Step 201 For any product line of any network device in the data transmission network, periodically determine the adjustment method of the rated bandwidth of the product line according to the preset reference health value of the product line and the real-time health value in the current cycle.
  • the health value is calculated and generated by the performance index of the network equipment and the business parameters of the product line.
  • the rated bandwidth adjustment device may periodically (the period may take a value of 30 minutes) obtain the real-time health value of all product lines of all network devices in the data transmission network in the current cycle, the health value may be determined by the network device Performance indicators (including CPU usage, IO usage, memory usage, etc.) and business parameters of the product line (including the number of business processes, business response time, etc.) are calculated and generated. Specifically, the health value can be run by each network device.
  • Performance indicators including CPU usage, IO usage, memory usage, etc.
  • business parameters of the product line including the number of business processes, business response time, etc.
  • the process is based on real-time calculation and performance based on performance indicators and business parameters and stored accordingly, and then periodically collected from each network device by the rated bandwidth adjustment device, or the rated bandwidth adjustment device can obtain the performance indicators and products of each network device After the service parameters of the line are temporarily calculated based on the performance indicators and service parameters. Afterwards, for any product line of any network device in the data transmission network, the rated bandwidth adjustment device can determine the product line rating based on the acquired real-time health value in the current period and the preset baseline health value of the product line Bandwidth adjustment method.
  • the health value can be used to determine whether the product line is healthy, and the rated bandwidth can be adjusted accordingly.
  • the processing in step 201 can be as follows: periodically calculate the average value of the real-time health value of the product line in the current period; If the value is less than the preset baseline health value of the product line, the adjustment method of the rated bandwidth of the product line is determined to be upward adjustment, otherwise it is downward adjustment.
  • the rated bandwidth adjustment device may periodically calculate the average value qos avg of the real-time health value qos real of the product line in the current period, that is:
  • t is the cycle length
  • the rated bandwidth adjustment device can determine the size of the average qos avg and the preset baseline health value qos bl of the product line. If qos avg ⁇ qos bl , it means that the product line is in a healthy state and the rated bandwidth of the product line can be increased , That is, the rated bandwidth adjustment method of the product line is up. If qos avg ⁇ qos bl , it means that the product line is in an unhealthy state, and the rated bandwidth of the product line needs to be adjusted down. .
  • step 202 if the adjustment method of the rated bandwidth is upward adjustment, the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated according to the fluctuation of the real-time health value of the product line in the current period.
  • the rated bandwidth adjustment device determines that the adjustment method of the rated bandwidth of a product line of a network device is upward, it can calculate the product line according to the fluctuation of the real-time health value of the product line in the current period The proportion of the corresponding rated bandwidth is increased in real time.
  • the real-time bandwidth data of the network device can be used to determine whether to perform the up adjustment.
  • part of the processing in step 202 can be as follows: If the real-time bandwidth of the network device is in the current period If the value is greater than the median value of the real-time bandwidth of the network equipment during the previous week and greater than the current corresponding rated bandwidth of the product line, then the real-time increase ratio of the rated bandwidth corresponding to the product line is calculated according to the fluctuation of the real-time health value of the product line in the current period Otherwise, the rated bandwidth adjustment mode is set to no adjustment.
  • the rated bandwidth adjustment device can obtain the real-time bandwidth median value of the network device to which the product line belongs within the current period Then determine the median real-time bandwidth Median real-time bandwidth of the network device during the previous week And the current corresponding rated bandwidth of the product line Between the size.
  • the rated bandwidth of the product line can be adjusted upwards, then the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line can be calculated according to the fluctuation of the real-time health value of the product line in the current period, otherwise the rated bandwidth of the product line will not be adjusted.
  • the rated bandwidth adjustment mode of the product line is set to no adjustment.
  • the real-time upward adjustment ratio of the rated bandwidth can be calculated through the function gradient.
  • the processing in step 202 can be as follows: According to the fluctuation of the real-time health value of the product line in the current period, the health value adjustment coefficient corresponding to the product line is established Real-time change function; calculate the latest smooth gradient corresponding to the real-time change function according to the derivative of the real-time change function to the real-time health value of the product line and the preset smoothing coefficient, and the historical smooth gradient corresponding to the real-time change function; according to the product in the historical period The historical upward adjustment ratio and latest smooth gradient of the rated bandwidth corresponding to the line are calculated, and the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated.
  • the rated bandwidth adjustment device can establish the real-time change function f of the health value adjustment coefficient corresponding to the product line after acquiring the fluctuation of the real-time health value of the product line in the current period:
  • t is the cycle length
  • qos real is the real-time health value
  • qos bl is the preset reference health value.
  • the ratio Tn and the latest smooth gradient can be adjusted upwards according to the history of the rated bandwidth corresponding to the product line in the historical period
  • T the current period
  • Tn the historical period
  • the corresponding processing can be as follows: According to the corresponding rating of the product line in the current cycle Bandwidth up-regulation times and preset up-decay attenuation rate, update the current up-stepping step control coefficient corresponding to the product line; according to the historical up-scaling ratio of the rated bandwidth corresponding to the product line in the historical period, the latest smooth gradient and the updated up-step The product of the long control coefficients calculates the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line.
  • the rated bandwidth adjustment device can count the number of times the rated bandwidth of each product line is increased in each cycle, so that when calculating the ratio T of the real-time increase of the rated bandwidth of the product line, the product line within the current cycle Corresponding to the rated bandwidth upward adjustment times iter and the preset upward attenuation rate decay, the current corresponding upward adjustment step control coefficient ⁇ of the product line is updated, namely:
  • Step 203 Adjust the ratio based on the real-time up-scaling ratio and the pre-stored rated bandwidth corresponding to the product line, and adjust the current corresponding rated bandwidth of the product line.
  • the rated bandwidth adjustment device may adjust the ratio based on the real-time upward adjustment ratio and the pre-stored rated bandwidth corresponding to the product line, and correspond to The rated bandwidth is adjusted. Specifically, the current period is T, the real-time up-ratio is ratio T , and the base up-ratio is ratio 0. The product line currently corresponds to the rated bandwidth of Then the adjusted rated bandwidth is
  • Step 204 If the rated bandwidth adjustment mode is downward adjustment, the current rated bandwidth corresponding to the product line is adjusted based on the preset downward adjustment ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period.
  • the rated bandwidth adjustment device determines that the rated bandwidth adjustment method of a product line of a certain network device is downward, it can be based on the preset downward ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period, Adjust the current corresponding rated bandwidth of the product line. Further, you can first calculate the product of the preset reduction ratio of the rated bandwidth and the current corresponding rated bandwidth of the product line, and then adjust the current corresponding rated bandwidth of the product line to the maximum between the product and the median real-time bandwidth of the network device in the current period value. Specifically, the current period is T, the preset downward adjustment ratio is ⁇ , and the current rated bandwidth of the product line is The median value of real-time bandwidth for the current period Then the adjusted rated bandwidth is
  • the method of adjusting the ratio based on the rated bandwidth corresponding to each product line of each network device by the rated bandwidth adjustment device may be adjusted as follows: determine the conventional bandwidth of the network device based on the historical bandwidth data of the network device every preset duration Value, and determine the conventional health value of the product line based on the historical health value of the product line; based on the conventional bandwidth value and the conventional health value, calculate and store the proportion of the adjusted bandwidth corresponding to the product line to increase the ratio.
  • the rated bandwidth adjustment device can obtain the historical bandwidth data of the network device (such as the acquisition) every preset duration (the preset duration can be 1 day) The historical bandwidth data of the past week) and the historical health value of the product line (such as obtaining the historical health value of the past week). Then, the rated bandwidth adjustment device may determine the conventional bandwidth value of the network device according to the acquired historical bandwidth data, and at the same time determine the conventional health value of the product line according to the historical health value of the product line. Furthermore, the rated bandwidth adjustment device can calculate the adjusted bandwidth corresponding to the above product line and adjust the ratio based on the above-mentioned conventional bandwidth value and conventional health value.
  • the above-mentioned method of determining the conventional bandwidth value of the network device may specifically be as follows: acquiring historical bandwidth data of all network devices in the data transmission network every preset time duration; according to the data distribution characteristics of the historical bandwidth data of each network device, Classify all network devices by EM clustering algorithm; for any category, use the preset quantile corresponding to the category to determine the general bandwidth of each network device under the category in all historical bandwidth data of the network devices under the category value.
  • the rated bandwidth adjustment device may obtain historical bandwidth data of all network devices in the data transmission network every preset time period, and then may determine the data distribution characteristics of the historical bandwidth data of each network device, and the data distribution characteristics may include history The maximum and minimum values of bandwidth data, kurtosis, skewness, mean, variance, etc. Afterwards, the rated bandwidth adjustment device can use these data distribution characteristics to classify all network devices by the EM clustering algorithm according to the number of different categories, and evaluate the classification results to select the appropriate number of categories.
  • the appropriateness of the number of categories can be obtained by the technical personnel based on experience and is configured in the rated bandwidth adjustment device.
  • the rated bandwidth adjustment device can summarize the historical bandwidth data of all network devices in the category, and select the bandwidth data pointed to by the preset quantile of the category among all the summarized historical bandwidth data And determine the bandwidth data as the normal bandwidth value of each network device in this category.
  • the above-mentioned way of determining the general health value of the product line may specifically be as follows: acquiring the historical health value of each product line of each network device in the data transmission network every preset duration; according to the historical health of each product line The data distribution characteristics of the values are used to classify all product lines through the EM clustering algorithm; for any category, the preset quantile corresponding to the category is used to determine each historical health value of the product line under the category Health value of each product line.
  • the rated bandwidth adjustment device may obtain the historical health value of each product line of each network device in the data transmission network every preset time period, and then may determine the data distribution characteristics of the historical health value of each product line. Data distribution characteristics may include maximum and minimum values of historical health values, kurtosis, skewness, mean, variance, and so on. Afterwards, the rated bandwidth adjustment device can use these data distribution characteristics to classify all product lines of all network devices through the EM clustering algorithm according to the number of different categories, and evaluate the classification results to select the appropriate number of categories Here, the appropriateness of the number of categories can be summarized by technicians based on experience and configured in the rated bandwidth adjustment device.
  • the rated bandwidth adjustment device can summarize the historical health values of all product lines under the category, and select the health value pointed to by the preset quantile of the category among all the historical health values obtained from the summary , And determine the health value as the normal health value of each product line under the category.
  • the method of adjusting the ratio based on the conventional bandwidth value and the conventional health value may be as follows: based on the conventional bandwidth value and the current rated bandwidth of the product line, calculate the bandwidth adjustment factor of the rated bandwidth corresponding to the product line; Value and the preset baseline health value of the product line, calculate the health value adjustment coefficient of the rated bandwidth corresponding to the product line; based on the bandwidth adjustment coefficient and the health value adjustment coefficient, and the preset upward adjustment limit factor, calculate and store the corresponding value of the product line Adjust the ratio based on the rated bandwidth.
  • the rated bandwidth adjustment device may first calculate the bandwidth adjustment factor of the rated bandwidth corresponding to the product line based on the normal bandwidth value bwi normal and the current corresponding rated bandwidth bwi nominal of the product line:
  • the rated bandwidth adjustment device can calculate the basis of the increase in the ratio of the rated bandwidth corresponding to the product line based on the bandwidth adjustment coefficient stability coe and the health value adjustment coefficient qos quantile , and the preset upward adjustment limit factor C 1 ⁇ (0, 1).
  • ratio 0 C 1 * tanh (qos coe * stability coe * ⁇ ).
  • any product line of any network device in the data transmission network periodically determine the adjustment method of the rated bandwidth of the product line according to the preset reference health value of the product line and the real-time health value in the current cycle, Among them, the health value is calculated and generated according to the performance indicators of the network equipment and the business parameters of the product line; if the rated bandwidth adjustment method is up, the corresponding bandwidth of the product line is calculated according to the fluctuation of the real-time health value of the product line in the current cycle Adjust the ratio in real time; adjust the ratio based on the real-time adjustment ratio and the pre-stored rated bandwidth corresponding to the product line, and adjust the current corresponding rated bandwidth of the product line.
  • the basic adjustment ratio is based on the historical bandwidth data of the network equipment and the product line
  • the historical health value is calculated; if the rated bandwidth adjustment method is down, the current rated bandwidth corresponding to the product line is adjusted based on the preset reduction ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period.
  • the adjustment method and the adjustment range of the rated bandwidth of each product line on each network device are dynamically determined in real time according to a preset algorithm, and the adjusted rated bandwidth can be more accurately reflected
  • the actual bandwidth load capacity of various network devices in the data transmission network when performing services of different product lines can be effectively applied to the actual planning and scheduling process of transmission network traffic. Therefore, the dynamically adjusted rated bandwidth is used to determine the transmission path. To a certain extent, the transmission quality of the data transmission network is guaranteed.
  • the adjustment of the rated bandwidth is dynamically adjusted by the rated bandwidth adjustment device based on the real-time data of the network device. There is no need to manually specify specific values, and it has strong self-adaptability, and the adjustment method does not depend on the initial set bandwidth value. The initial rated bandwidth is inaccurate, and the adjustment method disclosed in the present invention can also converge the rated bandwidth to a stable and reasonable range with a small number of adjustments. Therefore, the above-mentioned process of adjusting the rated bandwidth has strong convergence and robustness Sex.
  • an embodiment of the present invention also provides an apparatus for adjusting the rated bandwidth. As shown in FIG. 3, the apparatus includes:
  • the determination module 301 is configured to periodically determine the rated value of the product line for any product line of any network device in the data transmission network according to the preset reference health value of the product line and the real-time health value in the current period Bandwidth adjustment method;
  • the upward adjustment module 302 is configured to calculate the real-time upward adjustment ratio of the nominal bandwidth corresponding to the product line according to the fluctuation of the real-time health value of the product line during the current period if the nominal bandwidth adjustment mode is upward adjustment, based on Adjust the ratio based on the real-time upward adjustment ratio and the pre-stored rated bandwidth corresponding to the product line, and adjust the current corresponding rated bandwidth of the product line;
  • the downward adjustment module 303 is used to adjust the current rated bandwidth of the product line based on the preset downward adjustment ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period if the rated bandwidth adjustment mode is downward adjustment Make adjustments.
  • the device further includes a calculation module 304, which is used to:
  • an up-scaling ratio based on the rated bandwidth corresponding to the product line is calculated and stored.
  • calculation module 304 is specifically used to:
  • the conventional bandwidth value of each network device under the category is determined in all historical bandwidth data of the network devices under the category.
  • calculation module 304 is specifically used to:
  • all product lines are classified by the EM clustering algorithm
  • the general health value of each product line under the category is determined among all historical health values of the product line under the category.
  • calculation module 304 is specifically used to:
  • a base upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated and stored.
  • the determination module 301 is specifically used to:
  • the adjustment method of the rated bandwidth of the product line is upward adjustment, otherwise it is downward adjustment.
  • the upward adjustment module 302 is specifically used for:
  • the median real-time bandwidth of the network device in the current period is greater than the median real-time bandwidth of the network device in the previous week and greater than the current rated bandwidth corresponding to the product line, then according to the current period In the fluctuation of the real-time health value of the product line, calculate the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line, otherwise, set the rated bandwidth adjustment mode to no adjustment.
  • the upward adjustment module 302 is specifically used for:
  • the upward adjustment module 302 is specifically used for:
  • the real-time upward adjustment ratio of the rated bandwidth corresponding to the product line is calculated.
  • the health value is calculated and generated according to the performance index of the network device and the business parameters of the product line.
  • any product line of any network device in the data transmission network periodically determine the adjustment method of the rated bandwidth of the product line according to the preset reference health value of the product line and the real-time health value in the current cycle, Among them, the health value is calculated and generated according to the performance indicators of the network equipment and the business parameters of the product line; if the rated bandwidth adjustment method is up, the corresponding bandwidth of the product line is calculated according to the fluctuation of the real-time health value of the product line in the current cycle Adjust the ratio in real time; adjust the ratio based on the real-time adjustment ratio and the pre-stored rated bandwidth corresponding to the product line, and adjust the current corresponding rated bandwidth of the product line.
  • the basic adjustment ratio is based on the historical bandwidth data of the network equipment and the The historical health value is calculated; if the rated bandwidth adjustment method is down, the current rated bandwidth corresponding to the product line is adjusted based on the preset reduction ratio of the rated bandwidth and the real-time bandwidth data of the network device in the current period.
  • the adjustment method and the adjustment range of the rated bandwidth of each product line on each network device are dynamically determined in real time according to a preset algorithm, and the adjusted rated bandwidth can be more accurately reflected
  • the actual bandwidth load capacity of each network device in the data transmission network when performing services of different product lines can be effectively applied to the actual planning and scheduling process of the transmission network traffic. Therefore, the dynamically adjusted rated bandwidth is used to determine the transmission path. To a certain extent, the transmission quality of the data transmission network is guaranteed.
  • the adjustment of the rated bandwidth is dynamically adjusted by the rated bandwidth adjustment device based on the real-time data of the network device. There is no need to manually specify specific values, and it has strong self-adaptability, and the adjustment method does not depend on the initial set bandwidth value. The initial rated bandwidth is inaccurate, and the adjustment method disclosed in the present invention can also converge the rated bandwidth to a stable and reasonable range with a small number of adjustments. Therefore, the above-mentioned process of adjusting the rated bandwidth has strong convergence and robustness Sex.
  • the device for adjusting the rated bandwidth provided in the above embodiment only uses the division of the above functional modules as an example for adjusting the rated bandwidth.
  • the above functions can be allocated by different function modules as needed Completed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.
  • the device for adjusting the rated bandwidth provided by the foregoing embodiment and the method embodiment for adjusting the rated bandwidth belong to the same concept. For the specific implementation process, refer to the method embodiment, and details are not described here.
  • FIG. 5 is a schematic structural diagram of a rated bandwidth adjustment device provided by an embodiment of the present invention.
  • the rated bandwidth adjustment device 500 may have a relatively large difference due to different configurations or performances, and may include one or more central processors 522 (eg, one or more processors) and a memory 532, and one or more storage applications
  • the storage medium 530 of 552 or data 555 (for example, one or one mass storage device in Shanghai).
  • the memory 532 and the storage medium 530 may be short-term storage or persistent storage.
  • the program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the rated bandwidth adjusting device 500.
  • the central processor 522 may be configured to communicate with the storage medium 530 and execute a series of instruction operations in the storage medium 530 on the rated bandwidth adjustment device 500.
  • the rated bandwidth adjustment device 500 may also include one or more power supplies 529, one or more wired or wireless network interfaces 550, one or more input and output interfaces 558, one or more keyboards 556, and / or one or more than one Operating system 551, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc.
  • the rated bandwidth adjusting device 500 may include a memory, and one or more programs, where one or more programs are stored in the memory, and are configured to be executed by one or more processors to include the one or more programs. In order to carry out the above instruction to adjust the rated bandwidth.
  • the program may be stored in a computer-readable storage medium.
  • the mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

Landscapes

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

Abstract

本发明公开了一种调整额定带宽的方法和装置,属于数据传输技术领域。所述方法包括:对于数据传输网络中任一网络设备的产品线,周期性根据产品线的预设基准健康值和当前周期内的实时健康值,确定产品线的额定带宽调整方式;如果额定带宽调整方式为上调,则根据当前周期内产品线的实时健康值的波动情况,基于产品线对应的额定带宽的实时上调比例和预先存储的产品线对应的额定带宽的基础上调比例,对产品线当前对应的额定带宽进行调整;如果额定带宽调整方式为下调,则基于额定带宽的预设下调比例和当前周期内网络设备的实时带宽数据,对产品线当前对应的额定带宽进行调整。采用本发明,可以在一定程度上保证数据传输网络的传输质量。

Description

一种调整额定带宽的方法和装置 技术领域
本发明涉及数据传输技术领域,特别涉及一种调整额定带宽的方法和装置。
背景技术
网络设备的额定带宽是指网络设备正常工作状态下,单位时间内允许通过的网络流量的数据总量。网络流量在数据传输网络中传输的过程中,往往需要根据数据传输网络中各网络设备的额定带宽来确定各网络设备的当前带宽负载情况,进而可以确定网络流量的具体传输路径。
在将一批网络设备添加至数据传输网络来提供数据传输服务前,一般可以通过压力测试的方式对网络设备的额定带宽进行评估,具体的,技术人员可以先随机选取多个网络设备,然后在一段时间内检测选取出的每个网络设备所能传输的网络流量的最大数据量,进而可以将计算得到的上述多个网络设备的额定带宽的平均值或众数确定为本批网络设备的额定带宽。
在实现本发明的过程中,发明人发现现有技术至少存在以下问题:
其一,通过抽样测试得到的额定带宽无法准确地反映出每个网络设备各自的额定带宽;其二,网络设备在投入使用之后,往往会同时支持多类产品线(可以按照传输的数据类型分为文本传输类产品线、图片传输类产品线、视频传输类产品线等)的业务,网络设备的额定带宽也会随着多类产品线的业务负载的配比不同而出现变化;其三,随着时间推移,硬件老化或者故障等问题会导致网络设备的额定带宽下降。综上,压力测试所确定出的额定带宽将无法有效地应用于传输网络流量的实际规划和调度过程,采用上述额定带宽来确定传输路径,将会对数据传输网络的传输质量造成较大影响。
发明内容
为了解决现有技术的问题,本发明实施例提供了一种调整额定带宽的方法和装置。所述技术方案如下:
第一方面,提供了一种调整额定带宽的方法,所述方法包括:
对于数据传输网络中任一网络设备的任一产品线,周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式;
如果所述额定带宽调整方式为上调,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例;
基于所述实时上调比例和预先存储的所述产品线对应的额定带宽的基础上调比例,对所述产品线当前对应的额定带宽进行调整;
如果所述额定带宽调整方式为下调,则基于额定带宽的预设下调比例和所述当前周期内所述网络设备的实时带宽数据,对所述产品线当前对应的额定带宽进行调整。
可选的,所述方法还包括:
每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,并根据所述产品线的历史健康值确定所述产品线的常规健康值;
基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额定带宽的基础上调比例。
可选的,所述每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,包括:
每隔预设时长获取所述数据传输网络中所有网络设备的历史带宽数据;
根据每个所述网络设备的历史带宽数据的数据分布特征,通过EM聚类算法对所有所述网络设备进行分类;
对于任一类别,利用所述类别对应的预设分位数,在所述类别下的网络设备的所有历史带宽数据中确定所述类别下的每个网络设备的常规带宽值。
可选的,所述每隔预设时长根据所述产品线的历史健康值确定所述产品线的常规健康值,包括:
每隔预设时长获取所述数据传输网络中每个网络设备的每个产品线的历史健康值;
根据每个所述产品线的历史健康值的数据分布特征,通过EM聚类算法对所有产品线进行分类;
对于任一类别,利用所述类别对应的预设分位数,在所述类别下的产品线 的所有历史健康值中确定所述类别下的每个产品线的常规健康值。
可选的,所述基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额定带宽的基础上调比例,包括:
基于所述常规带宽值和所述产品线当前对应的额定带宽,计算所述产品线对应的额定带宽的带宽调整系数;
基于所述常规健康值和所述产品线的预设基准健康值,计算所述产品线对应的额定带宽的健康值调整系数;
基于所述带宽调整系数和所述健康值调整系数,以及预设的上调比例限制系数,计算并存储所述产品线对应的额定带宽的基础上调比例。
可选的,所述周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式,包括:
周期性计算当前周期内所述产品线的实时健康值的平均值;
如果所述平均值小于所述产品线的预设基准健康值,则确定所述产品线的额定带宽调整方式为上调,否则为下调。
可选的,所述根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,包括:
如果所述当前周期内所述网络设备的实时带宽中值大于所述上一周期内所述网络设备的实时带宽中值,且大于所述产品线当前对应的额定带宽,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,否则将所述额定带宽调整方式设置为不调整。
可选的,所述根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,包括:
根据所述当前周期内所述产品线的实时健康值的波动情况,建立所述产品线对应的健康值调整系数的实时变化函数;
根据所述实时变化函数对所述产品线的实时健康值的导数和预设的平滑系数,以及所述实时变化函数对应的历史平滑梯度,计算所述实时变化函数对应的最新平滑梯度;
根据历史周期内所述产品线对应的额定带宽的历史上调比例和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例。
可选的,所述根据历史周期内所述产品线对应的额定带宽的历史上调比例 和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例,包括:
根据当前周期内所述产品线对应的额定带宽上调次数和预设的上调衰减率,对所述产品线当前对应的上调步长控制系数进行更新;
根据历史周期内所述产品线对应的额定带宽的历史上调比例,以及所述最新平滑梯度与更新后的上调步长控制系数的乘积,计算所述产品线对应的额定带宽的实时上调比例。
可选的,所述健康值根据网络设备的性能指标和产品线的业务参数计算生成。
第二方面,提供了一种调整额定带宽的装置,所述装置包括:
确定模块,用于对于数据传输网络中任一网络设备的任一产品线,周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式;
上调模块,用于如果所述额定带宽调整方式为上调,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,基于所述实时上调比例和预先存储的所述产品线对应的额定带宽的基础上调比例,对所述产品线当前对应的额定带宽进行调整;
下调模块,用于如果所述额定带宽调整方式为下调,则基于额定带宽的预设下调比例和所述当前周期内所述网络设备的实时带宽数据,对所述产品线当前对应的额定带宽进行调整。
可选的,所述装置还包括计算模块,用于:
每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,并根据所述产品线的历史健康值确定所述产品线的常规健康值;
基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额定带宽的基础上调比例。
可选的,所述计算模块,具体用于:
每隔预设时长获取所述数据传输网络中所有网络设备的历史带宽数据;
根据每个所述网络设备的历史带宽数据的数据分布特征,通过EM聚类算法对所有所述网络设备进行分类;
对于任一类别,利用所述类别对应的预设分位数,在所述类别下的网络设 备的所有历史带宽数据中确定所述类别下的每个网络设备的常规带宽值。
可选的,所述计算模块,具体用于:
每隔预设时长获取所述数据传输网络中每个网络设备的每个产品线的历史健康值;
根据每个所述产品线的历史健康值的数据分布特征,通过EM聚类算法对所有产品线进行分类;
对于任一类别,利用所述类别对应的预设分位数,在所述类别下的产品线的所有历史健康值中确定所述类别下的每个产品线的常规健康值。
可选的,所述计算模块,具体用于:
基于所述常规带宽值和所述产品线当前对应的额定带宽,计算所述产品线对应的额定带宽的带宽调整系数;
基于所述常规健康值和所述产品线的预设基准健康值,计算所述产品线对应的额定带宽的健康值调整系数;
基于所述带宽调整系数和所述健康值调整系数,以及预设的上调比例限制系数,计算并存储所述产品线对应的额定带宽的基础上调比例。
可选的,所述确定模块,具体用于:
周期性计算当前周期内所述产品线的实时健康值的平均值;
如果所述平均值小于所述产品线的预设基准健康值,则确定所述产品线的额定带宽调整方式为上调,否则为下调。
可选的,所述上调模块,具体用于:
如果所述当前周期内所述网络设备的实时带宽中值大于所述上一周期内所述网络设备的实时带宽中值,且大于所述产品线当前对应的额定带宽,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,否则将所述额定带宽调整方式设置为不调整。
可选的,所述上调模块,具体用于:
根据所述当前周期内所述产品线的实时健康值的波动情况,建立所述产品线对应的健康值调整系数的实时变化函数;
根据所述实时变化函数对所述产品线的实时健康值的导数和预设的平滑系数,以及所述实时变化函数对应的历史平滑梯度,计算所述实时变化函数对应的最新平滑梯度;
根据历史周期内所述产品线对应的额定带宽的历史上调比例和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例。
可选的,所述上调模块,具体用于:
根据当前周期内所述产品线对应的额定带宽上调次数和预设的上调衰减率,对所述产品线当前对应的上调步长控制系数进行更新;
根据历史周期内所述产品线对应的额定带宽的历史上调比例,以及所述最新平滑梯度与更新后的上调步长控制系数的乘积,计算所述产品线对应的额定带宽的实时上调比例。
可选的,所述健康值根据网络设备的性能指标和产品线的业务参数计算生成。
第三方面,提供了一种额定带宽调整设备,所述额定带宽调整设备包括处理器和存储器,所述存储器中存储有至少一条指令、至少一段程序、代码集或指令集,所述至少一条指令、所述至少一段程序、所述代码集或指令集由所述处理器加载并执行以实现如第一方面所述的调整额定带宽的方法。
第四方面,提供了一种计算机可读存储介质,所述存储介质中存储有至少一条指令、至少一段程序、代码集或指令集,所述至少一条指令、所述至少一段程序、所述代码集或指令集由处理器加载并执行以实现如第一方面所述的调整额定带宽的方法。
本发明实施例提供的技术方案带来的有益效果是:
本发明实施例中,对于数据传输网络中任一网络设备的任一产品线,周期性根据产品线的预设基准健康值和当前周期内的实时健康值,确定产品线的额定带宽调整方式,其中,健康值根据网络设备的性能指标和产品线的业务参数计算生成;如果额定带宽调整方式为上调,则根据当前周期内产品线的实时健康值的波动情况,计算产品线对应的额定带宽的实时上调比例;基于实时上调比例和预先存储的产品线对应的额定带宽的基础上调比例,对产品线当前对应的额定带宽进行调整,其中,基础上调比例根据网络设备的历史带宽数据和产品线的历史健康值计算得到;如果额定带宽调整方式为下调,则基于额定带宽 的预设下调比例和当前周期内网络设备的实时带宽数据,对产品线当前对应的额定带宽进行调整。这样,通过网络设备的带宽数据和产品线的健康值,根据预设算法实时动态地确定各网络设备上各产品线的额定带宽的调整方式以及调整幅度,调整后的额定带宽可以较为准确地反映数据传输网络中各个网络设备执行不同产品线的业务时实际的带宽负载能力,可以有效地应用于传输网络流量的实际规划和调度过程,故而,采用动态调整后的额定带宽来确定传输路径,可以在一定程度上保证数据传输网络的传输质量。
此外,额定带宽的调整由额定带宽调整设备依托于网络设备的实时数据动态调节,无需人为指定具体数值,具备较强的自适应性,并且该调整方法不依赖初始设置的额定带宽值,即使设置初始额定带宽不准确,通过本发明公开的调整方法也可以在较少的调整次数下,将额定带宽收敛到一个稳定合理的范围,因此上述调整额定带宽的处理具备较强的收敛性和鲁棒性。
附图说明
为了更清楚地说明本发明实施例中的技术方案,下面将对实施例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发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的一种数据传输网络的场景架构示意图;
图2是本发明实施例提供的一种调整额定带宽的方法流程图;
图3是本发明实施例提供的一种调整额定带宽的装置结构示图;
图4是本发明实施例提供的一种调整额定带宽的装置结构示图;
图5是本发明实施例提供的一种额定带宽调整设备的结构示意图。
具体实施方式
为使本发明的目的、技术方案和优点更加清楚,下面将结合附图对本发明实施方式作进一步地详细描述。
本发明实施例提供了一种调整额定带宽的方法,该方法的可以运用在数据传输网络中,并具体可以由数据传输网络中的额定带宽调整设备来实现。其中,额定带宽调整设备可以是数据传输网络中用于调整各个网络设备上不同产品线 的额定带宽的设备,其可以通过对网络设备的带宽数据和各产品线的健康值进行分析,并基于分析结果以网络设备+产品线为单位实现额定带宽的调整处理。此处,数据传输网络中的每个网络设备可以在同一时刻支持至少一个产品线的业务,不同产品线可以用于传输不同数据类型的网络流量。上述网络设备的带宽数据和健康值的采集和存储可以由额定带宽调整设备来实现,也可以由数据传输网络中的数据采集设备和数据存储设备来分别完成。具体的场景架构可以如图1所示。上述额定带宽调整设备可以包括处理器、存储器、收发器,处理器可以用于进行下述流程中的调整额定带宽的处理,存储器可以用于存储下述处理过程中需要的数据以及产生的数据,收发器可以用于接收和发送下述处理过程中的相关数据。上述调整额定带宽、采集和存储带宽数据和健康值的处理均可以由部署在数据传输网络内部的分布式***来实现。本实施例中以“调整额定带宽、采集和存储数据的处理全部由单一的额定带宽调整设备来执行”为例进行说明,其它情况与之类似,不再赘述。
下面将结合具体实施方式,对图2所示的处理流程进行详细的说明,内容可以如下:
步骤201,对于数据传输网络中任一网络设备的任一产品线,周期性根据产品线的预设基准健康值和当前周期内的实时健康值,确定产品线的额定带宽调整方式。
其中,健康值由网络设备的性能指标和产品线的业务参数计算生成。
在实施中,额定带宽调整设备可以周期性(周期可以取值为30分钟)获取数据传输网络中所有网络设备的所有产品线在当前周期内的实时健康值,该健康值可以是由网络设备的性能指标(包括CPU使用率、IO使用率、内存使用率等)和产品线的业务参数(包括业务进程数目、业务响应时长等)计算生成,具体的,健康值可以由各个网络设备在运行的过程中基于性能指标和业务参数实时计算生成并相应存储,然后由额定带宽调整设备周期性从各个网络设备处采集得到,也可以由额定带宽调整设备在获取到各个网络设备的性能指标和各产品线的业务参数之后,临时基于性能指标和业务参数计算得到。之后,对于数据传输网络中的任一网络设备的任一产品线,额定带宽调整设备可以根据获取到的当前周期内的实时健康值和该产品线的预设基准健康值,确定产品线的额定带宽调整方式。
可选的,可以通过健康值来判断产品线是否健康,并依此调整额定带宽,相应的,步骤201的处理可以如下:周期性计算当前周期内产品线的实时健康值的平均值;如果平均值小于产品线的预设基准健康值,则确定产品线的额定带宽调整方式为上调,否则为下调。
在实施中,对于数据传输网络中的任一网络设备的任一产品线,额定带宽调整设备可以周期性计算当前周期内该产品线的实时健康值qos real的平均值qos avg,即:
Figure PCTCN2018119597-appb-000001
t为周期长度。
之后,额定带宽调整设备可以判断该平均值qos avg与产品线的预设基准健康值qos bl的大小,如果qos avg<qos bl,则表示产品线处于健康状态,可以上调该产品线的额定带宽,即确定该产品线的额定带宽调整方式为上调,如果qos avg≥qos bl,则表示产品线处于不健康状态,需要下调该产品线的额定带宽,即确定该产品线的额定带宽调整方式为下调。
步骤202,如果额定带宽调整方式为上调,则根据当前周期内产品线的实时健康值的波动情况,计算产品线对应的额定带宽的实时上调比例。
在实施中,额定带宽调整设备如果确定出某一网络设备的某一产品线的额定带宽调整方式为上调,则可以根据当前周期内该产品线的实时健康值的波动情况,计算出该产品线对应的额定带宽的实时上调比例。
可选的,在确定额定带宽调整方式为上调后,可以进一步通过网络设备的实时带宽数据来判断是否进行上调,相应的,步骤202的部分处理可以如下:如果当前周期内网络设备的实时带宽中值大于上一周期内网络设备的实时带宽中值,且大于产品线当前对应的额定带宽,则根据当前周期内产品线的实时健康值的波动情况,计算产品线对应的额定带宽的实时上调比例,否则将额定带宽调整方式设置为不调整。
在实施中,如果确定出产品线对应的额定带宽调整方式为上调,额定带宽调整设备可以获取当前周期内该产品线所属网络设备的实时带宽中值
Figure PCTCN2018119597-appb-000002
然后判断该实时带宽中值
Figure PCTCN2018119597-appb-000003
与上一周期内该网络设备的实时带宽中值
Figure PCTCN2018119597-appb-000004
以及该产品线当前对应的额定带宽
Figure PCTCN2018119597-appb-000005
之间的大小。如果当前周期内网络设备的实时带宽中值
Figure PCTCN2018119597-appb-000006
大于上一周期内网络设备的实时带宽中值
Figure PCTCN2018119597-appb-000007
且大于产品线当前对应的额定带宽
Figure PCTCN2018119597-appb-000008
Figure PCTCN2018119597-appb-000009
则认为可以上调该产品线的额定带宽,那么可以根据当前周期内产品线的实时健康值的波动情况,计算产品线对应的额定带宽的实时上调比例,否则不调整该产品线的额定带宽,即将产品线的额定带宽调整方式设置为不调整。
可选的,可以通过函数梯度来计算额定带宽的实时上调比例,相应的,步骤202的处理可以如下:根据当前周期内产品线的实时健康值的波动情况,建立产品线对应的健康值调整系数的实时变化函数;根据实时变化函数对产品线的实时健康值的导数和预设的平滑系数,以及实时变化函数对应的历史平滑梯度,计算实时变化函数对应的最新平滑梯度;根据历史周期内产品线对应的额定带宽的历史上调比例和最新平滑梯度,计算产品线对应的额定带宽的实时上调比例。
在实施中,额定带宽调整设备在获取到当前周期内产品线的实时健康值的波动情况后,可以依此建立该产品线对应的健康值调整系数的实时变化函数f:
Figure PCTCN2018119597-appb-000010
其中,t为周期长度,qos real为实时健康值,qos bl为预设基准健康值。之后,可以计算该实时变化函数f对产品线的实时健康值的导数d qos
Figure PCTCN2018119597-appb-000011
再根据该导数d qos和预设的平滑系数β,以及实时变化函数f对应的历史平滑梯度
Figure PCTCN2018119597-appb-000012
计算该实时变化函数对应的最新平滑梯度
Figure PCTCN2018119597-appb-000013
Figure PCTCN2018119597-appb-000014
这样,可以根据历史周期内产品线对应的额定带宽的历史上调比例ratio T-n和该最新平滑梯度
Figure PCTCN2018119597-appb-000015
计算产品线对应的额定带宽的实时上调比例ratio T,其中,T为当前周期,T-n为历史周期,具体的,当
Figure PCTCN2018119597-appb-000016
时,
Figure PCTCN2018119597-appb-000017
Figure PCTCN2018119597-appb-000018
时,ratio T=ratio T-n
可选的,在计算产品线的额定带宽的实时上调比例时,需要根据上调次数对额定带宽的调整步长进行一定程度的衰减处理,相应的处理可以如下:根据当前周期内产品线对应的额定带宽上调次数和预设的上调衰减率,对产品线当前对应的上调步长控制系数进行更新;根据历史周期内产品线对应的额定带宽的历史上调比例,以及最新平滑梯度与更新后的上调步长控制系数的乘积,计算产品线对应的额定带宽的实时上调比例。
在实施中,额定带宽调整设备可以在统计每个周期内各产品线的额定带宽 上调次数iter,这样,在计算产品线的额定带宽的实时上调比例ratio T时,可以先根据当前周期内产品线对应的额定带宽上调次数iter和预设的上调衰减率decay,对产品线当前对应的上调步长控制系数α进行更新,即:
Figure PCTCN2018119597-appb-000019
之后,额定带宽调整设备可以根据历史周期内产品线对应的额定带宽的历史上调比例ratio T-n,以及最新平滑梯度与更新后的上调步长控制系数的乘积
Figure PCTCN2018119597-appb-000020
计算产品线对应的额定带宽的实时上调比例ratio T,具体的,当
Figure PCTCN2018119597-appb-000021
时,
Figure PCTCN2018119597-appb-000022
Figure PCTCN2018119597-appb-000023
时,ratio T=ratio T-n
步骤203,基于实时上调比例和预先存储的产品线对应的额定带宽的基础上调比例,对产品线当前对应的额定带宽进行调整。
在实施中,额定带宽调整设备在计算得到产品线对应的额定带宽的实时上调比例后,可以基于该实时上调比例和预先存储的该产品线对应的额定带宽的基础上调比例,对产品线当前对应的额定带宽进行调整。具体的,当前周期为T,实时上调比例为ratio T,基础上调比例为ratio 0,产品线当前对应的额定带宽为
Figure PCTCN2018119597-appb-000024
则调整后的额定带宽为
Figure PCTCN2018119597-appb-000025
步骤204,如果额定带宽调整方式为下调,则基于额定带宽的预设下调比例和当前周期内网络设备的实时带宽数据,对产品线当前对应的额定带宽进行调整。
在实施中,额定带宽调整设备如果确定出某一网络设备的某一产品线的额定带宽调整方式为下调,则可以根据额定带宽的预设下调比例和当前周期内该网络设备的实时带宽数据,对该产品线当前对应的额定带宽进行调整。进一步的,可以先计算额定带宽的预设下调比例和产品线当前对应的额定带宽的乘积,然后将产品线当前对应的额定带宽调整为乘积与当前周期内网络设备的实时带宽中值间的最大值。具体的,当前周期为T,预设下调比例为ω,产品线当前的额定带宽为
Figure PCTCN2018119597-appb-000026
当前周期的实时带宽中值为
Figure PCTCN2018119597-appb-000027
则调整后的额定带宽为
Figure PCTCN2018119597-appb-000028
可选的,额定带宽调整设备确定每个网络设备的每个产品线对应的额定带宽的基础上调比例的方式具体可以如下:每隔预设时长根据网络设备的历史带宽数据确定网络设备的常规带宽值,并根据产品线的历史健康值确定产品线的常规健康值;基于常规带宽值和常规健康值,计算并存储产品线对应的额定带 宽的基础上调比例。
在实施中,对于数据传输网络中的任一网络设备的任一产品线,额定带宽调整设备可以每隔预设时长(预设时长可以为1天)获取该网络设备的历史带宽数据(如获取过去1周的历史带宽数据)和该产品线的历史健康值(如获取过去1周的历史健康值)。然后,额定带宽调整设备可以根据获取到的历史带宽数据确定网络设备的常规带宽值,并同时根据产品线的历史健康值确定产品线的常规健康值。进而,额定带宽调整设备可以基于上述常规带宽值和常规健康值,计算上述产品线对应的额定带宽的基础上调比例。
可选的,上述确定网络设备的常规带宽值的方式具体可以如下:每隔预设时长获取数据传输网络中所有网络设备的历史带宽数据;根据每个网络设备的历史带宽数据的数据分布特征,通过EM聚类算法对所有网络设备进行分类;对于任一类别,利用类别对应的预设分位数,在类别下的网络设备的所有历史带宽数据中确定类别下的每个网络设备的常规带宽值。
在实施中,额定带宽调整设备可以每隔预设时长获取数据传输网络中所有网络设备的历史带宽数据,然后可以确定每个网络设备的历史带宽数据的数据分布特征,该数据分布特征可以包括历史带宽数据的最大最小值、峰度、偏度、均值、方差等。之后,额定带宽调整设备可以使用这些数据分布特征,按照不同的类别个数,通过EM聚类算法对所有网络设备进行分类,并对分类结果进行评估从而选择出合适的类别个数,此处,类别个数合适与否可以由技术人员基于经验总结得到,并配置在额定带宽调整设备中。接下来,对于任一类别,额定带宽调整设备可以汇总该类别下的所有网络设备的历史带宽数据,并在汇总得到的所有历史带宽数据中选择该类别的预设分位数所指向的带宽数据,并将该带宽数据确定为该类别下的每个网络设备的常规带宽值。
可选的,上述确定产品线的常规健康值的方式具体可以如下:每隔预设时长获取数据传输网络中每个网络设备的每个产品线的历史健康值;根据每个产品线的历史健康值的数据分布特征,通过EM聚类算法对所有产品线进行分类;对于任一类别,利用类别对应的预设分位数,在类别下的产品线的所有历史健康值中确定类别下的每个产品线的常规健康值。
在实施中,额定带宽调整设备可以每隔预设时长获取数据传输网络中每个网络设备的每个产品线的历史健康值,然后可以确定每个产品线的历史健康值 的数据分布特征,该数据分布特征可以包括历史健康值的最大最小值、峰度、偏度、均值、方差等。之后,额定带宽调整设备可以使用这些数据分布特征,按照不同的类别个数,通过EM聚类算法对所有网络设备的所有产品线进行分类,并对分类结果进行评估从而选择出合适的类别个数,此处,类别个数合适与否可以由技术人员基于经验总结得到,并配置在额定带宽调整设备中。接下来,对于任一类别,额定带宽调整设备可以汇总该类别下的所有产品线的历史健康值,并在汇总得到的所有历史健康值中选择该类别的预设分位数所指向的健康值,并将该健康值确定为该类别下的每个产品线的常规健康值。
可选的,基于常规带宽值和常规健康值计算基础上调比例的方式具体可以如下:基于常规带宽值和产品线当前对应的额定带宽,计算产品线对应的额定带宽的带宽调整系数;基于常规健康值和产品线的预设基准健康值,计算产品线对应的额定带宽的健康值调整系数;基于带宽调整系数和健康值调整系数,以及预设的上调比例限制系数,计算并存储产品线对应的额定带宽的基础上调比例。
在实施中,额定带宽调整设备可以先基于常规带宽值bwi normal和产品线当前对应的额定带宽bwi nominal,计算产品线对应的额定带宽的带宽调整系数:
Figure PCTCN2018119597-appb-000029
再基于常规健康值qos quantile和产品线的预设基准健康值qos bl,计算产品线对应的额定带宽的健康值调整系数qos coe,具体的,当qos quantile<qos bl时,
Figure PCTCN2018119597-appb-000030
其中∈为归一化因子,当qos quantile≥qos bl时,qos coe=0。之后,额定带宽调整设备可以基于带宽调整系数stability coe和健康值调整系数qos quantile,以及预设的上调比例限制系数C 1∈(0,1],计算产品线对应的额定带宽的基础上调比例:ratio 0=C 1*tanh(qos coe*stability coe*∈)。
本发明实施例中,对于数据传输网络中任一网络设备的任一产品线,周期性根据产品线的预设基准健康值和当前周期内的实时健康值,确定产品线的额定带宽调整方式,其中,健康值根据网络设备的性能指标和产品线的业务参数计算生成;如果额定带宽调整方式为上调,则根据当前周期内产品线的实时健康值的波动情况,计算产品线对应的额定带宽的实时上调比例;基于实时上调比例和预先存储的产品线对应的额定带宽的基础上调比例,对产品线当前对应的额定带宽进行调整,其中,基础上调比例根据网络设备的历史带宽数据和产品线的历史健康值计算得到;如果额定带宽调整方式为下调,则基于额定带宽 的预设下调比例和当前周期内网络设备的实时带宽数据,对产品线当前对应的额定带宽进行调整。这样,通过网络设备的带宽数据和产品线的健康值,根据预设算法实时动态地确定各网络设备上各产品线的额定带宽的调整方式以及调整幅度,调整后的额定带宽可以较为准确地反映数据传输网络中各个网络设备执行不同产品线的业务时实际的带宽负载能力,可以有效地应用于传输网络流量的实际规划和调度过程,故而,采用动态调整后的额定带宽来确定传输路径,可以在一定程度上保证数据传输网络的传输质量。
此外,额定带宽的调整由额定带宽调整设备依托于网络设备的实时数据动态调节,无需人为指定具体数值,具备较强的自适应性,并且该调整方法不依赖初始设置的额定带宽值,即使设置初始额定带宽不准确,通过本发明公开的调整方法也可以在较少的调整次数下,将额定带宽收敛到一个稳定合理的范围,因此上述调整额定带宽的处理具备较强的收敛性和鲁棒性。
基于相同的技术构思,本发明实施例还提供了一种调整额定带宽的装置,如图3所示,所述装置包括:
确定模块301,用于对于数据传输网络中任一网络设备的任一产品线,周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式;
上调模块302,用于如果所述额定带宽调整方式为上调,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,基于所述实时上调比例和预先存储的所述产品线对应的额定带宽的基础上调比例,对所述产品线当前对应的额定带宽进行调整;
下调模块303,用于如果所述额定带宽调整方式为下调,则基于额定带宽的预设下调比例和所述当前周期内所述网络设备的实时带宽数据,对所述产品线当前对应的额定带宽进行调整。
可选的,如图4所示,所述装置还包括计算模块304,用于:
每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,并根据所述产品线的历史健康值确定所述产品线的常规健康值;
基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额 定带宽的基础上调比例。
可选的,所述计算模块304,具体用于:
每隔预设时长获取所述数据传输网络中所有网络设备的历史带宽数据;
根据每个所述网络设备的历史带宽数据的数据分布特征,通过EM聚类算法对所有所述网络设备进行分类;
对于任一类别,利用所述类别对应的预设分位数,在所述类别下的网络设备的所有历史带宽数据中确定所述类别下的每个网络设备的常规带宽值。
可选的,所述计算模块304,具体用于:
每隔预设时长获取所述数据传输网络中每个网络设备的每个产品线的历史健康值;
根据每个所述产品线的历史健康值的数据分布特征,通过EM聚类算法对所有产品线进行分类;
对于任一类别,利用所述类别对应的预设分位数,在所述类别下的产品线的所有历史健康值中确定所述类别下的每个产品线的常规健康值。
可选的,所述计算模块304,具体用于:
基于所述常规带宽值和所述产品线当前对应的额定带宽,计算所述产品线对应的额定带宽的带宽调整系数;
基于所述常规健康值和所述产品线的预设基准健康值,计算所述产品线对应的额定带宽的健康值调整系数;
基于所述带宽调整系数和所述健康值调整系数,以及预设的上调比例限制系数,计算并存储所述产品线对应的额定带宽的基础上调比例。
可选的,所述确定模块301,具体用于:
周期性计算当前周期内所述产品线的实时健康值的平均值;
如果所述平均值小于所述产品线的预设基准健康值,则确定所述产品线的额定带宽调整方式为上调,否则为下调。
可选的,所述上调模块302,具体用于:
如果所述当前周期内所述网络设备的实时带宽中值大于所述上一周期内所述网络设备的实时带宽中值,且大于所述产品线当前对应的额定带宽,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,否则将所述额定带宽调整方式设置为不调整。
可选的,所述上调模块302,具体用于:
根据所述当前周期内所述产品线的实时健康值的波动情况,建立所述产品线对应的健康值调整系数的实时变化函数;
根据所述实时变化函数对所述产品线的实时健康值的导数和预设的平滑系数,以及所述实时变化函数对应的历史平滑梯度,计算所述实时变化函数对应的最新平滑梯度;
根据历史周期内所述产品线对应的额定带宽的历史上调比例和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例。
可选的,所述上调模块302,具体用于:
根据当前周期内所述产品线对应的额定带宽上调次数和预设的上调衰减率,对所述产品线当前对应的上调步长控制系数进行更新;
根据历史周期内所述产品线对应的额定带宽的历史上调比例,以及所述最新平滑梯度与更新后的上调步长控制系数的乘积,计算所述产品线对应的额定带宽的实时上调比例。
可选的,所述健康值根据网络设备的性能指标和产品线的业务参数计算生成。
本发明实施例中,对于数据传输网络中任一网络设备的任一产品线,周期性根据产品线的预设基准健康值和当前周期内的实时健康值,确定产品线的额定带宽调整方式,其中,健康值根据网络设备的性能指标和产品线的业务参数计算生成;如果额定带宽调整方式为上调,则根据当前周期内产品线的实时健康值的波动情况,计算产品线对应的额定带宽的实时上调比例;基于实时上调比例和预先存储的产品线对应的额定带宽的基础上调比例,对产品线当前对应的额定带宽进行调整,其中,基础上调比例根据网络设备的历史带宽数据和产品线的历史健康值计算得到;如果额定带宽调整方式为下调,则基于额定带宽的预设下调比例和当前周期内网络设备的实时带宽数据,对产品线当前对应的额定带宽进行调整。这样,通过网络设备的带宽数据和产品线的健康值,根据预设算法实时动态地确定各网络设备上各产品线的额定带宽的调整方式以及调整幅度,调整后的额定带宽可以较为准确地反映数据传输网络中各个网络设备执行不同产品线的业务时实际的带宽负载能力,可以有效地应用于传输网络流量的实际规划和调度过程,故而,采用动态调整后的额定带宽来确定传输路径, 可以在一定程度上保证数据传输网络的传输质量。
此外,额定带宽的调整由额定带宽调整设备依托于网络设备的实时数据动态调节,无需人为指定具体数值,具备较强的自适应性,并且该调整方法不依赖初始设置的额定带宽值,即使设置初始额定带宽不准确,通过本发明公开的调整方法也可以在较少的调整次数下,将额定带宽收敛到一个稳定合理的范围,因此上述调整额定带宽的处理具备较强的收敛性和鲁棒性。
需要说明的是:上述实施例提供的调整额定带宽的装置在调整额定带宽时,仅以上述各功能模块的划分进行举例说明,实际应用中,可以根据需要而将上述功能分配由不同的功能模块完成,即将装置的内部结构划分成不同的功能模块,以完成以上描述的全部或者部分功能。另外,上述实施例提供的调整额定带宽的装置与调整额定带宽的方法实施例属于同一构思,其具体实现过程详见方法实施例,这里不再赘述。
图5是本发明实施例提供的额定带宽调整设备的结构示意图。该额定带宽调整设备500可因配置或性能不同而产生比较大的差异,可以包括一个或一个以***处理器522(例如,一个或一个以上处理器)和存储器532,一个或一个以上存储应用程序552或数据555的存储介质530(例如一个或一个以上海量存储设备)。其中,存储器532和存储介质530可以是短暂存储或持久存储。存储在存储介质530的程序可以包括一个或一个以上模块(图示没标出),每个模块可以包括对额定带宽调整设备500中的一系列指令操作。更进一步地,中央处理器522可以设置为与存储介质530通信,在额定带宽调整设备500上执行存储介质530中的一系列指令操作。
额定带宽调整设备500还可以包括一个或一个以上电源529,一个或一个以上有线或无线网络接口550,一个或一个以上输入输出接口558,一个或一个以上键盘556,和/或,一个或一个以上操作***551,例如Windows Server,Mac OS X,Unix,Linux,FreeBSD等等。
额定带宽调整设备500可以包括有存储器,以及一个或者一个以上的程序,其中一个或者一个以上程序存储于存储器中,且经配置以由一个或者一个以上处理器执行所述一个或者一个以上程序包含用于进行上述调整额定带宽的指 令。
本领域普通技术人员可以理解实现上述实施例的全部或部分步骤可以通过硬件来完成,也可以通过程序来指令相关的硬件完成,所述的程序可以存储于一种计算机可读存储介质中,上述提到的存储介质可以是只读存储器,磁盘或光盘等。
以上所述仅为本发明的较佳实施例,并不用以限制本发明,凡在本发明的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。

Claims (22)

  1. 一种调整额定带宽的方法,其特征在于,所述方法包括:
    对于数据传输网络中任一网络设备的任一产品线,周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式;
    如果所述额定带宽调整方式为上调,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例;
    基于所述实时上调比例和预先存储的所述产品线对应的额定带宽的基础上调比例,对所述产品线当前对应的额定带宽进行调整;
    如果所述额定带宽调整方式为下调,则基于额定带宽的预设下调比例和所述当前周期内所述网络设备的实时带宽数据,对所述产品线当前对应的额定带宽进行调整。
  2. 根据权利要求1所述的方法,其特征在于,所述方法还包括:
    每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,并根据所述产品线的历史健康值确定所述产品线的常规健康值;
    基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额定带宽的基础上调比例。
  3. 根据权利要求2所述的方法,其特征在于,所述每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,包括:
    每隔预设时长获取所述数据传输网络中所有网络设备的历史带宽数据;
    根据每个所述网络设备的历史带宽数据的数据分布特征,通过EM聚类算法对所有所述网络设备进行分类;
    对于任一类别,利用所述类别对应的预设分位数,在所述类别下的网络设备的所有历史带宽数据中确定所述类别下的每个网络设备的常规带宽值。
  4. 根据权利要求2所述的方法,其特征在于,所述每隔预设时长根据所述产品线的历史健康值确定所述产品线的常规健康值,包括:
    每隔预设时长获取所述数据传输网络中每个网络设备的每个产品线的历史健康值;
    根据每个所述产品线的历史健康值的数据分布特征,通过EM聚类算法对所有产品线进行分类;
    对于任一类别,利用所述类别对应的预设分位数,在所述类别下的产品线的所有历史健康值中确定所述类别下的每个产品线的常规健康值。
  5. 根据权利要求2所述的方法,其特征在于,所述基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额定带宽的基础上调比例,包括:
    基于所述常规带宽值和所述产品线当前对应的额定带宽,计算所述产品线对应的额定带宽的带宽调整系数;
    基于所述常规健康值和所述产品线的预设基准健康值,计算所述产品线对应的额定带宽的健康值调整系数;
    基于所述带宽调整系数和所述健康值调整系数,以及预设的上调比例限制系数,计算并存储所述产品线对应的额定带宽的基础上调比例。
  6. 根据权利要求1所述的方法,其特征在于,所述周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式,包括:
    周期性计算当前周期内所述产品线的实时健康值的平均值;
    如果所述平均值小于所述产品线的预设基准健康值,则确定所述产品线的额定带宽调整方式为上调,否则为下调。
  7. 根据权利要求1所述的方法,其特征在于,所述根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,包括:
    如果所述当前周期内所述网络设备的实时带宽中值大于所述上一周期内所述网络设备的实时带宽中值,且大于所述产品线当前对应的额定带宽,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的 额定带宽的实时上调比例,否则将所述额定带宽调整方式设置为不调整。
  8. 根据权利要求1所述的方法,其特征在于,所述根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,包括:
    根据所述当前周期内所述产品线的实时健康值的波动情况,建立所述产品线对应的健康值调整系数的实时变化函数;
    根据所述实时变化函数对所述产品线的实时健康值的导数和预设的平滑系数,以及所述实时变化函数对应的历史平滑梯度,计算所述实时变化函数对应的最新平滑梯度;
    根据历史周期内所述产品线对应的额定带宽的历史上调比例和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例。
  9. 根据权利要求8所述的方法,其特征在于,所述根据历史周期内所述产品线对应的额定带宽的历史上调比例和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例,包括:
    根据当前周期内所述产品线对应的额定带宽上调次数和预设的上调衰减率,对所述产品线当前对应的上调步长控制系数进行更新;
    根据历史周期内所述产品线对应的额定带宽的历史上调比例,以及所述最新平滑梯度与更新后的上调步长控制系数的乘积,计算所述产品线对应的额定带宽的实时上调比例。
  10. 根据权利要求1-9任一项所述的方法,其特征在于,所述健康值根据网络设备的性能指标和产品线的业务参数计算生成。
  11. 一种调整额定带宽的装置,其特征在于,所述装置包括:
    确定模块,用于对于数据传输网络中任一网络设备的任一产品线,周期性根据所述产品线的预设基准健康值和当前周期内的实时健康值,确定所述产品线的额定带宽调整方式;
    上调模块,用于如果所述额定带宽调整方式为上调,则根据所述当前周期 内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,基于所述实时上调比例和预先存储的所述产品线对应的额定带宽的基础上调比例,对所述产品线当前对应的额定带宽进行调整;
    下调模块,用于如果所述额定带宽调整方式为下调,则基于额定带宽的预设下调比例和所述当前周期内所述网络设备的实时带宽数据,对所述产品线当前对应的额定带宽进行调整。
  12. 根据权利要求11所述的装置,其特征在于,所述装置还包括计算模块,用于:
    每隔预设时长根据所述网络设备的历史带宽数据确定所述网络设备的常规带宽值,并根据所述产品线的历史健康值确定所述产品线的常规健康值;
    基于所述常规带宽值和所述常规健康值,计算并存储所述产品线对应的额定带宽的基础上调比例。
  13. 根据权利要求12所述的装置,其特征在于,所述计算模块,具体用于:
    每隔预设时长获取所述数据传输网络中所有网络设备的历史带宽数据;
    根据每个所述网络设备的历史带宽数据的数据分布特征,通过EM聚类算法对所有所述网络设备进行分类;
    对于任一类别,利用所述类别对应的预设分位数,在所述类别下的网络设备的所有历史带宽数据中确定所述类别下的每个网络设备的常规带宽值。
  14. 根据权利要求12所述的装置,其特征在于,所述计算模块,具体用于:
    每隔预设时长获取所述数据传输网络中每个网络设备的每个产品线的历史健康值;
    根据每个所述产品线的历史健康值的数据分布特征,通过EM聚类算法对所有产品线进行分类;
    对于任一类别,利用所述类别对应的预设分位数,在所述类别下的产品线的所有历史健康值中确定所述类别下的每个产品线的常规健康值。
  15. 根据权利要求12所述的装置,其特征在于,所述计算模块,具体用于:
    基于所述常规带宽值和所述产品线当前对应的额定带宽,计算所述产品线对应的额定带宽的带宽调整系数;
    基于所述常规健康值和所述产品线的预设基准健康值,计算所述产品线对应的额定带宽的健康值调整系数;
    基于所述带宽调整系数和所述健康值调整系数,以及预设的上调比例限制系数,计算并存储所述产品线对应的额定带宽的基础上调比例。
  16. 根据权利要求11所述的装置,其特征在于,所述确定模块,具体用于:
    周期性计算当前周期内所述产品线的实时健康值的平均值;
    如果所述平均值小于所述产品线的预设基准健康值,则确定所述产品线的额定带宽调整方式为上调,否则为下调。
  17. 根据权利要求11所述的装置,其特征在于,所述上调模块,具体用于:
    如果所述当前周期内所述网络设备的实时带宽中值大于所述上一周期内所述网络设备的实时带宽中值,且大于所述产品线当前对应的额定带宽,则根据所述当前周期内所述产品线的实时健康值的波动情况,计算所述产品线对应的额定带宽的实时上调比例,否则将所述额定带宽调整方式设置为不调整。
  18. 根据权利要求11所述的装置,其特征在于,所述上调模块,具体用于:
    根据所述当前周期内所述产品线的实时健康值的波动情况,建立所述产品线对应的健康值调整系数的实时变化函数;
    根据所述实时变化函数对所述产品线的实时健康值的导数和预设的平滑系数,以及所述实时变化函数对应的历史平滑梯度,计算所述实时变化函数对应的最新平滑梯度;
    根据历史周期内所述产品线对应的额定带宽的历史上调比例和所述最新平滑梯度,计算所述产品线对应的额定带宽的实时上调比例。
  19. 根据权利要求18所述的装置,其特征在于,所述上调模块,具体用于:
    根据当前周期内所述产品线对应的额定带宽上调次数和预设的上调衰减率,对所述产品线当前对应的上调步长控制系数进行更新;
    根据历史周期内所述产品线对应的额定带宽的历史上调比例,以及所述最新平滑梯度与更新后的上调步长控制系数的乘积,计算所述产品线对应的额定带宽的实时上调比例。
  20. 根据权利要求11-19任一项所述的装置,其特征在于,所述健康值根据网络设备的性能指标和产品线的业务参数计算生成。
  21. 一种额定带宽调整设备,其特征在于,所述额定带宽调整设备包括处理器和存储器,所述存储器中存储有至少一条指令、至少一段程序、代码集或指令集,所述至少一条指令、所述至少一段程序、所述代码集或指令集由所述处理器加载并执行以实现如权利要求1至10任一所述的调整额定带宽的方法。
  22. 一种计算机可读存储介质,其特征在于,所述存储介质中存储有至少一条指令、至少一段程序、代码集或指令集,所述至少一条指令、所述至少一段程序、所述代码集或指令集由处理器加载并执行以实现如权利要求1至10任一所述的调整额定带宽的方法。
PCT/CN2018/119597 2018-11-07 2018-12-06 一种调整额定带宽的方法和装置 WO2020093502A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP18939669.0A EP3876485A4 (en) 2018-11-07 2018-12-06 NOMINAL BANDWIDTH ADJUSTMENT METHOD AND DEVICE
US17/044,608 US20210111961A1 (en) 2018-11-07 2018-12-06 Method and device for adjusting rated bandwidth

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811320541.7A CN109067607B (zh) 2018-11-07 2018-11-07 一种调整额定带宽的方法和装置
CN201811320541.7 2018-11-07

Publications (1)

Publication Number Publication Date
WO2020093502A1 true WO2020093502A1 (zh) 2020-05-14

Family

ID=64789316

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/119597 WO2020093502A1 (zh) 2018-11-07 2018-12-06 一种调整额定带宽的方法和装置

Country Status (4)

Country Link
US (1) US20210111961A1 (zh)
EP (1) EP3876485A4 (zh)
CN (1) CN109067607B (zh)
WO (1) WO2020093502A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110048968B (zh) * 2019-04-12 2021-06-22 网宿科技股份有限公司 一种域名带宽调整方法及装置
US10958592B2 (en) 2019-04-12 2021-03-23 Wangsu Science & Technology Co., Ltd. Domain name bandwidth adjustment method and apparatus
CN112203102A (zh) * 2020-09-27 2021-01-08 平安国际智慧城市科技股份有限公司 网络监控的方法、装置、电子设备及介质
CN115883375A (zh) * 2022-12-02 2023-03-31 迈普通信技术股份有限公司 网络带宽处理方法、装置、设备及存储介质
CN117412083B (zh) * 2023-11-07 2024-05-14 南月(广州)机器人科技有限公司 一种用于竞技产品教学的物联网视频传输方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009127734A1 (en) * 2008-04-18 2009-10-22 Thomson Licensing Network apparatus and controlling method therefore
CN104158761A (zh) * 2014-08-05 2014-11-19 华为技术有限公司 一种分流流量的方法和装置
CN107920031A (zh) * 2016-10-11 2018-04-17 中国电信股份有限公司 接入带宽自适应调整方法和装置

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1983877A (zh) * 2004-04-14 2007-06-20 烽火通信科技股份有限公司 基于以太网无源光网络上行链路带宽动态分配方法和装置
CN100396009C (zh) * 2006-02-23 2008-06-18 华为技术有限公司 带宽控制方法、***、接入控制设备、用户档案管理设备
CN102497640B (zh) * 2011-12-09 2014-03-19 西安电子科技大学 密集局域网环境宽带信道的动态配置方法
CN103491556B (zh) * 2012-06-13 2017-06-20 华为技术服务有限公司 一种网络调整的方法及装置
CN102724193B (zh) * 2012-06-14 2015-01-21 西安交通大学 针对IP网络环境中Streaming业务生存性进行控制的方法
CN104168100B (zh) * 2014-07-24 2017-07-28 重庆邮电大学 一种时分波分混合复用无源光网络的动态资源调度方法
CN106161283B (zh) * 2015-03-31 2019-12-06 新华三技术有限公司 报文传输方法及装置
JP6466279B2 (ja) * 2015-08-05 2019-02-06 アラクサラネットワークス株式会社 通信装置
CN105376805B (zh) * 2015-11-24 2018-06-08 上海师范大学 一种基于预测业务带宽可调的异构无线网络负载均衡方法
CN105871753A (zh) * 2015-11-30 2016-08-17 乐视致新电子科技(天津)有限公司 路由器带宽管理方法、装置及路由器
CN105656810B (zh) * 2015-12-25 2020-02-07 北京奇虎科技有限公司 一种更新应用程序的方法及装置

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009127734A1 (en) * 2008-04-18 2009-10-22 Thomson Licensing Network apparatus and controlling method therefore
CN104158761A (zh) * 2014-08-05 2014-11-19 华为技术有限公司 一种分流流量的方法和装置
CN107920031A (zh) * 2016-10-11 2018-04-17 中国电信股份有限公司 接入带宽自适应调整方法和装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3876485A4 *

Also Published As

Publication number Publication date
CN109067607A (zh) 2018-12-21
EP3876485A1 (en) 2021-09-08
US20210111961A1 (en) 2021-04-15
CN109067607B (zh) 2020-09-18
EP3876485A4 (en) 2021-12-22

Similar Documents

Publication Publication Date Title
WO2020093502A1 (zh) 一种调整额定带宽的方法和装置
WO2020220739A1 (zh) 一种请求控制方法、相关设备及计算机存储介质
CN110880984B (zh) 基于模型的流量异常监测方法、装置、设备及存储介质
CN105281981B (zh) 网络服务的数据流量监控方法和装置
WO2020062439A1 (zh) 一种监控带宽状态的方法和装置
CN109194984B (zh) 一种基于边缘计算的视频帧调度方法
WO2021004063A1 (zh) 一种缓存服务器的带宽调度方法及装置
WO2019144560A1 (zh) 一种检测网络质量的方法和***
US7467291B1 (en) System and method for calibrating headroom margin
CN109819057A (zh) 一种负载均衡方法及***
CN112231075A (zh) 一种基于云服务的服务器集群负载均衡控制方法及***
CN106210129B (zh) 一种基于Web服务器配置的限流方法及***
CN105577958B (zh) 用于调整分流策略和分流用户请求的方法、装置及***
US20180295044A1 (en) Automated management of service instances
CN111200526A (zh) 网络设备的监控***及方法
EP3148150B1 (en) Method and device for determining transmission buffer size
CN114339135A (zh) 一种负载均衡方法、装置、电子设备和存储介质
CN109617758A (zh) 节点网络质量计算方法及装置、服务器、计算机存储介质
CN111901134B (zh) 一种基于循环神经网络模型rnn的预测网络质量的方法和装置
CN117177296B (zh) 基于5g网络的公共监控视频实时传输与远程播放方法
CN110245019B (zh) 一种自适应***资源的线程并发方法及装置
CN117221069A (zh) 一种基于微服务器架构的管理方法及装置
CN106940673A (zh) 一种监测项间隔智能调整方法及***
WO2020082518A1 (zh) 一种识别带宽需求突发的方法和装置
CN111754218A (zh) 支付方式推荐方法和装置

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: 18939669

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2018939669

Country of ref document: EP

Effective date: 20210603