WO2019034065A1 - 一种智能调度方法和装置及其计算机可读存储介质和计算机设备 - Google Patents

一种智能调度方法和装置及其计算机可读存储介质和计算机设备 Download PDF

Info

Publication number
WO2019034065A1
WO2019034065A1 PCT/CN2018/100557 CN2018100557W WO2019034065A1 WO 2019034065 A1 WO2019034065 A1 WO 2019034065A1 CN 2018100557 W CN2018100557 W CN 2018100557W WO 2019034065 A1 WO2019034065 A1 WO 2019034065A1
Authority
WO
WIPO (PCT)
Prior art keywords
node
bandwidth
running
value
guaranteed
Prior art date
Application number
PCT/CN2018/100557
Other languages
English (en)
French (fr)
Inventor
符立佳
苗辉
Original Assignee
贵州白山云科技股份有限公司
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 贵州白山云科技股份有限公司 filed Critical 贵州白山云科技股份有限公司
Publication of WO2019034065A1 publication Critical patent/WO2019034065A1/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
    • H04L41/32Specific management aspects for broadband networks
    • 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
    • H04L41/06Management of faults, events, alarms or notifications
    • H04L41/069Management of faults, events, alarms or notifications using logs of notifications; Post-processing of notifications
    • 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
    • H04L41/50Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5003Managing SLA; Interaction between SLA and QoS
    • H04L41/5009Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
    • 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
    • H04L41/50Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5003Managing SLA; Interaction between SLA and QoS
    • H04L41/5019Ensuring fulfilment of SLA
    • H04L41/5025Ensuring fulfilment of SLA by proactively reacting to service quality change, e.g. by reconfiguration after service quality degradation or upgrade
    • 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
    • H04L41/50Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5029Service quality level-based billing, e.g. dependent on measured service level customer is charged more or less

Definitions

  • the present invention relates to, but is not limited to, the field of computer networks, and in particular, to an intelligent scheduling method and apparatus, and computer readable storage medium and computer apparatus.
  • the current intelligent scheduling system has the following problems:
  • the current intelligent scheduling system automatically schedules according to availability and quality of service, without considering the bandwidth cost.
  • the node service with the best quality of service is preferentially selected, which is prone to cost waste.
  • Cost-based bandwidth adjustment needs manual judgment and manual completion, and the judgment and adjustment efficiency are poor, and the labor cost is high.
  • the bandwidth running high scheduling will be automatically adjusted only after the service quality is degraded, or manually adjusted after the bandwidth is full, the adjustment efficiency and the effect are poor, and the cost is high; and the service quality after the running high and the pre-adjustment time will be decline.
  • the present invention is directed to solving the problems described above.
  • an intelligent scheduling method comprising:
  • node bandwidth resources are scheduled.
  • the method before the step of determining the expected value of the node bandwidth, the method further includes:
  • the original coverage scheme combined with the IP availability detection result and the service quality of each IP address, the first coverage scheme, the network region-service node mapping table, the node IP information table, and the node bandwidth variation parameter are generated.
  • the network area-service node mapping table is a correspondence between the network area and the nodes whose service quality is up to standard.
  • the node bandwidth change parameter is the rate of change of the node's next time point bandwidth compared with the current value.
  • the bandwidth value of the next time point of the node bandwidth is the average value of the historical data or manually configured.
  • the node bandwidth change parameter is calculated according to the following expression. :
  • Node bandwidth change parameter (node next time point bandwidth - node current bandwidth) / node current bandwidth +1,
  • the node IP information table includes a node name, a guaranteed bandwidth, a standard charging bandwidth, an egress bandwidth, an uncharged high point, a charging point, and an IP list, where the guaranteed bandwidth is the lowest bandwidth that should be provided according to the payment situation.
  • the guaranteed bandwidth has a first floating interval, the guaranteed bandwidth is smaller than the first floating interval minimum, and the outgoing bandwidth is greater than the first floating interval maximum.
  • the step of determining a node bandwidth expected value comprises:
  • the node bandwidth expected value is calculated according to the following expression:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter
  • the total bandwidth value of the specified network area is calculated by accessing the log, and the total bandwidth of the network area is calculated.
  • the value is assigned to each service IP according to the resolution ratio of each service IP configured in the network area, and the bandwidth value of each service IP in the specified network area is obtained.
  • the service IP in the designated node is in each network area.
  • the bandwidth values within the sum are added to obtain the current bandwidth value of the node, and the expected bandwidth of the node is calculated according to the following expression:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter.
  • determining, according to the first coverage scheme, the bandwidth usage level of each node includes:
  • the running full threshold is calculated by the following expression:
  • Run full threshold exit bandwidth * run full factor
  • the first running overbandwidth value is calculated according to the following expression:
  • the expected bandwidth value of the node is greater than or equal to a maximum value of the first floating interval of the standard charging bandwidth and less than the running full threshold, determining that the bandwidth usage attribute of the corresponding node is “running high” and marking the second running
  • the ultra-bandwidth value, the second run-over bandwidth value is calculated according to the following expression:
  • Second running over bandwidth value node bandwidth expected value - standard charging bandwidth
  • the expected bandwidth value of the node is greater than or equal to the guaranteed bandwidth and less than the minimum value of the first floating interval, determining that the bandwidth usage attribute of the corresponding node is “running low” and marking the third running over bandwidth value, the first The three-run superbandwidth value is calculated according to the following expression:
  • the node bandwidth expected value is less than the guaranteed bandwidth, it is determined that the bandwidth usage attribute of the corresponding node is “not guaranteed to be guaranteed”, and the fourth running over bandwidth value is marked, and the fourth running over bandwidth value is calculated according to the following expression. get:
  • the step of adjusting the bandwidth configuration carried by each node according to the determination result of the bandwidth usage level of the node, the step of obtaining the second coverage scheme includes:
  • the step of scheduling the bandwidth of the first running ultra-bandwidth value to the non-"running full” node is performed by selecting a node whose bandwidth usage attribute is "running full":
  • a node that is the same network area as the "running full” node service and whose bandwidth usage attribute is "not guaranteed” is used as a backup node, and part of the first coverage scheme is selected.
  • the IP of the "running full” node is replaced with the backup node IP until the bandwidth of the "not reaching the bottom” node is fully filled to exceed the guaranteed bandwidth or all "running full” nodes become "running high” nodes;
  • the "low running” node When the bandwidth of the "not reaching the bottom” node is fully filled to exceed the guaranteed bandwidth, and there is still a “running full” node, the "low running” node is selected as the backup node, and the part of the first coverage scheme is "running".
  • the IP of the full node is replaced with the backup node IP until the bandwidth of the "low running” node is completely filled to exceed the minimum value of the first floating interval or all "running full” nodes become "running high” nodes;
  • a full alarm is sent when the "normal" node is fully filled to exceed the maximum value of the first floating interval and there is still a "running full” node.
  • the step of scheduling the bandwidth of the second running overbandwidth value to the node that is not “running full” and not “running high” by selecting the node whose bandwidth usage attribute is "running high” includes:
  • the network area-service node mapping table select a node that is the same network area as the "high” node service, and the bandwidth usage attribute is "not guaranteed” as a backup node, and part of the first coverage scheme
  • the IP of the "Run High” node is replaced with the backup node IP until the bandwidth of the "Unreached” node is fully filled to exceed the guaranteed bandwidth or all "Run High” nodes become "normal” nodes;
  • the "low running” node is selected as the backup node, and the part of the first coverage scheme is "running".
  • the IP of the high node is replaced with the backup node IP until the "low" node is fully filled to exceed the minimum value of the first floating interval or all "high” nodes become "normal” nodes;
  • the node with the highest number of uncharged high points in the “normal” node is selected as the backup node.
  • the IP of the part of the "high” node in the first coverage scheme is replaced with the backup node IP until the "normal" node is completely filled to exceed the running full threshold or all "high” nodes become "normal” nodes. ;
  • a run-high alarm is sent until the "normal" node is fully filled to exceed the run-out threshold and there is still a "run-high” node.
  • the step of selecting a node whose bandwidth usage attribute is “not guaranteed”, and scheduling a part of the bandwidth of the node that is not “guaranteed to the bottom” to the node that “has not reached the guarantee” includes:
  • the network area-service node mapping table selecting a node that is the same network area as the "not guaranteed bottom” node service and the bandwidth usage attribute is "down" as the source node, and the part of the source node in the first coverage scheme is The IP is replaced with the "Unreached” node IP until the bandwidth of the "Unreached” node is fully filled to exceed the guaranteed bandwidth or the bandwidth of all "down” nodes is reduced to the guaranteed bandwidth.
  • the method further includes:
  • the bandwidth consumption cost is determined for the second coverage scheme. Specifically include:
  • an intelligent scheduling apparatus comprising:
  • a node bandwidth prediction module configured to determine a node bandwidth expected value according to the first coverage scheme
  • a bandwidth usage evaluation module configured to determine, according to the expected bandwidth value of the node, a bandwidth usage level of each node in the first coverage scheme
  • a second solution configuration module configured to adjust a bandwidth configuration carried by each node according to a determination result of a bandwidth usage level of the node, to obtain a second coverage plan
  • the resource scheduling module is configured to schedule the node bandwidth resource according to the second coverage scheme.
  • the device further comprises:
  • the first scheme configuration module is configured to generate a first coverage scheme, a network region-service node mapping table, a node IP information table, and a node bandwidth variation parameter according to the original coverage scheme, combining the IP availability detection result and the service quality of each IP address.
  • the network area-service node mapping table is a correspondence between the network area and the nodes whose service quality is up to standard.
  • the node bandwidth change parameter is the rate of change of the node's next time point bandwidth compared with the current value.
  • the bandwidth value of the next time point of the node bandwidth is the average value of the historical data or manually configured.
  • the node bandwidth change parameter is calculated according to the following expression. :
  • Node bandwidth change parameter (node next time point bandwidth - node current bandwidth) / node current bandwidth +1,
  • the node IP information table includes a node name, a guaranteed bandwidth, a standard charging bandwidth, an egress bandwidth, an uncharged high point, a charging point, and an IP list, where the guaranteed bandwidth is the lowest bandwidth that should be provided according to the payment situation.
  • the guaranteed bandwidth has a first floating interval, the guaranteed bandwidth is smaller than the first floating interval minimum, and the outgoing bandwidth is greater than the first floating interval maximum.
  • the node bandwidth prediction module includes:
  • a change detecting unit configured to compare the original coverage scheme and the first coverage scheme
  • a first bandwidth calculation unit configured to calculate a node bandwidth expected value according to the following expression when the network area configuration corresponding to all IPs in the specified node in the original coverage scheme and the first coverage scheme does not change:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter
  • a second bandwidth calculation unit configured to calculate a total of the specified network area by using an access log when the network area configuration corresponding to all IPs in the specified node is changed in the first coverage scheme compared to the original coverage scheme
  • the bandwidth value, the total bandwidth value of the network area is allocated to each service IP according to the resolution ratio of each service IP configured in the network area, and the bandwidth value of each service IP in the specified network area is obtained, and the node IP information table is used to specify the node.
  • the service IP in the network adds the bandwidth values in each network area to obtain the current bandwidth value of the node, and calculates the expected bandwidth of the node according to the following expression:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter.
  • the bandwidth usage assessment module includes:
  • a first evaluation unit configured to determine, when the expected bandwidth value of the node is greater than or equal to a full threshold, that the bandwidth usage attribute of the corresponding node is “running full”, and marking the first running overbandwidth value, the running full threshold Calculated by the following expression:
  • Run full threshold exit bandwidth * run full factor
  • the first running overbandwidth value is calculated according to the following expression:
  • a second evaluation unit configured to determine, when the node bandwidth expected value is greater than or equal to a maximum value of the first floating interval of the standard charging bandwidth, and less than the running full threshold, determining that the bandwidth usage attribute of the corresponding node is “high” And marking the second running overbandwidth value, the second running overbandwidth value being calculated according to the following expression:
  • Second running over bandwidth value node bandwidth expected value - standard charging bandwidth
  • a third evaluation unit configured to determine, when the node bandwidth expected value is greater than or equal to the guaranteed bandwidth and less than the minimum value of the first floating interval, determining that the bandwidth usage attribute of the corresponding node is “down” and marking the third run
  • the ultra-bandwidth value, the third run-over bandwidth value is calculated according to the following expression:
  • a fourth evaluation unit configured to determine, when the expected bandwidth value of the node is less than the guaranteed bandwidth, that the bandwidth usage attribute of the corresponding node is “not guaranteed”, and marking the fourth running overbandwidth value, the fourth running super
  • the bandwidth value is calculated according to the following expression:
  • the second solution configuration module includes:
  • Running a full adjustment unit for selecting a node whose bandwidth usage attribute is “running full”, and scheduling the bandwidth of the first running overbandwidth value to a non-running full node;
  • the running height adjusting unit is configured to select a node whose bandwidth usage attribute is “running high”, and schedule the bandwidth of the second running over bandwidth value to a node that is not “running full” and not “running high”;
  • the under-guaranteed adjustment unit is used to select a node whose bandwidth usage attribute is “not guaranteed”, and to allocate a part of the bandwidth of a node that is not “guaranteed to the bottom” to the node that “has not reached the guarantee”.
  • the running full adjustment unit comprises:
  • the first level runs through the adjustment subunit for referring to the network area-service node mapping table, and selects a node that is the same network area as the “running full” node service, and the bandwidth usage attribute is “not guaranteed to be guaranteed” as a backup. Node, replacing the IP of the part of the "running full” node in the first coverage scheme with the backup node IP until the bandwidth of the "not reaching the bottom” node is fully filled to exceed the guaranteed bandwidth or all "running full” nodes become For the "run high” node;
  • the second level runs through the adjustment subunit, and is used to fill the bandwidth of the "not guaranteed bottom” node to exceed the guaranteed bandwidth, and when there is still a "running full” node, the "low running” node is selected as the backup node.
  • the third stage runs over the adjustment subunit, and is used to fill the bandwidth of the “low running” node until the minimum value of the first floating interval is exceeded, and when there is still a “running full” node, the “normal” node is selected.
  • the node with the highest number of billing high points is used as the backup node, and the IP of the part of the "running full” node in the first coverage scheme is replaced with the backup node IP until the "normal” node is completely filled to exceed the first float.
  • the interval maximum or all "running full” nodes become "running high” nodes;
  • the full alarm sub-unit is configured to send a full alarm when the "normal" node is fully filled to exceed the maximum value of the first floating interval and there is still a "running full” node.
  • the running height adjusting unit comprises:
  • the first level running height adjusting subunit is configured to refer to the network area-service node mapping table, and select a node that is the same network area as the “high running” node service, and the bandwidth usage attribute is “not guaranteed to be guaranteed” as a backup. Node, replacing the IP of the part of the "high” node in the first coverage scheme with the backup node IP until the bandwidth of the "not reaching the bottom” node is fully filled to exceed the guaranteed bandwidth or all "high” nodes become Is a "normal” node;
  • the second-level running height adjusting sub-unit is configured to select a “low running” node as a backup node when all the bandwidths of the “not reaching the bottom” node are filled up to exceed the guaranteed bandwidth and the “high running” node still exists.
  • the third-level running height adjusting sub-unit is configured to select the "normal" node in the "normal” node when the "low running” node is fully filled to exceed the minimum value of the first floating interval and there is still a "running high” node.
  • the node with the highest number of points is used as the backup node, and the IP of the part of the "running” node in the first coverage scheme is replaced with the IP of the backup node until the "normal" node is fully filled to exceed the running full threshold or all
  • the running high alarm sub-unit is configured to send a running high alarm until the "normal" node is fully filled to exceed the running full threshold and the "high" node still exists.
  • the non-guaranteed adjustment unit comprises:
  • the first level of the under-guaranteed sub-unit is used to refer to the network area-service node mapping table, and select a node that has the same network area as the "non-guaranteed" node service and the bandwidth usage attribute is "normal” as the source node, and The IP of part of the source node in the first coverage scheme is replaced by the "not reaching the bottom” node IP until the bandwidth of the "not guaranteed bottom” node is fully filled to exceed the guaranteed bandwidth or all "normal” nodes become "running” low”;
  • the second-level under-guaranteed sub-unit is configured to refer to the network area-service node mapping table, and select a node that is in the same network area as the “not guaranteed bottom” node service and whose bandwidth usage attribute is “down” as the source node. Replace the IP of part of the source node in the first coverage scheme with the “under-guarantee” node IP until the bandwidth of the “not guaranteed” node is fully filled to exceed the guaranteed bandwidth or all “down” node bandwidth drops. To the guaranteed bandwidth.
  • the device further comprises:
  • a second solution evaluation module configured to: when all the “down” node bandwidths are reduced to the guaranteed bandwidth during the process of adjusting the first coverage scheme to obtain the second coverage scheme, and the “non-guaranteed” node still exists, the second The coverage scheme performs bandwidth cost waste determination, specifically for:
  • a computer readable storage medium having stored thereon a computer program, the program being implemented by a processor to implement the steps of the above method.
  • a computer apparatus comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, the processor implementing the program The steps of the above method.
  • the first coverage scheme is simulated, and the bandwidth usage of the node is predicted according to the simulation result, and then the first coverage scheme is further adjusted to obtain a second coverage scheme that is more compatible with the actual bandwidth usage.
  • the system automatically determines and adjusts the bandwidth usage, and solves the problem that the bandwidth configuration adjustment manual operation is inefficient and the error probability is high.
  • the standard billing bandwidth generation and application mechanism is formulated, which solves the problem of lack of reasonable billing bandwidth standard confirmation mechanism. Based on the quality of service scheduling, the optimal cost-performance scheduling method and system for bandwidth cost calculation is provided, and a standard charging bandwidth calculation method based on big data analysis is provided to realize automatic adjustment and scheduling of bandwidth running to ensure high bandwidth running. The sudden increase will not increase the bandwidth cost.
  • FIG. 1 exemplarily shows a flow of an intelligent scheduling method according to Embodiment 1 of the present invention
  • FIG. 2 exemplarily shows a relationship between parameter values related to bandwidth and node bandwidth usage attributes in a node IP information table
  • FIG. 3 exemplarily shows a specific flow of step 104 in FIG. 1;
  • FIG. 4 exemplarily shows a specific flow of step 301 in FIG. 3;
  • FIG. 5 exemplarily shows a specific flow of step 302 in FIG. 3;
  • FIG. 6 exemplarily shows a specific flow of step 303 in FIG. 3;
  • FIG. 7 exemplarily shows the specific flow of step 1 in FIG. 1;
  • FIG. 8 exemplarily shows a structure of an intelligent scheduling apparatus according to Embodiment 2 of the present invention.
  • FIG. 9 exemplarily shows the structure of the node bandwidth prediction module 801 of FIG. 8;
  • FIG. 10 exemplarily shows the structure of the bandwidth usage evaluation module 802 of FIG. 8;
  • FIG. 11 exemplarily shows the structure of the second scenario configuration module 803 of FIG. 8;
  • FIG. 12 exemplarily shows the structure of the running full adjustment unit 1101 in FIG. 11;
  • FIG. 13 exemplarily shows the structure of the running height adjusting unit 1102 of FIG. 11;
  • Fig. 14 exemplarily shows the structure of the unsupported adjustment unit 1103 of Fig. 11.
  • the current intelligent scheduling system has the following problems:
  • the current intelligent scheduling system automatically schedules according to availability and quality of service, without considering the bandwidth cost.
  • the node service with the best quality of service is preferentially selected, which is prone to cost waste.
  • Cost-based bandwidth adjustment needs manual judgment and manual completion, and the judgment and adjustment efficiency are poor, and the labor cost is high.
  • the bandwidth running high scheduling will be automatically adjusted only after the service quality is degraded, or manually adjusted after the bandwidth is full, the adjustment efficiency and the effect are poor, and the cost is high; and the service quality after the running high and the pre-adjustment time will be decline.
  • an embodiment of the present invention provides an intelligent scheduling method and apparatus, performing a simulation operation on a first coverage scheme, predicting a bandwidth usage of a node according to a simulation result, and further adjusting the first coverage scheme to obtain and
  • the second coverage scheme which is more compatible with the actual bandwidth, realizes automatic judgment and adjustment of the bandwidth usage of the system, and solves the problem that the bandwidth configuration adjustment manual operation is inefficient and the error probability is high.
  • the standard billing bandwidth generation and application mechanism is formulated, which solves the problem of lack of reasonable billing bandwidth standard confirmation mechanism.
  • An embodiment of the present invention provides an intelligent scheduling method.
  • the process of adjusting the bandwidth configuration in the scheduling solution by using the method is as shown in FIG. 1 , and includes:
  • Step 101 Generate a first coverage scheme, a network region-service node mapping table, a node IP information table, and a node bandwidth variation parameter according to the original coverage scheme, combining the IP availability detection result and the service quality of each IP address.
  • the network area-service node mapping table is a correspondence between the network area and the nodes whose service quality is up to standard
  • the node bandwidth change parameter is the rate of change of the node's next time point bandwidth compared with the current value.
  • the bandwidth value of the next time point of the node bandwidth is the average value of the historical data or manually configured.
  • the node bandwidth change parameter is calculated according to the following expression. :
  • Node bandwidth change parameter (node next time point bandwidth - node current bandwidth) / node current bandwidth +1,
  • the node IP information table includes a node name, a guaranteed bandwidth, a standard charging bandwidth, an egress bandwidth, an uncharged high point, a charging point, and an IP list, where the guaranteed bandwidth is the lowest bandwidth that should be provided according to the payment situation.
  • the guaranteed bandwidth has a first floating interval, the guaranteed bandwidth is smaller than the first floating interval minimum, and the outgoing bandwidth is greater than the first floating interval maximum.
  • Figure 2 is a schematic diagram showing the relationship of each parameter value.
  • Step 102 Determine, according to the first coverage scheme, a node bandwidth expected value.
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter
  • the designated node is the original coverage scheme and all nodes that provide services related to the first coverage scheme.
  • the total bandwidth value of the specified network area is calculated by accessing the log, and the network area is determined.
  • the total bandwidth value is allocated to each service IP according to the resolution ratio of each service IP configured in the network area, and the bandwidth value of each service IP in the specified network area is obtained.
  • the service IPs in the designated node are respectively The bandwidth values in the network area are added to obtain the current bandwidth value of the node, and the expected bandwidth of the node is calculated according to the following expression:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter.
  • Step 103 Determine, according to the expected bandwidth value of the node, a bandwidth usage level of each node in the first coverage scheme.
  • the bandwidth usage level is embodied by the bandwidth usage attribute, and the bandwidth usage attribute has four values: “running full”, “running high”, “running low”, and “not reaching the guaranteed bottom”.
  • the meanings of the four values are as follows:
  • Run full threshold exit bandwidth * run full factor
  • the first running overbandwidth value is calculated according to the following expression:
  • Second running over bandwidth value node bandwidth expected value - standard charging bandwidth
  • the third running superbandwidth value is calculated according to the following expression:
  • the third running overband value is a negative value
  • Fourth running super bandwidth value node bandwidth expected value - standard charging bandwidth
  • the fourth running overband value is a negative value.
  • Step 104 Adjust a bandwidth configuration carried by each node according to a determination result of a bandwidth usage level of the node, to obtain a second coverage plan.
  • Step 301 Select a node whose bandwidth usage attribute is “running full”, and schedule the bandwidth of the first running overbandwidth value to a non-running full node;
  • Step 401 Refer to the network area-service node mapping table, select a node that is the same network area as the “running full” node service, and use a node whose bandwidth usage attribute is “not guaranteed” as a backup node, and the first coverage.
  • the IP of the part of the "running full” node in the scheme is replaced with the backup node IP until the bandwidth of the "non-guaranteed” node is fully filled to exceed the guaranteed bandwidth or all "running full” nodes become "running high” nodes;
  • Step 402 When the bandwidth of the “not reaching the bottom” node is completely filled to exceed the guaranteed bandwidth, and the “running full” node still exists, the “low running” node is selected as the backup node, and the first coverage plan is used. The IP of the part of the "running full” node is replaced with the backup node IP until the bandwidth of the "low running” node is completely filled to exceed the minimum value of the first floating interval or all "running full” nodes become "running high” nodes;
  • Step 403 When the bandwidth of the “low running” node is completely filled to exceed the minimum value of the first floating interval, and there is still a “running full” node, the number of uncharged high points in the “normal” node is selected.
  • the node replaces the IP of the part of the "running full” node in the first coverage scheme with the backup node IP until the "normal" node is fully filled to exceed the maximum value of the first floating interval or all "running full "The node becomes a "run high” node;
  • Step 404 Send a full alarm when the "normal" node is fully filled to exceed the maximum value of the first floating interval and there is still a “running full” node.
  • Step 302 Select a node whose bandwidth usage attribute is “running high”, and schedule the bandwidth of the second running ultra-bandwidth value to a node that is not “running full” and not “running high”;
  • Step 501 Referring to the network area-service node mapping table, select, as the backup node, a node that is in the same network area as the “running” node service and whose bandwidth usage attribute is “not guaranteed”, and the first coverage.
  • the IP of the part of the "running high” node in the scheme is replaced with the backup node IP until the bandwidth of the "not reaching the bottom” node is completely filled to exceed the guaranteed bandwidth or all "running high” nodes become "normal” nodes;
  • Step 502 When all the bandwidths of the “not guaranteed bottom” node are filled to exceed the guaranteed bandwidth and the “high” node still exists, the “low running” node is selected as the backup node, and the first coverage plan is used. The IP of the part of the "running high” node is replaced with the backup node IP until the "low running” node is completely filled to exceed the minimum value of the first floating interval or all "running high” nodes become "normal” nodes;
  • Step 503 When the “low running” node is completely filled to exceed the minimum value of the first floating interval and there is still a “running high” node, the node with the highest number of uncharged high points in the “normal” node is selected as the backup. Node, replacing the IP of the part of the "high” node in the first coverage scheme with the backup node IP until the "normal” node is completely filled to exceed the running full threshold or all "running high” nodes become " Normal” node;
  • Step 504 A high alarm is sent until the "normal" node is fully filled to exceed the running full threshold and the "high” node still exists.
  • Step 303 Select a node whose bandwidth usage attribute is “not guaranteed to be guaranteed”, and schedule a part of the bandwidth of the node that is not “guaranteed to the bottom” to the node that “has not reached the guarantee”;
  • Step 601 Referring to the network area-service node mapping table, select a node that is in the same network area as the “non-guaranteed” node service and the bandwidth usage attribute is “normal” as the source node, and part of the source in the first coverage scheme.
  • the IP of the node is replaced by the "Unreached” node IP until the bandwidth of the "Unreached” node is fully filled to exceed the guaranteed bandwidth or all "normal” nodes become “down”;
  • Step 602 Referring to the network area-service node mapping table, select a node that has the same network area as the “not guaranteed bottom” node service and the bandwidth usage attribute is “down” as the source node, and part of the first coverage plan.
  • the IP of the source node is replaced with the "Unreached” node IP until the bandwidth of the "Unreached” node is fully filled to exceed the guaranteed bandwidth or the bandwidth of all "down” nodes is reduced to the guaranteed bandwidth.
  • Step 105 Perform scheduling on a node bandwidth resource according to the second coverage scheme.
  • Step 106 Perform bandwidth cost on the second coverage scheme when all the “down” node bandwidths are reduced to the guaranteed bandwidth and the “non-guaranteed” node still exists in the process of adjusting the first coverage scheme to obtain the second coverage scheme. Waste judgment
  • Step 701 Determine a bandwidth usage attribute of each node in the second coverage scheme.
  • Step 702 When a specified node that is greater than the first running overbandwidth value or the second running overbandwidth value of the charging point occurs in the evaluation period, it is determined that the bandwidth is full and an alarm is sent;
  • Step 703 When a fourth running overbandwidth value whose absolute value is greater than the number of charging points occurs in the evaluation period, it is determined that the bottom line is wasted and an alarm is sent;
  • Step 704 When a specified node whose absolute value is greater than the third running ultra-bandwidth value of the charging point is found in the evaluation period, it is determined that the standard charging bandwidth is set too high, and an alarm notification is sent.
  • the embodiment of the invention provides an intelligent scheduling device, which has the structure shown in FIG. 8 and includes:
  • a node bandwidth prediction module 801 configured to determine a node bandwidth expected value according to the first coverage scheme
  • a bandwidth usage evaluation module 802 configured to determine, according to the expected bandwidth value of the node, a bandwidth usage level of each node in the first coverage scheme
  • the second solution configuration module 803 is configured to adjust a bandwidth configuration carried by each node according to a determination result of a bandwidth usage level of the node, to obtain a second coverage plan.
  • the resource scheduling module 804 is configured to schedule the node bandwidth resource according to the second coverage scheme.
  • the device further comprises:
  • the first scheme configuration module 805 is configured to generate a first coverage scheme, a network region-service node mapping table, a node IP information table, and a node bandwidth variation parameter according to the original coverage scheme, combining the IP availability detection result and the service quality of each IP address.
  • the network area-service node mapping table is a correspondence between the network area and the nodes whose service quality is up to standard.
  • the node bandwidth change parameter is the rate of change of the node's next time point bandwidth compared with the current value.
  • the bandwidth value of the next time point of the node bandwidth is the average value of the historical data or manually configured.
  • the node bandwidth change parameter is calculated according to the following expression. :
  • Node bandwidth change parameter (node next time point bandwidth - node current bandwidth) / node current bandwidth +1,
  • the node IP information table includes a node name, a guaranteed bandwidth, a standard charging bandwidth, an egress bandwidth, an uncharged high point, a charging point, and an IP list, where the guaranteed bandwidth is the lowest bandwidth that should be provided according to the payment situation.
  • the guaranteed bandwidth has a first floating interval, the guaranteed bandwidth is smaller than the first floating interval minimum, and the outgoing bandwidth is greater than the first floating interval maximum.
  • the structure of the node bandwidth prediction module 801 is as shown in FIG. 9 and includes:
  • a change detecting unit 901 configured to compare the original coverage scheme and the first coverage scheme
  • the first bandwidth calculation unit 902 is configured to calculate a node bandwidth expected value according to the following expression when there is no change in the network area configuration corresponding to all IPs in the specified node in the original coverage scheme and the first coverage scheme:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter
  • a second bandwidth calculation unit 903 configured to calculate a specified network area by using an access log when the network area configuration corresponding to all IPs in the specified node is changed in the first coverage scheme compared to the original coverage scheme.
  • the total bandwidth value is allocated to each service IP according to the resolution ratio of each service IP configured according to the network area, and the bandwidth value of each service IP in the specified network area is combined with the node IP information table.
  • the service IPs in the nodes add the bandwidth values in the respective network areas to obtain the current bandwidth value of the node, and calculate the expected bandwidth of the node according to the following expression:
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter.
  • the structure of the bandwidth usage assessment module 802 is as shown in FIG. 10, and includes:
  • the first evaluation unit 1001 is configured to determine, when the expected bandwidth value of the node is greater than or equal to the full threshold, that the bandwidth usage attribute of the corresponding node is “running full” and mark the first running overbandwidth value, the running full threshold.
  • Run full threshold exit bandwidth * run full factor
  • the first running overbandwidth value is calculated according to the following expression:
  • the second evaluation unit 1002 is configured to determine that the bandwidth usage attribute of the corresponding node is “running” when the expected bandwidth value of the node is greater than or equal to a maximum value of the first floating interval of the standard charging bandwidth and less than the running full threshold. High", and marking the second run over bandwidth value, the second run over bandwidth value is calculated according to the following expression:
  • Second running over bandwidth value node bandwidth expected value - standard charging bandwidth
  • the third evaluation unit 1003 is configured to determine, when the node bandwidth expected value is greater than or equal to the guaranteed bandwidth and less than the minimum value of the first floating interval, determining that the bandwidth usage attribute of the corresponding node is “down” and marking the third Running the overbandwidth value, the third runover bandwidth value is calculated according to the following expression:
  • the fourth evaluation unit 1004 is configured to determine, when the node bandwidth expected value is less than the guaranteed bandwidth, that the bandwidth usage attribute of the corresponding node is “not guaranteed”, and mark the fourth running over bandwidth value, the fourth running
  • the superbandwidth value is calculated according to the following expression:
  • the structure of the second solution configuration module 803 is as shown in FIG. 11 and includes:
  • the full-running adjustment unit 1101 is configured to select a node whose bandwidth usage attribute is “running full”, and schedule the bandwidth of the first running over-bandwidth value to a non-running full node;
  • the running height adjusting unit 1102 is configured to select a node whose bandwidth usage attribute is “running high”, and schedule the bandwidth of the second running ultra-bandwidth value to a node that is not “running full” and not “running high”;
  • the bottom-up adjustment unit 1103 is configured to select a node whose bandwidth usage attribute is “not guaranteed to be guaranteed”, and to allocate a partial bandwidth of a node that is not “guaranteed to the bottom” to the node that “has not reached the guarantee”.
  • the structure of the run full adjustment unit 1101 is as shown in FIG. 12, and includes:
  • the first level runs through the adjustment sub-unit 1201, and is configured to refer to the network area-service node mapping table, select a node that has the same network area as the “running full” node service, and the bandwidth usage attribute is “not guaranteed to be guaranteed”.
  • the backup node replaces the IP of the part of the "running full” node in the first coverage scheme with the backup node IP until the bandwidth of the "non-guaranteed" node is fully filled to exceed the guaranteed bandwidth or all "running full” nodes become a "high” node;
  • the second stage runs through the adjustment sub-unit 1202, and is configured to select the “running low” node as the backup node when the bandwidth of the “not reaching the bottom” node is completely filled to exceed the guaranteed bandwidth, and the “running full” node still exists.
  • the node becomes a "run high" node;
  • the third stage runs over the adjustment subunit 1203, and is used to select the "normal" node when the bandwidth of the "low running” node is completely filled to exceed the minimum value of the first floating interval, and there is still a "running full” node.
  • the node with the highest number of uncharged high points is used as the backup node, and the IP of the part of the "running full” node in the first coverage scheme is replaced with the backup node IP until the "normal" node is completely filled to exceed the first
  • the maximum value of the floating interval or all "running full” nodes becomes the "high" node;
  • the full alarm sub-unit 1204 is configured to send a full alarm when the "normal" node is fully filled to exceed the maximum value of the first floating interval and there is still a "running full” node.
  • the structure of the running height adjusting unit 1102 is as shown in FIG. 13 and includes:
  • the first level running height adjustment sub-unit 1301 is configured to refer to the network area-service node mapping table, and select a node that has the same network area as the “high-running” node service and whose bandwidth usage attribute is “not guaranteed”.
  • the backup node replaces the IP of the part of the "high” node in the first coverage scheme with the backup node IP until the bandwidth of the "non-guaranteed” node is completely filled to exceed the guaranteed bandwidth or all "high” nodes Becomes a "normal” node;
  • the second-level running height adjusting sub-unit 1302 is configured to select the “low running” node as the backup node when all the bandwidths of the “not reaching the bottom” node are filled to exceed the guaranteed bandwidth and the “high” node still exists. And replacing the IP of the part of the "high” node in the first coverage scheme with the backup node IP until the "low running” node is completely filled to exceed the minimum value of the first floating interval or all "high” nodes become Is a "normal" node;
  • the third-level running height adjusting sub-unit 1303 is configured to select the "normal" node in the "normal” node when the "low running” node is fully filled to exceed the minimum value of the first floating interval and there is still a "running high” node.
  • the node with the highest number of nodes as the backup node replaces the IP of the part of the "high” node in the first coverage scheme with the backup node IP until the "normal" node is fully filled to exceed the running full threshold or All "run high” nodes become "normal” nodes;
  • the running high alarm sub-unit 1304 is configured to send a running high alarm until the "normal" node is fully filled to exceed the running full threshold and the "high" node still exists.
  • the structure of the non-guaranteed adjustment unit 1103 is as shown in FIG. 14 and includes:
  • the first-level non-guaranteed adjustment sub-unit 1401 is configured to refer to the network area-service node mapping table, and select a node that has the same network area as the “not guaranteed bottom” node service and the bandwidth usage attribute is “normal” as the source node. Replace the IP of part of the source nodes in the first coverage scheme with the "under-guaranteed” node IP until the bandwidth of the "non-guaranteed” node is fully filled to exceed the guaranteed bandwidth or all "normal” nodes become " Run low"
  • the second level non-guaranteed adjustment sub-unit 1402 is configured to refer to the network area-service node mapping table, and select a node that has the same network area as the "not guaranteed bottom” node service and whose bandwidth usage attribute is "down” as the source node. Replace the IP of part of the source nodes in the first coverage scheme with the "under-guaranteed” node IP until the bandwidth of the "non-guaranteed” node is fully filled to exceed the guaranteed bandwidth or all "low-low” node bandwidths. Drop to the guaranteed bandwidth.
  • the device further comprises:
  • a second solution evaluation module 806, configured to: when all the “down” node bandwidths are reduced to the guaranteed bandwidth during the process of adjusting the first coverage plan, and the “non-guaranteed” node still exists,
  • the second coverage scheme determines the bandwidth cost waste, specifically for:
  • An embodiment of the present invention provides an intelligent scheduling method, which receives a first coverage scheme, a network area-service node mapping table, and a node IP information table, and sequentially performs a bandwidth of a guaranteed bandwidth, a bandwidth running, and a standard charging model to ensure each The charging bandwidth of the node is greater than the guaranteed bandwidth, the high bandwidth is scheduled to the node that does not reach the standard charging bandwidth, or the remaining uncharged high-point node service is used to ensure that the charging bandwidth of each node is within 5% of the standard charging bandwidth ( 5% is the experience value, which can be adjusted according to the actual situation) to achieve the purpose of cost control.
  • Guaranteed bandwidth the bandwidth to be paid regardless of whether it is used or not, so this part should be used first.
  • the charging method of the node bandwidth is generally monthly billing, and the billing method has three peaks and 95 billing.
  • B.95 billing is a bandwidth point every 5 minutes, according to the 95% peak charging, that is, 433 high points will not be charged;
  • the bandwidth of each node runs averaging, and the charging bandwidth is 60%-80% of the total export bandwidth is the optimal cost performance model; further rational application of the non-billing high point to service traffic burst and run high, The goal of achieving the best price/performance ratio.
  • the original coverage scheme is the correspondence between the network area and the node IP.
  • the coverage scheme that the current scheduling module is using is the original version before the first coverage scheme is calculated based on the availability and the quality of service;
  • the first coverage scheme is the correspondence between the network area and the node IP, and in the embodiment of the present invention, the coverage scheme based on the IP availability and the service quality calculation
  • Iii Network Area-Service Node Mapping Table: Correspondence between the network area and the nodes with the quality of service standards. In the embodiment of the present invention, the network area based on IP availability and service quality is used to meet the service quality standard. List of nodes;
  • Node bandwidth change parameter the rate of change of the bandwidth of the node at the next time point compared with the current value.
  • the bandwidth value of the next time point of the node bandwidth is the average value of the historical data or manually configured.
  • Node bandwidth change parameter (node next time point bandwidth - node current bandwidth) / node current bandwidth +1
  • Node IP information table basic information of the node, including node name, guaranteed bandwidth, standard charging bandwidth, egress bandwidth, number of uncharged high points, billing points, IP list
  • Standard billing bandwidth unit M, set by historical bandwidth data analysis and manual configuration
  • F.IP list list of IPs in the node
  • the business process is as follows:
  • the standard charging bandwidth in the node IP information table is obtained by summing the basic bandwidth and the incremental bandwidth;
  • the basic bandwidth is obtained by averaging the synchronous bandwidth of the previous charging period, and the response bandwidth is obtained.
  • the same bandwidth of the previous charging period is the same day of the week, and the bandwidth is the current charging bandwidth. For example, all Tuesdays averaged last month, which is the base bandwidth of the existing billing cycle on Tuesday;
  • Incremental bandwidth is manually set by the resource administrator according to the project conditions
  • the current bandwidth value of the node is calculated as follows: the total bandwidth value of the specified network area is calculated through the access log; and the total bandwidth value of the network area is calculated.
  • the resolution ratio of each service IP is allocated to each service IP, and the bandwidth value of each service IP in the specified network area is obtained; combined with the node IP information table, the service IP in the designated node is in each network area.
  • the bandwidth values are added to obtain the current bandwidth value of the node;
  • Node bandwidth expected value node current bandwidth value * node bandwidth change parameter
  • Ii Determine whether the expected bandwidth of the node is greater than the standard charging bandwidth in the node IP information table, and the greater than 5% (ie, the exit bandwidth of 1.05 bits) and less than the full threshold (the full threshold is the exit bandwidth *0.91) If yes, the node attribute is marked as "running high” and the second running overband value is "node bandwidth expected value - standard charging bandwidth".
  • Iii determining whether the expected bandwidth of the node is lower than the guaranteed bandwidth in the node IP information table. If yes, the node attribute is marked as “not guaranteed”, and the fourth running overband value is “node bandwidth expected value-guaranteed bandwidth” ".
  • Iv. Determine whether the expected bandwidth of the node is lower than the standard charging bandwidth (more than 5%) in the node IP information table and higher than the guaranteed bandwidth. If it is satisfied, the attribute of the changed node is marked as “low running”, and the third running is exceeded.
  • the bandwidth value is "node bandwidth expected value - standard charging bandwidth”.
  • the node v. determining that the expected bandwidth of the node is greater than the guaranteed bandwidth, and the difference from the standard charging bandwidth is within plus or minus 5% (the range of the first floating interval is 5% of the standard charging bandwidth), and if so, the node is marked It is normal.
  • A Referring to the network area-service node mapping table, select the node with the same network area as the running full node service, and the node attribute is “not guaranteed” as the backup node, and pass the IP of the partial full node in the coverage scheme. Replace the IP address of the backup node, and realize the scheduling of the bandwidth from the running full node to the node that has not reached the guaranteed bottom until the guaranteed bottom is full or the full node becomes "running high";
  • A Referring to the network area-service node mapping table, select the node with the same network area as the running node service and the node attribute is “not guaranteed” as the backup node, and pass the IP of the partial full node in the coverage scheme. Replace the IP address of the backup node, and realize the scheduling of the bandwidth from the running full node to the node that has not reached the guaranteed bottom until the guaranteed bottom is full or the running high node becomes "normal";
  • the node with the highest number of uncharged high points in the optional normal node is selected as the backup node, which will run the value.
  • the bandwidth is scheduled to the backup node until the normal node runs to 91% of the bandwidth exit or the running node becomes "normal";
  • A. Referring to the network area-service node mapping table, select the node that is the same network area as the node service that does not reach the guaranteed bottom, and the node attribute is "normal" as the source node, and replace the IP of part of the source node in the coverage scheme. In the manner that the IP of the guaranteed node is not reached, the bandwidth is dispatched from the normal attribute node to the node that has not reached the guaranteed bottom until the guaranteed bottom is fully filled or the normal node becomes "running low";
  • Embodiments of the present invention also provide a computer readable storage medium having stored thereon a computer program that, when executed by a processor, implements the steps of the above method.
  • Embodiments of the present invention also provide a computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor performing the steps of the method.
  • the first coverage scheme is simulated, and the bandwidth usage of the node is predicted according to the simulation result, and then the first coverage scheme is further adjusted to obtain a second coverage scheme that is more compatible with the actual bandwidth usage.
  • the system automatically determines and adjusts the bandwidth usage, and solves the problem that the bandwidth configuration adjustment manual operation is inefficient and the error probability is high.
  • the standard billing bandwidth generation and application mechanism is formulated, which solves the problem of lack of reasonable billing bandwidth standard confirmation mechanism. Based on the quality of service scheduling, the optimal cost-performance scheduling method and system for bandwidth cost calculation is provided, and a standard charging bandwidth calculation method based on big data analysis is provided to realize automatic adjustment and scheduling of bandwidth running to ensure high bandwidth running. The sudden increase will not increase the bandwidth cost.
  • the invention provides an intelligent scheduling method and device and computer readable storage medium and computer device thereof, which realize automatic judgment and adjustment of bandwidth usage, and solve the problem that the bandwidth configuration adjustment manual operation efficiency is low and the error probability is high.
  • the standard billing bandwidth generation and application mechanism is formulated, which solves the problem of lack of reasonable billing bandwidth standard confirmation mechanism.
  • the optimal cost-performance scheduling method and system for bandwidth cost calculation is provided, and a standard charging bandwidth calculation method based on big data analysis is provided to realize automatic adjustment and scheduling of bandwidth running to ensure high bandwidth running. The sudden increase will not increase the bandwidth cost.

Landscapes

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

Abstract

本发明的实施例提供了一种智能调度方法和装置及其计算机可读存储介质和计算机设备。所公开的方法包括:根据第一覆盖方案,确定节点带宽预期值;根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;根据所述第二覆盖方案,对节点带宽资源进行调度。本发明提供的技术方案适用于内容分发网络带宽资源调度,实现了了系统对带宽使用情况自动的判断和调整。

Description

一种智能调度方法和装置及其计算机可读存储介质和计算机设备
本申请要求在2017年8月15日提交中国专利局、申请号为201710695232.7、发明名称为“一种智能调度方法和装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本发明涉及但不限于计算机网络领域,尤其涉及一种智能调度方法和装置及其计算机可读存储介质和计算机设备。
背景技术
目前互联网高速发展,智能调度平台被广泛用于将用户的请求引导至最优的边缘服务器以获得最佳服务。随着网民对网络访问质量要求的不断提升以及带宽成本在整体运营成本中的比例不断增加,如何实现一种基于高性价比服务的智能调度平台成为运营的重要部分。
当前的智能调度系统存在如下问题:
a)当前智能调度系统根据可用性和服务质量进行自动调度,没有考虑带宽成本,在可用性容错和服务质量调优中,会优先选择服务质量最好的节点服务,容易发生成本浪费的情况。
b)基于成本的带宽调整需要人工判断并手动完成,判断、调整效率差,人工成本高。
c)带宽跑高调度只有在服务质量下降后才会自动调整,或者人工查看带宽跑满后手工调整,调整效率和效果差、成本高;且在跑高后和调整前的时间会使服务质量下降。
d)对计费带宽的标准制定缺乏一套可行机制。
发明内容
本发明旨在解决上面描述的问题。
根据本发明的第一方面,提供了一种智能调度方法,包括:
根据第一覆盖方案,确定节点带宽预期值;
根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;
根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;
根据所述第二覆盖方案,对节点带宽资源进行调度。
优选的,根据第一覆盖方案,确定节点带宽预期值的步骤之前还包括:
根据原始覆盖方案,结合IP可用性检测结果与各IP地址的服务质量,生成第一覆盖方案、网络区域-服务节点映射表、节点IP信息表和节点带宽变化参数,
其中,
网络区域-服务节点映射表为网络区域和服务质量达标的节点间的对应关系,
节点带宽变化参数为节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置,所述节点带宽变化参数根据以下表达式计算得到:
节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带宽+1,
节点IP信息表包含节点名称、保底带宽、标准计费带宽、出口带宽、未计费高点个数、计费点数、IP列表,所述保底带宽为根据付费情况节点最低应提供的带宽,所述标准计费带宽为基础带宽与增量带宽之和,所述基础带宽通过对上一个计费周期的同期带宽求平均值得到,所述出口带宽>=所述标准计费带宽>=所述保底带宽,所述标准计费带宽具有第一浮动区间,所述保底带宽小于所述第一浮动区间最小值,所述出口带宽大于所述第一浮动区间最大值。
优选的,根据第一覆盖方案,确定节点带宽预期值的步骤包括:
对比所述原始覆盖方案和所述第一覆盖方案;
在所述原始覆盖方案和所述第一覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化时,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
在相较于所述原始覆盖方案,所述第一覆盖方案中指定节点内所有IP对应的网络区域配置发生了变化时,通过访问日志计算出指定网络区域的总带宽值,将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值,结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数。
优选的,根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平的步骤包括:
在所述节点带宽预期值大于等于跑满门限值时,判定相应节点的带宽使用属性为“跑满”,并标记第一跑超带宽值,所述跑满门限值通过以下表达式计算得到:
跑满门限值=出口带宽*跑满系数,
所述第一跑超带宽值根据以下表达式计算得到:
第一跑超带宽值=节点带宽预期值-标准计费带宽;
在所述节点带宽预期值大于等于所述标准计费带宽的第一浮动区间最大值且小于所述跑满门限值时,判定相应节点的带宽使用属性为“跑高”,并标记第二跑超带宽值,所述第二跑超带宽值根据以下表达式计算得到:
第二跑超带宽值=节点带宽预期值-标准计费带宽;
在所述节点带宽预期值在所述第一浮动区间内时,判定相应节点的带宽使用属性为“正常”;
在所述节点带宽预期值大于等于所述保底带宽且小于所述第一浮动区间最小值时,判定相应节点的带宽使用属性为“跑低”,并标记第三跑超带宽值,所述第三跑超带宽值根据以下表达式计算得到:
第三跑超带宽值=节点带宽预期值-标准计费带宽;
在所述节点带宽预期值小于所述保底带宽时,判定相应节点的带宽使用属性为“未达到保底”,并标记第四跑超带宽值,所述第四跑超带宽值根据以下 表达式计算得到:
第四跑超带宽值=节点带宽预期值-标准计费带宽。
优选的,根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案的步骤包括:
选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点;
选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点;
选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点。
优选的,选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点的步骤包括:
参照所述网络区域-服务节点映射表,选取与所述“跑满”节点服务相同网络区域,且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑满”节点变为“跑高”节点;
在“未达到保底”节点的带宽全部填满至超过所述保底带宽,且仍存在“跑满”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值或者全部“跑满”节点变为“跑高”节点;
在“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值,且仍存在“跑满”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述第一浮动区间最大值或者全部“跑满”节点变为“跑高”节点;
在“正常”节点全部填满至超过所述第一浮动区间最大值且仍存在“跑满”节点时,发送跑满告警。
优选的,选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽 调度到非“跑满”且非“跑高”节点的步骤包括:
参照所述网络区域-服务节点映射表,选取与所述“跑高”节点服务相同网络区域、且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑高”节点变为“正常”节点;
在全部“未达到保底”节点的带宽全部填满至超过所述保底带宽且仍存在“跑高”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“跑低”节点全部填满至超过所述第一浮动区间最小值或者全部“跑高”节点变为“正常”节点;
在“跑低”节点全部填满至超过所述第一浮动区间最小值且仍存在“跑高”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述跑满门限值或全部“跑高”节点变为“正常”节点;
直到“正常”节点全部填满至超过所述跑满门限值且仍存在“跑高”节点时,发送跑高告警。
优选的,选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点的步骤包括:
参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“正常”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“正常”节点变为“跑低”;
参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“跑低”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑低”节点带宽降至保底带宽。
优选的,该方法还包括:
在所述第一覆盖方案调整得到所述第二覆盖方案过程中全部“跑低”节点带宽降至保底带宽且仍存在“未达到保底”节点时,对第二覆盖方案进行带宽 成本浪费判定,具体包括:
确定所述第二覆盖方案下各个节点的带宽使用属性;
在评测周期内出现大于计费点数的第一跑超带宽值或第二跑超带宽值的指定节点时,判定为带宽跑满,发送告警;
在评测周期内出现绝对值大于计费点数的第四跑超带宽值时,判定为保底浪费,发送告警;
在评测周期内出现绝对值大于计费点数的第三跑超带宽值的指定节点时,判定为标准计费带宽设置过高,发送告警通知。
根据本发明的另一方面,提供了一种智能调度装置,包括:
节点带宽预测模块,用于根据第一覆盖方案,确定节点带宽预期值;
带宽使用情况评估模块,用于根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;
第二方案配置模块,用于根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;
资源调度模块,用于根据所述第二覆盖方案,对节点带宽资源进行调度。
优选的,该装置还包括:
第一方案配置模块,用于根据原始覆盖方案,结合IP可用性检测结果与各IP地址的服务质量,生成第一覆盖方案、网络区域-服务节点映射表、节点IP信息表和节点带宽变化参数,
其中,
网络区域-服务节点映射表为网络区域和服务质量达标的节点间的对应关系,
节点带宽变化参数为节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置,所述节点带宽变化参数根据以下表达式计算得到:
节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带宽+1,
节点IP信息表包含节点名称、保底带宽、标准计费带宽、出口带宽、未计 费高点个数、计费点数、IP列表,所述保底带宽为根据付费情况节点最低应提供的带宽,所述标准计费带宽为基础带宽与增量带宽之和,所述基础带宽通过对上一个计费周期的同期带宽求平均值得到,所述出口带宽>=所述标准计费带宽>=所述保底带宽,所述标准计费带宽具有第一浮动区间,所述保底带宽小于所述第一浮动区间最小值,所述出口带宽大于所述第一浮动区间最大值。
优选的,所述节点带宽预测模块包括:
变化检测单元,用于对比所述原始覆盖方案和所述第一覆盖方案;
第一带宽计算单元,用于在所述原始覆盖方案和所述第一覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化时,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
第二带宽计算单元,用于在相较于所述原始覆盖方案,所述第一覆盖方案中指定节点内所有IP对应的网络区域配置发生了变化时,通过访问日志计算出指定网络区域的总带宽值,将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值,结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数。
优选的,所述带宽使用情况评估模块包括:
第一评估单元,用于在所述节点带宽预期值大于等于跑满门限值时,判定相应节点的带宽使用属性为“跑满”,并标记第一跑超带宽值,所述跑满门限值通过以下表达式计算得到:
跑满门限值=出口带宽*跑满系数,
所述第一跑超带宽值根据以下表达式计算得到:
第一跑超带宽值=节点带宽预期值-标准计费带宽;
第二评估单元,用于在所述节点带宽预期值大于等于所述标准计费带宽的第一浮动区间最大值且小于所述跑满门限值时,判定相应节点的带宽使用属性为“跑高”,并标记第二跑超带宽值,所述第二跑超带宽值根据以下表达式计算得到:
第二跑超带宽值=节点带宽预期值-标准计费带宽;
在所述节点带宽预期值在所述第一浮动区间内时,判定相应节点的带宽使用属性为“正常”;
第三评估单元,用于在所述节点带宽预期值大于等于所述保底带宽且小于所述第一浮动区间最小值时,判定相应节点的带宽使用属性为“跑低”,并标记第三跑超带宽值,所述第三跑超带宽值根据以下表达式计算得到:
第三跑超带宽值=节点带宽预期值-标准计费带宽;
第四评估单元,用于在所述节点带宽预期值小于所述保底带宽时,判定相应节点的带宽使用属性为“未达到保底”,并标记第四跑超带宽值,所述第四跑超带宽值根据以下表达式计算得到:
第四跑超带宽值=节点带宽预期值-标准计费带宽。
优选的,所述第二方案配置模块包括:
跑满调节单元,用于选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点;
跑高调节单元,用于选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点;
未达保底调节单元,用于选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点。
优选的,所述跑满调节单元包括:
第一级跑满调节子单元,用于参照所述网络区域-服务节点映射表,选取与所述“跑满”节点服务相同网络区域,且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑满”节点变为“跑高”节点;
第二级跑满调节子单元,用于在“未达到保底”节点的带宽全部填满至超过所述保底带宽,且仍存在“跑满”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值或者全部“跑满”节点变为“跑高”节点;
第三级跑满调节子单元,用于在“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值,且仍存在“跑满”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述第一浮动区间最大值或者全部“跑满”节点变为“跑高”节点;
跑满告警子单元,用于在“正常”节点全部填满至超过所述第一浮动区间最大值且仍存在“跑满”节点时,发送跑满告警。
优选的,所述跑高调节单元包括:
第一级跑高调节子单元,用于参照所述网络区域-服务节点映射表,选取与所述“跑高”节点服务相同网络区域、且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑高”节点变为“正常”节点;
第二级跑高调节子单元,用于在全部“未达到保底”节点的带宽全部填满至超过所述保底带宽且仍存在“跑高”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“跑低”节点全部填满至超过所述第一浮动区间最小值或者全部“跑高”节点变为“正常”节点;
第三级跑高调节子单元,用于在“跑低”节点全部填满至超过所述第一浮动区间最小值且仍存在“跑高”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述跑满门限值或全部“跑高”节点变为“正常”节点;
跑高告警子单元,用于直到“正常”节点全部填满至超过所述跑满门限值且仍存在“跑高”节点时,发送跑高告警。
优选的,所述未达保底调节单元包括:
第一级未达保底调节子单元,用于参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“正常”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底” 节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“正常”节点变为“跑低”;
第二级未达保底调节子单元,用于参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“跑低”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑低”节点带宽降至保底带宽。
优选的,该装置还包括:
第二方案评估模块,用于在所述第一覆盖方案调整得到所述第二覆盖方案过程中全部“跑低”节点带宽降至保底带宽且仍存在“未达到保底”节点时,对第二覆盖方案进行带宽成本浪费判定,具体用于:
确定所述第二覆盖方案下各个节点的带宽使用属性,
在评测周期内出现大于计费点数的第一跑超带宽值或第二跑超带宽值的指定节点时,判定为带宽跑满,发送告警,
在评测周期内出现绝对值大于计费点数的第四跑超带宽值时,判定为保底浪费,发送告警,
在评测周期内出现绝对值大于计费点数的第三跑超带宽值的指定节点时,判定为标准计费带宽设置过高,发送告警通知。
根据本发明的另一方面,提供了一种计算机可读存储介质,所述存储介质上存储有计算机程序,所述程序被处理器执行时实现上述方法的步骤。
根据本发明的另一方面,提供了一种计算机设备,包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序,所述处理器执行所述程序时实现上述方法的步骤。
本发明所公开的上述技术方案,对第一覆盖方案进行模拟运行,根据模拟结果预测节点带宽使用情况,进而对第一覆盖方案进行进一步调整,得到与带宽使用实际情况更契合的第二覆盖方案,实现了系统对带宽使用情况自动的判断和调整,解决了带宽配置调整人工操作效率低下、出错概率高的问题。且制定了标准计费带宽的生成和应用机制,解决了缺乏合理计费带宽标准确认机制的问题。在根据服务质量调度基础上,增加带宽成本计算的最优性价比调度方 法和系统,提供一种基于大数据分析的标准计费带宽计算方法,实现带宽跑高的自动调整和调度,确保带宽跑高、突增不会增加带宽成本。
参照附图来阅读对于示例性实施例的以下描述,本发明实施例的其他特性特征和优点将变得清晰。
附图说明
并入到说明书中并且构成说明书的一部分的附图示出了本发明的实施例,并且与描述一起用于解释本发明的原理。在这些附图中,类似的附图标记用于表示类似的要素。下面描述中的附图是本发明的一些实施例,而不是全部实施例。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,可以根据这些附图获得其他的附图。
图1示例性地示出了本发明的实施例一提供的一种智能调度方法的流程;
图2示例性地示出了节点IP信息表中涉及带宽的各参数值与节点带宽使用属性之间的关系;
图3示例性地示出了图1中步骤104的具体流程;
图4示例性地示出了图3中步骤301的具体流程;
图5示例性地示出了图3中步骤302的具体流程;
图6示例性地示出了图3中步骤303的具体流程;
图7示例性地示出了图1中步骤1的具体流程;
图8示例性地示出了本发明的实施例二提供的一种智能调度装置的结构;
图9示例性地示出了图8中节点带宽预测模块801的结构;
图10示例性地示出了图8中带宽使用情况评估模块802的结构;
图11示例性地示出了图8中第二方案配置模块803的结构;
图12示例性地示出了图11中跑满调节单元1101的结构;
图13示例性地示出了图11中跑高调节单元1102的结构;
图14示例性地示出了图11中未达保底调节单元1103的结构。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例是本发明一部分实施例,而不是全部的实施例。基于本发明中的实施例,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,都属于本发明保护的范围。需要说明的是,在不冲突的情况下,本申请中的实施例及实施例中的特征可以相互任意组合。
当前的智能调度系统存在如下问题:
a)当前智能调度系统根据可用性和服务质量进行自动调度,没有考虑带宽成本,在可用性容错和服务质量调优中,会优先选择服务质量最好的节点服务,容易发生成本浪费的情况。
b)基于成本的带宽调整需要人工判断并手动完成,判断、调整效率差,人工成本高。
c)带宽跑高调度只有在服务质量下降后才会自动调整,或者人工查看带宽跑满后手工调整,调整效率和效果差、成本高;且在跑高后和调整前的时间会使服务质量下降。
d)对计费带宽的标准制定缺乏一套可行机制。
为了解决上述问题,本发明的实施例提供了一种智能调度方法和装置,对第一覆盖方案进行模拟运行,根据模拟结果预测节点带宽使用情况,进而对第一覆盖方案进行进一步调整,得到与带宽使用实际情况更契合的第二覆盖方案,实现了系统对带宽使用情况自动的判断和调整,解决了带宽配置调整人工操作效率低下、出错概率高的问题。且制定了标准计费带宽的生成和应用机制,解决了缺乏合理计费带宽标准确认机制的问题。
首先结合附图,对本发明的实施例一进行说明。
本发明实施例提供了一种智能调度方法,使用该方法进行调度方案中带宽配置调整的流程如图1所示,包括:
步骤101、根据原始覆盖方案,结合IP可用性检测结果与各IP地址的服务质量,生成第一覆盖方案、网络区域-服务节点映射表、节点IP信息表和节点 带宽变化参数;
其中,网络区域-服务节点映射表为网络区域和服务质量达标的节点间的对应关系,
节点带宽变化参数为节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置,所述节点带宽变化参数根据以下表达式计算得到:
节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带宽+1,
节点IP信息表包含节点名称、保底带宽、标准计费带宽、出口带宽、未计费高点个数、计费点数、IP列表,所述保底带宽为根据付费情况节点最低应提供的带宽,所述标准计费带宽为基础带宽与增量带宽之和,所述基础带宽通过对上一个计费周期的同期带宽求平均值得到,所述出口带宽>=所述标准计费带宽>=所述保底带宽,所述标准计费带宽具有第一浮动区间,所述保底带宽小于所述第一浮动区间最小值,所述出口带宽大于所述第一浮动区间最大值。
图2为各参数值关系示意图。
步骤102、根据第一覆盖方案,确定节点带宽预期值;
本步骤中,首先对比所述原始覆盖方案和所述第一覆盖方案,根据对比结果,进行如下处理:
1、在所述原始覆盖方案和所述第一覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化时,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数,
所述指定节点即为所述原始覆盖方案及第一覆盖方案涉及的提供服务的全部节点。
2、在相较于所述原始覆盖方案,所述第一覆盖方案中指定节点内所有IP对应的网络区域配置发生了变化时,通过访问日志计算出指定网络区域的总带宽值,将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值,结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数。
步骤103、根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;
本发明实施例中,将带宽使用水平通过带宽使用属性体现,带宽使用属性有“跑满”、“跑高”、“跑低”、“未达到保底”四种取值。对四种取值含义说明如下:
1、在所述节点带宽预期值大于等于跑满门限值时,判定相应节点的带宽使用属性为“跑满”,并标记第一跑超带宽值,所述跑满门限值通过以下表达式计算得到:
跑满门限值=出口带宽*跑满系数,
所述第一跑超带宽值根据以下表达式计算得到:
第一跑超带宽值=节点带宽预期值-标准计费带宽;
2、在所述节点带宽预期值大于等于所述标准计费带宽的第一浮动区间最大值且小于所述跑满门限值时,判定相应节点的带宽使用属性为“跑高”,并标记第二跑超带宽值,所述第二跑超带宽值根据以下表达式计算得到:
第二跑超带宽值=节点带宽预期值-标准计费带宽;
3、在所述节点带宽预期值在所述第一浮动区间内时,判定相应节点的带宽使用属性为“正常”;
4、在所述节点带宽预期值大于等于所述保底带宽且小于所述第一浮动区间最小值时,判定相应节点的带宽使用属性为“跑低”,并标记第三跑超带宽值,所述第三跑超带宽值根据以下表达式计算得到:
第三跑超带宽值=节点带宽预期值-标准计费带宽,
第三跑超带宽值为负值;
5、在所述节点带宽预期值小于所述保底带宽时,判定相应节点的带宽使用属性为“未达到保底”,并标记第四跑超带宽值,所述第四跑超带宽值根据以下表达式计算得到:
第四跑超带宽值=节点带宽预期值-标准计费带宽,
第四跑超带宽值为负值。
步骤104、根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;
本步骤具体如图3所示,包括:
步骤301、选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点;
本步骤具体如图4所示,包括:
步骤401、参照所述网络区域-服务节点映射表,选取与所述“跑满”节点服务相同网络区域,且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑满”节点变为“跑高”节点;
步骤402、在“未达到保底”节点的带宽全部填满至超过所述保底带宽,且仍存在“跑满”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值或者全部“跑满”节点变为“跑高”节点;
步骤403、在“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值,且仍存在“跑满”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述第一浮动区间最大值或者全部“跑满”节点变为“跑高”节点;
步骤404、在“正常”节点全部填满至超过所述第一浮动区间最大值且仍存在“跑满”节点时,发送跑满告警。
步骤302、选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点;
本步骤具体如图5所示,包括:
步骤501、参照所述网络区域-服务节点映射表,选取与所述“跑高”节点服务相同网络区域、且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑高”节点变为 “正常”节点;
步骤502、在全部“未达到保底”节点的带宽全部填满至超过所述保底带宽且仍存在“跑高”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“跑低”节点全部填满至超过所述第一浮动区间最小值或者全部“跑高”节点变为“正常”节点;
步骤503、在“跑低”节点全部填满至超过所述第一浮动区间最小值且仍存在“跑高”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述跑满门限值或全部“跑高”节点变为“正常”节点;
步骤504、直到“正常”节点全部填满至超过所述跑满门限值且仍存在“跑高”节点时,发送跑高告警。
步骤303、选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点;
本步骤具体如图6所示,包括:
步骤601、参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“正常”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“正常”节点变为“跑低”;
步骤602、参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“跑低”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑低”节点带宽降至保底带宽。
步骤105、根据所述第二覆盖方案,对节点带宽资源进行调度。
步骤106、在所述第一覆盖方案调整得到所述第二覆盖方案过程中全部“跑低”节点带宽降至保底带宽且仍存在“未达到保底”节点时,对第二覆盖方案进行带宽成本浪费判定;
本步骤具体如图7所示,包括:
步骤701、确定所述第二覆盖方案下各个节点的带宽使用属性;
步骤702、在评测周期内出现大于计费点数的第一跑超带宽值或第二跑超带宽值的指定节点时,判定为带宽跑满,发送告警;
步骤703、在评测周期内出现绝对值大于计费点数的第四跑超带宽值时,判定为保底浪费,发送告警;
步骤704、在评测周期内出现绝对值大于计费点数的第三跑超带宽值的指定节点时,判定为标准计费带宽设置过高,发送告警通知。
下面结合附图,对本发明的实施例二进行说明。
本发明实施例提供了一种智能调度装置,其结构如图8所示,包括:
节点带宽预测模块801,用于根据第一覆盖方案,确定节点带宽预期值;
带宽使用情况评估模块802,用于根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;
第二方案配置模块803,用于根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;
资源调度模块804,用于根据所述第二覆盖方案,对节点带宽资源进行调度。
优选的,该装置还包括:
第一方案配置模块805,用于根据原始覆盖方案,结合IP可用性检测结果与各IP地址的服务质量,生成第一覆盖方案、网络区域-服务节点映射表、节点IP信息表和节点带宽变化参数,
其中,
网络区域-服务节点映射表为网络区域和服务质量达标的节点间的对应关系,
节点带宽变化参数为节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置,所述节点带宽变化参数根据以下表达式计算得到:
节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带 宽+1,
节点IP信息表包含节点名称、保底带宽、标准计费带宽、出口带宽、未计费高点个数、计费点数、IP列表,所述保底带宽为根据付费情况节点最低应提供的带宽,所述标准计费带宽为基础带宽与增量带宽之和,所述基础带宽通过对上一个计费周期的同期带宽求平均值得到,所述出口带宽>=所述标准计费带宽>=所述保底带宽,所述标准计费带宽具有第一浮动区间,所述保底带宽小于所述第一浮动区间最小值,所述出口带宽大于所述第一浮动区间最大值。
优选的,所述节点带宽预测模块801的结构如图9所示,包括:
变化检测单元901,用于对比所述原始覆盖方案和所述第一覆盖方案;
第一带宽计算单元902,用于在所述原始覆盖方案和所述第一覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化时,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
第二带宽计算单元903,用于在相较于所述原始覆盖方案,所述第一覆盖方案中指定节点内所有IP对应的网络区域配置发生了变化时,通过访问日志计算出指定网络区域的总带宽值,将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值,结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值,根据以下表达式计算节点带宽预期值:
节点带宽预期值=节点当前的带宽值*节点带宽变化参数。
优选的,所述带宽使用情况评估模块802的结构如图10所示,包括:
第一评估单元1001,用于在所述节点带宽预期值大于等于跑满门限值时,判定相应节点的带宽使用属性为“跑满”,并标记第一跑超带宽值,所述跑满门限值通过以下表达式计算得到:
跑满门限值=出口带宽*跑满系数,
所述第一跑超带宽值根据以下表达式计算得到:
第一跑超带宽值=节点带宽预期值-标准计费带宽;
第二评估单元1002,用于在所述节点带宽预期值大于等于所述标准计费带 宽的第一浮动区间最大值且小于所述跑满门限值时,判定相应节点的带宽使用属性为“跑高”,并标记第二跑超带宽值,所述第二跑超带宽值根据以下表达式计算得到:
第二跑超带宽值=节点带宽预期值-标准计费带宽;
在所述节点带宽预期值在所述第一浮动区间内时,判定相应节点的带宽使用属性为“正常”;
第三评估单元1003,用于在所述节点带宽预期值大于等于所述保底带宽且小于所述第一浮动区间最小值时,判定相应节点的带宽使用属性为“跑低”,并标记第三跑超带宽值,所述第三跑超带宽值根据以下表达式计算得到:
第三跑超带宽值=节点带宽预期值-标准计费带宽;
第四评估单元1004,用于在所述节点带宽预期值小于所述保底带宽时,判定相应节点的带宽使用属性为“未达到保底”,并标记第四跑超带宽值,所述第四跑超带宽值根据以下表达式计算得到:
第四跑超带宽值=节点带宽预期值-标准计费带宽。
优选的,所述第二方案配置模块803的结构如图11所示,包括:
跑满调节单元1101,用于选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点;
跑高调节单元1102,用于选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点;
未达保底调节单元1103,用于选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点。
优选的,所述跑满调节单元1101的结构如图12所示,包括:
第一级跑满调节子单元1201,用于参照所述网络区域-服务节点映射表,选取与所述“跑满”节点服务相同网络区域,且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑满”节点变为“跑高”节点;
第二级跑满调节子单元1202,用于在“未达到保底”节点的带宽全部填满 至超过所述保底带宽,且仍存在“跑满”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值或者全部“跑满”节点变为“跑高”节点;
第三级跑满调节子单元1203,用于在“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值,且仍存在“跑满”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述第一浮动区间最大值或者全部“跑满”节点变为“跑高”节点;
跑满告警子单元1204,用于在“正常”节点全部填满至超过所述第一浮动区间最大值且仍存在“跑满”节点时,发送跑满告警。
优选的,所述跑高调节单元1102的结构如图13所示,包括:
第一级跑高调节子单元1301,用于参照所述网络区域-服务节点映射表,选取与所述“跑高”节点服务相同网络区域、且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑高”节点变为“正常”节点;
第二级跑高调节子单元1302,用于在全部“未达到保底”节点的带宽全部填满至超过所述保底带宽且仍存在“跑高”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“跑低”节点全部填满至超过所述第一浮动区间最小值或者全部“跑高”节点变为“正常”节点;
第三级跑高调节子单元1303,用于在“跑低”节点全部填满至超过所述第一浮动区间最小值且仍存在“跑高”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述跑满门限值或全部“跑高”节点变为“正常”节点;
跑高告警子单元1304,用于直到“正常”节点全部填满至超过所述跑满门限值且仍存在“跑高”节点时,发送跑高告警。
优选的,所述未达保底调节单元1103的结构如图14所示,包括:
第一级未达保底调节子单元1401,用于参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“正常”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“正常”节点变为“跑低”;
第二级未达保底调节子单元1402,用于参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“跑低”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑低”节点带宽降至保底带宽。
优选的,该装置还包括:
第二方案评估模块806,用于在所述第一覆盖方案调整得到所述第二覆盖方案过程中全部“跑低”节点带宽降至保底带宽且仍存在“未达到保底”节点时,对第二覆盖方案进行带宽成本浪费判定,具体用于:
确定所述第二覆盖方案下各个节点的带宽使用属性,
在评测周期内出现大于计费点数的第一跑超带宽值或第二跑超带宽值的指定节点时,判定为带宽跑满,发送告警,
在评测周期内出现绝对值大于计费点数的第四跑超带宽值时,判定为保底浪费,发送告警,
在评测周期内出现绝对值大于计费点数的第三跑超带宽值的指定节点时,判定为标准计费带宽设置过高,发送告警通知。
下面结合附图,对本发明的实施例三进行说明。
本发明实施例提供了一种智能调度方法,接收第一覆盖方案、网络区域-服务节点映射表和节点IP信息表,依次进行保底带宽、带宽跑高、标准计费模型的运行,确保每个节点的计费带宽大于保底带宽、跑高带宽调度到未达到标准计费带宽的节点或使用剩余未计费高点数节点服务、确保各节点计费带宽在标准计费带宽上下5%之内(5%为经验值,可根据实际情况调整),达到成本控 制的目的。
1、带宽成本的关键点:
i.保底带宽:不管是否使用,都要付费的带宽,所以这部分应该优先使用
ii.计费方式:节点带宽的计费方式,一般都是月度计费,计费方式有去三峰、95计费等。
A.去三峰为每5分钟一个带宽点,每天一个峰值,按照第四高的峰值付费;
B.95计费为每5分钟一个带宽点,按照第95%的峰值计费,也就是433个高点不会被计费;
C.每个节点带宽跑平均,且计费带宽在总出口带宽的60%-80%为最优性价比的模型;进一步合理应用不计费的高点来服务流量突增和跑高,就能达到最佳性价比的目标。
iii.带宽跑满:
节点带宽在超过出口带宽91%(经验值)的时候,会影响服务质量,所以当节点带宽超过91%时,需要进行调整
2、关键数据和文件:
i.原始覆盖方案:覆盖方案为网络区域和节点IP的对应关系,本发明实施例中是指当前调度模块正在使用的覆盖方案,为第一覆盖方案基于可用性和服务质量计算前的原始版本;
ii.第一覆盖方案:覆盖方案为网络区域和节点IP的对应关系,本发明实施例中是指基于IP可用性和服务质量运算得出的覆盖方案
例如:
联通北京1.1.1.1、1.1.1.2
电信上海2.2.2.1、2.2.2.2
iii.网络区域-服务节点映射表:网络区域和服务质量达标的节点间的对应关系,本发明实施例中是指基于IP可用性和服务质量运算得出的服务某个网络区域达到服务质量标准的节点列表;
例如:
联通北京node1、node2
电信上海node3、node4
iv.节点带宽变化参数:节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置
节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带宽+1
v.节点IP信息表:节点的基本信息,包括节点名称、保底带宽、标准计费带宽、出口带宽、未计费高点个数、计费点数、IP列表
A.保底带宽:单位M,节点采购时确定的值
B.标准计费带宽:单位M,由历史带宽数据分析和人工配置设置
C.出口带宽:单位M,节点采购时确定的值
D.未计费高点个数:单位个,每天可用的非计费高点个数,当天没有用完可以累加到下一天,超标使用需要在当月剩余天数里扣除。
E.计费点数:每天最少达到“正常“的点数
F.IP列表:节点内IP的列表
例如:
node1 1000 6000 10000 433 433 1.1.1.1、1.1.1.2
业务流程如下:
1、获取第一覆盖方案、网络区域-服务节点映射表、节点IP信息表、节点带宽变化参数;模拟使用第一覆盖方案进行调度,计算各个节点的带宽预期值。
i.获取原始覆盖方案、第一覆盖方案、网络区域-服务节点映射表和节点IP信息表、节点带宽变化参数;
A.节点IP信息表中的标准计费带宽,由基础带宽和增量带宽两部分求和得出;
B.基础带宽通过对上一个计费周期的同期带宽求平均值,得到响应的值,上一个计费周期的同期带宽为星期数相同的天记为同期,带宽为当天的计费带宽。例如上个月所有的周二求平均值,为现有计费周期周二的基础带宽;
C.增量带宽由资源管理员根据项目情况,人工设置;
ii.获取各个节点的带宽预期值:
A.对比原始覆盖方案和第一覆盖方案,如果覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化,则直接获取该节点当前的带宽值,节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
B.如果覆盖方案中指定节点内所有IP对应的网络区域配置发生变化,则节点当前的带宽值计算方式为:则通过访问日志计算出指定网络区域的总带宽值;并将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值;结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值;
节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
2、判断在第一覆盖方案调度下,是否存在带宽成本浪费的情况
i.判断节点带宽预期值是否大于节点IP信息表中的出口带宽*0.91;如果大于则将该节点属性标为“跑满“,并标记第一跑超带宽值,第一跑超带宽值为“节点带宽预期值-标准计费带宽”;
ii.判断节点带宽预期值是否大于节点IP信息表中的标准计费带宽,大于的幅度超过5%(即1.05位的出口带宽)且小于跑满门限值(跑满门限值为出口带宽*0.91);如果满足,则将该节点属性标为“跑高”,第二跑超带宽值为“节点带宽预期值-标准计费带宽”。
iii.判断节点带宽预期值是否低于节点IP信息表中的保底带宽,如果满足,则将该节点属性标为“未达到保底”,第四跑超带宽值为“节点带宽预期值-保底带宽”。
iv.判断节点带宽预期值是否低于节点IP信息表中的标准计费带宽(超过5%)且高于保底带宽,如果满足,则该改节点属性标为“跑低”,第三跑超带宽值为“节点带宽预期值-标准计费带宽”。
v.判断节点带宽预期值大于保底带宽,且与标准计费带宽差异在正负5%以内(第一浮动区间的范围为标准计费带宽上下浮动5%),如果满足,则将该节点标为正常。
vi.将所有节点判断结果,形成带宽情况判定列表,格式为“节点名属性第N跑超带宽”。
3、调整第一覆盖方案,形成第二覆盖方案,解决带宽成本浪费的问题。
i.选取属性为“跑满”的节点信息,通过修改覆盖方案,将跑超值的带宽调度到非跑满节点;
A.参照网络区域-服务节点映射表,选取与所述跑满节点服务相同网络区域,且节点属性为“未达到保底”的节点作为备份节点,在覆盖方案里通过将部分跑满节点的IP替换成备份节点IP的方式,实现将带宽从跑满节点调度到未达到保底的节点,直到保底全部填满或者跑满节点变为“跑高”;
B.如果全部可选未达到保底节点均已经跑到保底,且节点仍为跑满状态,则继续选用跑低节点作为备份节点,进行调度,直到可选跑低节点全部变为正常或跑满节点变为“跑高”为止;
C.如果全部可选跑低节点均已经跑到正常,且节点仍为跑满状态,则选用可选的正常节点中未计费高点个数最多的节点作为备份节点,将跑超值的带宽调度到备份节点,直到正常节点跑至带宽出口的91%或跑满节点变为“跑高”为止;
D.如果全部可选正常节点均跑至带宽出口91%,且节点仍为跑满状态,则发送跑满告警给资源管理员,人工介入处理。
ii.选取属性为“跑高”的节点信息,通过修改覆盖方案,将跑超值的带宽调度到非跑满、跑高节点;
A.参照网络区域-服务节点映射表,选取与所述跑高节点服务相同网络区域、且节点属性为“未达到保底”的节点作为备份节点,在覆盖方案里通过将部分跑满节点的IP替换成备份节点IP的方式,实现将带宽从跑满节点调度到未达到保底的节点,直到保底全部填满或者跑高节点变为“正常”;
B.如果全部可选未达到保底节点均已经跑到保底,且节点仍为跑高状态,则继续选用跑低节点作为备份节点,进行调度,直到可选跑低节点全部变为正常或跑高节点变为“正常”为止;
C.如果全部可选跑低节点均已经跑到正常,且节点仍为跑高状态,则选用可选的正常节点中未计费高点个数最多的节点作为备份节点,将跑超值的带宽 调度到备份节点,直到正常节点跑至带宽出口的91%或跑高节点变为“正常”为止;
D.如果全部可选正常节点均跑至带宽出口91%,且节点仍为跑高状态,则发送跑高告警给资源管理员,人工介入处理。
iii.选取属性为“未达到保底”的节点信息,通过修改覆盖方案,将达到保底的节点带宽调度到未达到保底的节点;
A.参照网络区域-服务节点映射表,选取与所述未达到保底的节点服务相同网络区域、且节点属性为“正常”的节点作为源节点,在覆盖方案里通过将部分源节点的IP替换成未达到保底节点IP的方式,实现将带宽从正常属性节点调度到未达到保底的节点,直到保底全部填满或者正常节点变为“跑低”;
B.参照网络区域-服务节点映射表,选取与所述未达到保底的节点服务相同网络区域、且节点属性为“跑低”的节点作为源节点,在覆盖方案里通过将部分源节点的IP替换成未达到保底节点IP的方式,实现将带宽从跑低属性节点调度到未达到保底的节点,直到保底全部填满或者跑低节点带宽降至保底带宽值;
C.如果跑低节点带宽降至保底带宽值但节点仍未达到保底,则继续执行下一步判定。
4、判定第二覆盖方案是否存在带宽成本浪费的情况;
i.同上方法,获取基于第二覆盖方案的各个节点的带宽预期值,并将将所有节点判断结果,形成带宽情况判定列表,格式为“节点名属性跑超带宽”;
ii.判断近24小时,指定节点有大于计费点数的跑高或跑满带宽值,则判定为带宽跑满,发送告警通知资源管理员接入;
iii.判断近24小时,指定节点有大于计费点数的未达到保底的带宽值,则判定为保底浪费,发送告警通知资源管理员接入;
iv.判断近24小时,指定节点有大于计费点数的跑低带宽值,则判定为“标准计费带宽”设置过高,发送告警通知资源管理员接入;
5、将第二覆盖方案发送给调度模块,进行调度服务。将用户访问引导至指定的服务IP。
本发明的实施例还提供了一种计算机可读存储介质,此存储介质上存储有 计算机程序,所述程序被处理器执行时实现上述方法的步骤。
本发明的实施例还提供了一种计算机设备,包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序,处理器执行此程序时实现上述方法的步骤。
本发明所公开的上述技术方案,对第一覆盖方案进行模拟运行,根据模拟结果预测节点带宽使用情况,进而对第一覆盖方案进行进一步调整,得到与带宽使用实际情况更契合的第二覆盖方案,实现了系统对带宽使用情况自动的判断和调整,解决了带宽配置调整人工操作效率低下、出错概率高的问题。且制定了标准计费带宽的生成和应用机制,解决了缺乏合理计费带宽标准确认机制的问题。在根据服务质量调度基础上,增加带宽成本计算的最优性价比调度方法和系统,提供一种基于大数据分析的标准计费带宽计算方法,实现带宽跑高的自动调整和调度,确保带宽跑高、突增不会增加带宽成本。
上面描述的内容可以单独地或者以各种方式组合起来实施,而这些变型方式都在本发明的保护范围之内。
本领域普通技术人员可以理解上述方法中的全部或部分步骤可通过程序来指令相关硬件完成,所述程序可以存储于计算机可读存储介质中,如只读存储器、磁盘或光盘等。可选地,上述实施例的全部或部分步骤也可以使用一个或多个集成电路来实现,相应地,上述实施例中的各模块/单元可以采用硬件的形式实现,也可以采用软件功能模块的形式实现。本发明不限制于任何特定形式的硬件和软件的结合。
需要说明的是,在本文中,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物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物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……”限定的要素,并不排除在包括所述要素的物品或者设备中还存在另外的相同要素。
以上实施例仅用以说明本发明的技术方案而非限制,仅仅参照较佳实施例对本发明进行了详细说明。本领域的普通技术人员应当理解,可以对本发明的技术方案进行修改或者等同替换,而不脱离本发明技术方案的精神和范围,均应涵盖在本发明的权利要求范围当中。
工业实用性
本发明提供了一种智能调度方法和装置及其计算机可读存储介质和计算机设备,实现了对带宽使用情况自动的判断和调整,解决了带宽配置调整人工操作效率低下、出错概率高的问题。且制定了标准计费带宽的生成和应用机制,解决了缺乏合理计费带宽标准确认机制的问题。在根据服务质量调度基础上,增加带宽成本计算的最优性价比调度方法和系统,提供一种基于大数据分析的标准计费带宽计算方法,实现带宽跑高的自动调整和调度,确保带宽跑高、突增不会增加带宽成本。

Claims (20)

  1. 一种智能调度方法,其特征在于,包括:
    根据第一覆盖方案,确定节点带宽预期值;
    根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;
    根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;
    根据所述第二覆盖方案,对节点带宽资源进行调度。
  2. 根据权利要求1所述的智能调度方法,其特征在于,根据第一覆盖方案,确定节点带宽预期值的步骤之前还包括:
    根据原始覆盖方案,结合IP可用性检测结果与各IP地址的服务质量,生成第一覆盖方案、网络区域-服务节点映射表、节点IP信息表和节点带宽变化参数,
    其中,
    网络区域-服务节点映射表为网络区域和服务质量达标的节点间的对应关系,
    节点带宽变化参数为节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置,所述节点带宽变化参数根据以下表达式计算得到:
    节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带宽+1,
    节点IP信息表包含节点名称、保底带宽、标准计费带宽、出口带宽、未计费高点个数、计费点数、IP列表,所述保底带宽为根据付费情况节点最低应提供的带宽,所述标准计费带宽为基础带宽与增量带宽之和,所述基础带宽通过对上一个计费周期的同期带宽求平均值得到,所述出口带宽>=所述标准计费带宽>=所述保底带宽,所述标准计费带宽具有第一浮动区间,所述保底带宽小于所述第一浮动区间最小值,所述出口带宽大于所述第一浮动区间最大值。
  3. 根据权利要求2所述的智能调度方法,其特征在于,根据第一覆盖方案, 确定节点带宽预期值的步骤包括:
    对比所述原始覆盖方案和所述第一覆盖方案;
    在所述原始覆盖方案和所述第一覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化时,根据以下表达式计算节点带宽预期值:
    节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
    在相较于所述原始覆盖方案,所述第一覆盖方案中指定节点内所有IP对应的网络区域配置发生了变化时,通过访问日志计算出指定网络区域的总带宽值,将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值,结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值,根据以下表达式计算节点带宽预期值:
    节点带宽预期值=节点当前的带宽值*节点带宽变化参数。
  4. 根据权利要求3所述的智能调度方法,其特征在于,根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平的步骤包括:
    在所述节点带宽预期值大于等于跑满门限值时,判定相应节点的带宽使用属性为“跑满”,并标记第一跑超带宽值,所述跑满门限值通过以下表达式计算得到:
    跑满门限值=出口带宽*跑满系数,
    所述第一跑超带宽值根据以下表达式计算得到:
    第一跑超带宽值=节点带宽预期值-标准计费带宽;
    在所述节点带宽预期值大于等于所述标准计费带宽的第一浮动区间最大值且小于所述跑满门限值时,判定相应节点的带宽使用属性为“跑高”,并标记第二跑超带宽值,所述第二跑超带宽值根据以下表达式计算得到:
    第二跑超带宽值=节点带宽预期值-标准计费带宽;
    在所述节点带宽预期值在所述第一浮动区间内时,判定相应节点的带宽使用属性为“正常”;
    在所述节点带宽预期值大于等于所述保底带宽且小于所述第一浮动区间最小值时,判定相应节点的带宽使用属性为“跑低”,并标记第三跑超带宽值,所述第三跑超带宽值根据以下表达式计算得到:
    第三跑超带宽值=节点带宽预期值-标准计费带宽;
    在所述节点带宽预期值小于所述保底带宽时,判定相应节点的带宽使用属性为“未达到保底”,并标记第四跑超带宽值,所述第四跑超带宽值根据以下表达式计算得到:
    第四跑超带宽值=节点带宽预期值-标准计费带宽。
  5. 根据权利要求4所述的智能调度方法,其特征在于,根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案的步骤包括:
    选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点;
    选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点;
    选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点。
  6. 根据权利要求5所述的智能调度方法,其特征在于,选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点的步骤包括:
    参照所述网络区域-服务节点映射表,选取与所述“跑满”节点服务相同网络区域,且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑满”节点变为“跑高”节点;
    在“未达到保底”节点的带宽全部填满至超过所述保底带宽,且仍存在“跑满”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值或者全部“跑满”节点变为“跑高”节点;
    在“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值,且仍存在“跑满”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到 “正常”节点全部填满至超过所述第一浮动区间最大值或者全部“跑满”节点变为“跑高”节点;
    在“正常”节点全部填满至超过所述第一浮动区间最大值且仍存在“跑满”节点时,发送跑满告警。
  7. 根据权利要求5所述的智能调度方法,其特征在于,选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点的步骤包括:
    参照所述网络区域-服务节点映射表,选取与所述“跑高”节点服务相同网络区域、且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑高”节点变为“正常”节点;
    在全部“未达到保底”节点的带宽全部填满至超过所述保底带宽且仍存在“跑高”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“跑低”节点全部填满至超过所述第一浮动区间最小值或者全部“跑高”节点变为“正常”节点;
    在“跑低”节点全部填满至超过所述第一浮动区间最小值且仍存在“跑高”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述跑满门限值或全部“跑高”节点变为“正常”节点;
    直到“正常”节点全部填满至超过所述跑满门限值且仍存在“跑高”节点时,发送跑高告警。
  8. 根据权利要求5所述的智能调度方法,其特征在于,选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点的步骤包括:
    参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“正常”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“正常”节点变为“跑低”;
    参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相 同网络区域且带宽使用属性为“跑低”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑低”节点带宽降至保底带宽。
  9. 根据权利要求5-8任一所述的智能调度方法,其特征在于,该方法还包括:
    在所述第一覆盖方案调整得到所述第二覆盖方案过程中全部“跑低”节点带宽降至保底带宽且仍存在“未达到保底”节点时,对第二覆盖方案进行带宽成本浪费判定,具体包括:
    确定所述第二覆盖方案下各个节点的带宽使用属性;
    在评测周期内出现大于计费点数的第一跑超带宽值或第二跑超带宽值的指定节点时,判定为带宽跑满,发送告警;
    在评测周期内出现绝对值大于计费点数的第四跑超带宽值时,判定为保底浪费,发送告警;
    在评测周期内出现绝对值大于计费点数的第三跑超带宽值的指定节点时,判定为标准计费带宽设置过高,发送告警通知。
  10. 一种智能调度装置,其特征在于,包括:
    节点带宽预测模块,用于根据第一覆盖方案,确定节点带宽预期值;
    带宽使用情况评估模块,用于根据所述节点带宽预期值,判定在第一覆盖方案下,各节点的带宽使用水平;
    第二方案配置模块,用于根据对节点的带宽使用水平的判定结果,调整各节点承载的带宽配置,得到第二覆盖方案;
    资源调度模块,用于根据所述第二覆盖方案,对节点带宽资源进行调度。
  11. 根据权利要求10所述的智能调度装置,其特征在于,该装置还包括:
    第一方案配置模块,用于根据原始覆盖方案,结合IP可用性检测结果与各IP地址的服务质量,生成第一覆盖方案、网络区域-服务节点映射表、节点IP信息表和节点带宽变化参数,
    其中,
    网络区域-服务节点映射表为网络区域和服务质量达标的节点间的对应关系,
    节点带宽变化参数为节点下一个时间点带宽与当前值对比的变化率,节点带宽下一个时间点的带宽值为历史数据的平均值或手动配置,所述节点带宽变化参数根据以下表达式计算得到:
    节点带宽变化参数=(节点下一个时间点带宽-节点当前带宽)/节点当前带宽+1,
    节点IP信息表包含节点名称、保底带宽、标准计费带宽、出口带宽、未计费高点个数、计费点数、IP列表,所述保底带宽为根据付费情况节点最低应提供的带宽,所述标准计费带宽为基础带宽与增量带宽之和,所述基础带宽通过对上一个计费周期的同期带宽求平均值得到,所述出口带宽>=所述标准计费带宽>=所述保底带宽,所述标准计费带宽具有第一浮动区间,所述保底带宽小于所述第一浮动区间最小值,所述出口带宽大于所述第一浮动区间最大值。
  12. 根据权利要求11所述的智能调度装置,其特征在于,所述节点带宽预测模块包括:
    变化检测单元,用于对比所述原始覆盖方案和所述第一覆盖方案;
    第一带宽计算单元,用于在所述原始覆盖方案和所述第一覆盖方案中指定节点内所有IP对应的网络区域配置均没有变化时,根据以下表达式计算节点带宽预期值:
    节点带宽预期值=节点当前的带宽值*节点带宽变化参数;
    第二带宽计算单元,用于在相较于所述原始覆盖方案,所述第一覆盖方案中指定节点内所有IP对应的网络区域配置发生了变化时,通过访问日志计算出指定网络区域的总带宽值,将网络区域总带宽值,按照网络区域配置各个服务IP的解析比例关系,分配给每个服务IP,得到各个服务IP在指定网络区域的带宽值,结合节点IP信息表,将指定节点内的服务IP在各个网络区域内的带宽值相加,得到节点当前的带宽值,根据以下表达式计算节点带宽预期值:
    节点带宽预期值=节点当前的带宽值*节点带宽变化参数。
  13. 根据权利要求12所述的智能调度装置,其特征在于,所述带宽使用情况评估模块包括:
    第一评估单元,用于在所述节点带宽预期值大于等于跑满门限值时,判定相应节点的带宽使用属性为“跑满”,并标记第一跑超带宽值,所述跑满门限 值通过以下表达式计算得到:
    跑满门限值=出口带宽*跑满系数,
    所述第一跑超带宽值根据以下表达式计算得到:
    第一跑超带宽值=节点带宽预期值-标准计费带宽;
    第二评估单元,用于在所述节点带宽预期值大于等于所述标准计费带宽的第一浮动区间最大值且小于所述跑满门限值时,判定相应节点的带宽使用属性为“跑高”,并标记第二跑超带宽值,所述第二跑超带宽值根据以下表达式计算得到:
    第二跑超带宽值=节点带宽预期值-标准计费带宽;
    在所述节点带宽预期值在所述第一浮动区间内时,判定相应节点的带宽使用属性为“正常”;
    第三评估单元,用于在所述节点带宽预期值大于等于所述保底带宽且小于所述第一浮动区间最小值时,判定相应节点的带宽使用属性为“跑低”,并标记第三跑超带宽值,所述第三跑超带宽值根据以下表达式计算得到:
    第三跑超带宽值=节点带宽预期值-标准计费带宽;
    第四评估单元,用于在所述节点带宽预期值小于所述保底带宽时,判定相应节点的带宽使用属性为“未达到保底”,并标记第四跑超带宽值,所述第四跑超带宽值根据以下表达式计算得到:
    第四跑超带宽值=节点带宽预期值-标准计费带宽。
  14. 根据权利要求13所述的智能调度装置,其特征在于,所述第二方案配置模块包括:
    跑满调节单元,用于选取带宽使用属性为“跑满”的节点,将第一跑超带宽值的带宽调度到非“跑满”节点;
    跑高调节单元,用于选取带宽使用属性为“跑高”的节点,将第二跑超带宽值的带宽调度到非“跑满”且非“跑高”节点;
    未达保底调节单元,用于选取带宽使用属性为“未达到保底”的节点,将非“未达到保底”的节点的部分带宽调度到该“未达到保底”的节点。
  15. 根据权利要求14所述的智能调度装置,其特征在于,所述跑满调节单元包括:
    第一级跑满调节子单元,用于参照所述网络区域-服务节点映射表,选取与所述“跑满”节点服务相同网络区域,且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑满”节点变为“跑高”节点;
    第二级跑满调节子单元,用于在“未达到保底”节点的带宽全部填满至超过所述保底带宽,且仍存在“跑满”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值或者全部“跑满”节点变为“跑高”节点;
    第三级跑满调节子单元,用于在“跑低”节点的带宽全部填满至超过所述第一浮动区间最小值,且仍存在“跑满”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑满”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述第一浮动区间最大值或者全部“跑满”节点变为“跑高”节点;
    跑满告警子单元,用于在“正常”节点全部填满至超过所述第一浮动区间最大值且仍存在“跑满”节点时,发送跑满告警。
  16. 根据权利要求14所述的智能调度装置,其特征在于,所述跑高调节单元包括:
    第一级跑高调节子单元,用于参照所述网络区域-服务节点映射表,选取与所述“跑高”节点服务相同网络区域、且带宽使用属性为“未达到保底”的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑高”节点变为“正常”节点;
    第二级跑高调节子单元,用于在全部“未达到保底”节点的带宽全部填满至超过所述保底带宽且仍存在“跑高”节点时,选用“跑低”节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“跑低”节点全部填满至超过所述第一浮动区间最小值或者全部“跑高”节点变为“正常”节点;
    第三级跑高调节子单元,用于在“跑低”节点全部填满至超过所述第一浮 动区间最小值且仍存在“跑高”节点时,选用“正常”节点中未计费高点个数最多的节点作为备份节点,将所述第一覆盖方案中部分“跑高”节点的IP替换成备份节点IP,直到“正常”节点全部填满至超过所述跑满门限值或全部“跑高”节点变为“正常”节点;
    跑高告警子单元,用于直到“正常”节点全部填满至超过所述跑满门限值且仍存在“跑高”节点时,发送跑高告警。
  17. 根据权利要求14所述的智能调度装置,其特征在于,所述未达保底调节单元包括:
    第一级未达保底调节子单元,用于参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“正常”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“正常”节点变为“跑低”;
    第二级未达保底调节子单元,用于参照所述网络区域-服务节点映射表,选取与“未达到保底”节点服务相同网络区域且带宽使用属性为“跑低”的节点作为源节点,将所述第一覆盖方案中部分源节点的IP替换成“未达到保底”节点IP,直到“未达到保底”节点的带宽全部填满至超过所述保底带宽或者全部“跑低”节点带宽降至保底带宽。
  18. 根据权利要求14-17任一所述的智能调度装置,其特征在于,该装置还包括:
    第二方案评估模块,用于在所述第一覆盖方案调整得到所述第二覆盖方案过程中全部“跑低”节点带宽降至保底带宽且仍存在“未达到保底”节点时,对第二覆盖方案进行带宽成本浪费判定,具体用于:
    确定所述第二覆盖方案下各个节点的带宽使用属性,
    在评测周期内出现大于计费点数的第一跑超带宽值或第二跑超带宽值的指定节点时,判定为带宽跑满,发送告警,
    在评测周期内出现绝对值大于计费点数的第四跑超带宽值时,判定为保底浪费,发送告警,
    在评测周期内出现绝对值大于计费点数的第三跑超带宽值的指定节点时, 判定为标准计费带宽设置过高,发送告警通知。
  19. 一种计算机可读存储介质,所述存储介质上存储有计算机程序,所述程序被处理器执行时实现权利要求1至9中任意一项所述方法的步骤。
  20. 一种计算机设备,包括存储器、处理器及存储在所述存储器上并可在所述处理器上运行的计算机程序,所述处理器执行所述程序时实现权利要求1至9中任意一项所述方法的步骤。
PCT/CN2018/100557 2017-08-15 2018-08-15 一种智能调度方法和装置及其计算机可读存储介质和计算机设备 WO2019034065A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710695232.7A CN107707382B (zh) 2017-08-15 2017-08-15 一种智能调度方法和装置
CN201710695232.7 2017-08-15

Publications (1)

Publication Number Publication Date
WO2019034065A1 true WO2019034065A1 (zh) 2019-02-21

Family

ID=61171032

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/100557 WO2019034065A1 (zh) 2017-08-15 2018-08-15 一种智能调度方法和装置及其计算机可读存储介质和计算机设备

Country Status (2)

Country Link
CN (1) CN107707382B (zh)
WO (1) WO2019034065A1 (zh)

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107707382B (zh) * 2017-08-15 2018-10-23 贵州白山云科技有限公司 一种智能调度方法和装置
CN110493046B (zh) * 2018-02-27 2022-08-16 贵州白山云科技股份有限公司 一种cdn网络中节点服务器带宽资源的调度方法和系统
CN110381170A (zh) * 2018-02-27 2019-10-25 贵州白山云科技股份有限公司 一种cdn网络中调度节点服务器的业务的方法和系统
CN109067926A (zh) * 2018-10-11 2018-12-21 平安科技(深圳)有限公司 负载均衡方法、装置、计算机设备及存储介质
CN109412977B (zh) * 2018-10-18 2022-05-17 网宿科技股份有限公司 一种域名带宽调节方法及相关设备
CN111464323A (zh) * 2019-01-18 2020-07-28 北京沃东天骏信息技术有限公司 节点带宽的调度方法和调度装置
CN110474852B (zh) * 2019-08-01 2023-06-20 网宿科技股份有限公司 一种带宽调度方法及装置
CN111355609B (zh) * 2020-02-18 2023-04-07 山西知脉信息技术有限公司 一种互联网数据中心带宽资源冗余调整方法
CN114095443B (zh) * 2020-08-24 2023-11-03 百度在线网络技术(北京)有限公司 用于管控流量的方法、装置、电子设备及可读存储介质
CN112491564B (zh) * 2020-11-18 2023-06-02 北京金山云网络技术有限公司 网络设备的带宽计费值检测方法、装置、设备及介质
CN114500285B (zh) * 2021-12-28 2024-01-02 天翼云科技有限公司 一种带宽调度方法及装置
CN117834956B (zh) * 2023-12-18 2024-06-21 北京联广通网络科技有限公司 一种资源分发回源的直播视频加速系统及方法
CN118200146A (zh) * 2024-05-16 2024-06-14 杭州网银互联科技股份有限公司 Idc跨机房网络带宽调度方法及控制器

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012061533A1 (en) * 2010-11-02 2012-05-10 Qualcomm Incorporated Systems and methods for communicating in a network
CN102571607A (zh) * 2012-02-28 2012-07-11 大连梯耐德网络技术有限公司 一种网络流速分配系统
CN103856346A (zh) * 2012-12-06 2014-06-11 深圳市腾讯计算机系统有限公司 节点调度方法、装置和系统
CN104410582A (zh) * 2014-12-10 2015-03-11 国家电网公司 一种基于流量预测的电力通信网流量均衡方法
CN107707382A (zh) * 2017-08-15 2018-02-16 贵州白山云科技有限公司 一种智能调度方法和装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101448321B (zh) * 2008-10-28 2010-09-08 北京邮电大学 异构无线网络频谱资源共享的方法及装置
CN101867503B (zh) * 2010-06-09 2012-12-26 清华大学 一种跨域bgp路由策略部署的方法
CN102137023B (zh) * 2011-04-14 2014-01-29 中国人民解放军空军工程大学 基于可用带宽预测的组播拥塞控制方法
CN103533032B (zh) * 2013-09-26 2019-03-05 北京奇虎科技有限公司 带宽调节装置及方法
CN103825775B (zh) * 2013-12-31 2017-04-05 广东工业大学 自适应探测包长的多跳无线网络可用带宽实时检测方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012061533A1 (en) * 2010-11-02 2012-05-10 Qualcomm Incorporated Systems and methods for communicating in a network
CN102571607A (zh) * 2012-02-28 2012-07-11 大连梯耐德网络技术有限公司 一种网络流速分配系统
CN103856346A (zh) * 2012-12-06 2014-06-11 深圳市腾讯计算机系统有限公司 节点调度方法、装置和系统
CN104410582A (zh) * 2014-12-10 2015-03-11 国家电网公司 一种基于流量预测的电力通信网流量均衡方法
CN107707382A (zh) * 2017-08-15 2018-02-16 贵州白山云科技有限公司 一种智能调度方法和装置

Also Published As

Publication number Publication date
CN107707382B (zh) 2018-10-23
CN107707382A (zh) 2018-02-16

Similar Documents

Publication Publication Date Title
WO2019034065A1 (zh) 一种智能调度方法和装置及其计算机可读存储介质和计算机设备
Jalaparti et al. Dynamic pricing and traffic engineering for timely inter-datacenter transfers
WO2020224022A1 (zh) 一种资源调度方法及系统
CN107395733B (zh) 地理分布交互服务云资源协同优化方法
CN110858161A (zh) 资源分配方法、装置、系统、设备和介质
Wang et al. Pricing network resources for adaptive applications
US20200012523A1 (en) System and method for scheduling workload based on a credit-based mechanism
CN104065663A (zh) 一种基于混合云调度模型的自动伸缩、费用优化的内容分发服务方法
CN103699445A (zh) 一种任务调度方法、装置及系统
WO2020015578A1 (zh) 一种调度缓存节点的方法、装置、系统、介质及设备
CN103095846A (zh) 云计算资源的用户个性化调度方法和系统
WO2024094104A1 (zh) 一种动态反馈加权云存储资源调度方法、装置及设备
CN104009939A (zh) 一种服务资源分配方法和系统
WO2021259246A1 (zh) 资源调度方法和装置、电子设备、计算机可读存储介质
CN101383853B (zh) 一种直连节点数量控制方法及网络实体装置
CN104994150B (zh) 一种面向云视频服务的请求分配方法
WO2023151268A1 (zh) 一种业务分流方法、装置及系统
CN111694517A (zh) 分布式数据迁移方法、系统和电子设备
CN114844791B (zh) 基于大数据的云服务自动管理分配方法、系统及存储介质
WO2013082742A1 (zh) 资源调度方法、装置和系统
CN110196773B (zh) 统一调度计算资源的多时间尺度安全校核系统及方法
US11838389B2 (en) Service deployment method and scheduling apparatus
CN110475255B (zh) 网络负荷预测方法和装置
WO2020024443A1 (zh) 资源调度方法、装置、计算机设备及计算机可读存储介质
CN108563504A (zh) 一种资源管控方法和装置

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: 18846127

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18846127

Country of ref document: EP

Kind code of ref document: A1