A CDN node traffic scheduling method, computer equipment and storage medium

Abstract

This invention discloses a CDN node traffic scheduling method, computer device, and storage medium, relating to the field of network scheduling technology. The method includes receiving content access requests, extracting request association information and content classification identifiers corresponding to the content access requests, generating a request record to be scheduled, retrieving historical scheduling records of edge regions based on the request record to obtain candidate edge region distribution information, guiding the request traffic corresponding to the request record to be scheduled to the target CDN node for content distribution, obtaining actual service feedback, and updating the historical scheduling records of edge regions based on the actual service feedback to correct the target candidate region sequence. This invention improves the stability, consistency, and traceability of target CDN node selection by solidifying the expected service period, trusted service level, target candidate region ranking, and corresponding CDN node into the same permission determination path.

Description

Technical Field

[0001] This invention relates to the field of network scheduling technology, and in particular to a CDN node traffic scheduling method, computer equipment, and storage medium. Background Technology

[0002] As content delivery networks (CDNs) evolve towards edge deployment, dynamic scheduling, and refined distribution, edge regions and corresponding CDN nodes distributed across different areas are gradually forming the edge network portion of the CDN. Coordinated scheduling processes, encompassing content access request identification, candidate region selection, load prediction, permission control, and service feedback correction, have become a crucial development direction for CDN node traffic scheduling technology.

[0003] Existing CDN node traffic scheduling schemes in edge network scenarios mainly rely on geographical proximity or instantaneous service status, which do not make sufficient use of request semantic features, historical performance, and time-based reliability. This results in inaccurate judgment of the carrying capacity and reliability of candidate regions during the determination of target CDN nodes, thereby affecting the stability of request traffic scheduling and the accuracy of target CDN node determination. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a CDN node traffic scheduling method to solve the problem of insufficient reliability in the target CDN node acceptance judgment in CDN node traffic scheduling in edge network scenarios.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a CDN node traffic scheduling method, which includes receiving a content access request, extracting the request association information and content classification identifier corresponding to the content access request, generating a request record to be scheduled, and retrieving historical scheduling records of edge regions based on the request record to be scheduled to obtain candidate edge region distribution information. Based on the distribution information of candidate edge regions and the records of scheduling requests, the location differences of each candidate edge region are determined, and the candidate edge regions are sorted in combination with the historical scheduling records of the edge regions to form a sequence of target candidate regions; Based on the target candidate region sequence, historical scheduling records of edge regions, and the starting service status of corresponding CDN nodes, predictive acceptance information and predictive reliability information are obtained within a preset scheduling period. Based on the predicted acceptance information and the predicted confidence information, the predicted confidence gate is determined for the target candidate area to generate pre-occupation permission information. A revocable pre-reservation token is generated based on the pre-reservation license information. The revocable pre-reservation token is then associated with the CDN node corresponding to the target candidate region sequence to obtain the target CDN node. The request traffic corresponding to the pending request record is directed to the target CDN node for content distribution, actual service feedback is obtained, and the historical scheduling record of the edge area is updated based on the actual service feedback to correct the target candidate area sequence.

[0007] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps for obtaining the candidate edge region distribution information are as follows: Receive content access requests, extract source address, access path and content identifier, record access time, determine session context, and form request association information and content classification identifier; Associate the request association information with the content classification identifier to generate a record of requests to be scheduled. Based on the pending scheduling request record, retrieve the historical scheduling records of the edge area, match the historical acceptance path and historical service status in the historical scheduling records of the edge area, determine the distribution location and historical service status of the candidate edge area, and form the candidate edge area distribution information.

[0008] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps for forming the target candidate region sequence are as follows: The source location is determined based on the source address in the request record to be scheduled, and the distribution location of each candidate edge region in the candidate edge region distribution information is extracted. The distribution location of each candidate edge region is compared with the source location to obtain the location difference of each candidate edge region. The candidate edge regions are initially sorted according to their positional differences to obtain an initial candidate region sequence. Based on the historical scheduling records of the edge regions, the historical success rate, historical content hit rate and historical transmission continuity rate of each candidate edge region are determined, and the initial candidate region sequence is corrected to form the target candidate region sequence.

[0009] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps for obtaining the predicted acceptance information and the predicted reliability information are as follows: Based on the target candidate region sequence, obtain the starting service status, historical acceptance status and historical service status of the corresponding CDN node within the preset scheduling period; By performing periodic acceptance analysis on the target candidate region sequence based on the initial service status, historical acceptance status, and historical service status, the predicted acceptance information corresponding to the target candidate region sequence is obtained. The predicted acceptance information is used to verify the corresponding historical acceptance status and historical service status, and the predicted reliability information corresponding to the target candidate region sequence is output.

[0010] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps for generating pre-occupancy license information are as follows: The expected acceptance period and corresponding reliable acceptance level of the target candidate region within the preset scheduling cycle are determined by predicting acceptance information and predicting reliable acceptance information. Using the expected acceptance period and the credible acceptance level, perform prediction confidence gate determination on the target candidate area to determine the corresponding gate release status or gate restriction status; The target candidate area in the gate release state is marked as the pre-occupancy permission area and a pre-occupancy permission range is formed. The target candidate area in the gate restriction state is marked as the restriction area. Based on the pre-occupancy permission area, the pre-occupancy permission range and the restriction area, pre-occupancy permission information is generated.

[0011] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps of generating a revocable pre-occupancy token based on the pre-occupancy license information and associating the revocable pre-occupancy token with the CDN node corresponding to the target candidate region sequence to obtain the target CDN node are as follows: Based on the candidate edge regions within the pre-owned licensed area, a revocable pre-owned token is generated one by one, and the revocable pre-owned token is associated with the CDN node corresponding to the target candidate region sequence to form a token association relationship. Perform restriction verification on the token association based on the restricted area, remove the revocable pre-claimed token corresponding to the restricted area, and output the verified revocable pre-claimed token; Based on the target candidate region sequence, the verified revocable pre-occupied tokens are matched sequentially to generate valid token association relationships; Based on the valid token association relationship, the CDN node with the highest priority in the target candidate region sequence is determined, and the target CDN node is obtained.

[0012] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps for obtaining actual service feedback are as follows: Based on the target CDN node, the request traffic corresponding to the request to be scheduled is guided to the target CDN node, the target CDN node is used to perform content distribution on the request traffic, and the content response information corresponding to the request traffic is recorded; The system identifies the status of response reception, content hit, and transmission continuity in the content response information, forming acceptance status, content hit status, and transmission status, and generates actual service feedback.

[0013] As a preferred embodiment of the CDN node traffic scheduling method of the present invention, the specific steps of updating the historical scheduling records of the edge region and correcting the target candidate region sequence based on actual service feedback are as follows: Based on actual service feedback, the acceptance status, content hit status and transmission status of the target CDN node are written into the historical scheduling record of the edge area. Based on the historical scheduling records of edge regions after actual service feedback is written, identify the service offset and region failure status of each candidate edge region in the target candidate region sequence; Based on the service offset, the sorting position of each candidate edge region in the target candidate region sequence is adjusted to form the adjusted candidate region sorting order. Based on the failure status of the regions, the failed candidate edge regions in the adjusted candidate region sorting order are restricted and marked to form a corrected target candidate region sequence.

[0014] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein the computer program, when executed by the processor, implements any step of the CDN node traffic scheduling method as described in the first aspect of the present invention.

[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the CDN node traffic scheduling method as described in the first aspect of the present invention.

[0016] The beneficial effects of this invention are as follows: Based on the predicted acceptance information and the predicted credibility information, the predicted confidence level gate is used to determine the target candidate region and generate pre-occupancy license information. The expected acceptance period, credibility acceptance level, target candidate region ranking and corresponding CDN node are solidified into the same license determination path, which improves the stability, consistency and traceability of the target CDN node selection, makes the pre-occupancy license region and the restricted region have a clear basis for division, and forms constraints on the determination offset caused by node switching disturbance, state fluctuation and time period change. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the CDN node traffic scheduling method.

[0019] Figure 2 A flowchart for forming a sequence of target candidate regions.

[0020] Figure 3 A flowchart for generating pre-reservation license information.

[0021] Figure 4 A flowchart for obtaining the target CDN node. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a CDN node traffic scheduling method, including the following steps: S1: Receive content access requests, extract the request association information and content classification identifier corresponding to the content access requests, generate a request record to be scheduled, and retrieve the historical scheduling records of the edge area based on the request record to be scheduled to obtain the distribution information of the candidate edge area.

[0026] S1.1: Receive content access requests, extract source address, access path and content identifier, record access time, determine session context, and form request association information and content classification identifier.

[0027] Receive content access requests, extract the network address information, resource request location, session-related content, and content location content corresponding to the content access request, and record the arrival time of the content access request.

[0028] Extract the source address from the network address information and determine the access time as the time the request arrives. When the network address information contains forwarding address content, the first valid public IP address is taken as the source address; When the network address information does not contain forwarding address content, the connection address corresponding to the establishment of the connection is taken as the source address.

[0029] Extract the access path based on the location of the resource request, remove the protocol content, host content, and additional tags that are not related to the location of the target content from the access path, retain the resource access hierarchy, and unify the expression of the access path.

[0030] Based on the access path and target content location, the target content location field is extracted, and the access path is placed at the beginning and the target content location field is placed at the end. If there are multiple target content location fields, they are arranged in the order of appearance in the content access request and then concatenated in order. The standardized resource location content formed by the concatenation is determined as the content identifier.

[0031] Based on the current content access request, if the content access request contains a session identifier, the session identifier is read and determined as the session context; When a content access request does not contain a session identifier, the source address, access path, and content identifier are extracted. Related access content with the same source address, access path belonging to the same resource directory, and content identifier corresponding to the same standardized resource location content is filtered out, and the related access content is determined as the session context.

[0032] The request association information is formed by associating and combining the source address, access time, access path, session context, and content identifier.

[0033] The content business category is determined based on the access path, the content object category is determined based on the content identifier, and the combination of the content business category and the content object category is used as the content classification identifier.

[0034] S1.2: Associate the request association information and content classification identifier to generate a request record to be scheduled.

[0035] Based on request association information and content classification identifier, the source address, access time, access path, session context, content identifier, and content classification identifier are written into the same record in sequence to form the record content.

[0036] The request record number is formed by sequentially concatenating the source address, access time, access path, and content identifier, and serves as the unique record identifier for the current content access request.

[0037] The request record number is associated with the record content and written to form a request record to be scheduled.

[0038] The request record number is associated with the record content and written to form a request record to be scheduled.

[0039] S1.3: Retrieve historical scheduling records of the edge area based on the request to be scheduled, match the historical acceptance path and historical service status in the historical scheduling records of the edge area, determine the distribution location and historical service status of the candidate edge area, and form the distribution information of the candidate edge area.

[0040] Based on the pending request record, the source address, access path, content identifier, and content category identifier are extracted and used as search criteria to retrieve the historical scheduling record corresponding to the current content access request from the historical scheduling records of the edge area.

[0041] Based on the historical scheduling records of the edge region, the edge region identifier, edge region distribution location, historical connection path, historical connection status and historical service status are read. Among them, the historical service status includes historical content hit status and historical transmission status.

[0042] The access path, content identifier, and content category identifier in the pending scheduling request record are compared and matched with the historical acceptance path, historical content identifier, and historical content category identifier in the historical scheduling records of each edge area.

[0043] The access path and historical routing path are standardized, and the directory hierarchy path retained after standardization is extracted. When the directory hierarchy path corresponding to the access path is the same as the directory hierarchy path corresponding to the historical host path, it is determined to be the same resource directory as the historical host path. Filter historical scheduling records of edge areas that correspond to the same resource directory in terms of historical access paths, the same standardized resource location content in terms of historical content identifiers, and whose historical content classification identifiers are consistent with the current content classification identifiers.

[0044] The retained historical scheduling records of edge areas are categorized according to edge area identifiers. The distribution location of edge areas corresponding to each edge area identifier is read, and the occurrence frequency of each historical service status corresponding to each edge area identifier is counted.

[0045] The historical service status that appears most frequently is determined as the historical service status of the corresponding candidate edge region.

[0046] The edge region identifier, edge region distribution location, and historical service status are correlated to form candidate edge region distribution information.

[0047] S2: Based on the distribution information of candidate edge regions and the scheduling request records, determine the location differences of each candidate edge region, and sort the candidate edge regions in combination with the historical scheduling records of the edge regions to form a sequence of target candidate regions.

[0048] S2.1: Determine the source location based on the source address in the request record to be scheduled, extract the distribution location of each candidate edge region from the candidate edge region distribution information, compare the distribution location of each candidate edge region with the source location, and obtain the positional difference of each candidate edge region.

[0049] Read the edge region identifier and edge region distribution location from the candidate edge region distribution information; the edge region distribution location includes the location of the prefecture-level administrative division and the longitude and latitude values.

[0050] The location information of the prefecture-level administrative divisions corresponding to each candidate edge area is compared with the location information of the prefecture-level administrative divisions corresponding to the source location. When the candidate edge area and the source location correspond to the same prefecture-level administrative division, the location difference is determined as the location difference within the same area; When the candidate edge area and the source location correspond to different prefecture-level administrative divisions, the location difference is determined as a difference between adjacent areas if there is a public administrative division boundary between the candidate edge area and the source location based on the prefecture-level administrative division boundary information; if there is a public administrative division boundary, the location difference is determined as a difference between adjacent areas if there is no public administrative division boundary; if there is no public administrative division boundary, the location difference is determined as a difference between non-adjacent areas if there is no public administrative division boundary.

[0051] Distance calculations are performed on the longitude and latitude values ​​of the source location and the longitude and latitude values ​​of each candidate edge region to generate the geographical distance between each candidate edge region and the source location, expressed as: ; In the formula, For the first The geographical distance (in kilometers) between each candidate edge region and its source location. Represents the Earth's radius (unit: kilometers); This represents the latitude value of the distribution location corresponding to the source location. For the first The latitude values ​​of the distribution locations corresponding to each candidate edge region. This represents the longitude value of the distribution location corresponding to the source location. For the first The longitude values ​​corresponding to the distribution locations of each candidate edge region. This indicates the source location.

[0052] The location differences within the same region, the location differences between adjacent regions, and the location differences between non-adjacent regions are determined as the first priority, the second priority, and the third priority, respectively. Among the candidate edge regions in the same location difference category, the sorting order is determined according to the geographical distance from smallest to largest to obtain the location differences of each candidate edge region.

[0053] S2.2: Sort the candidate edge regions according to their positional differences to obtain an initial candidate region sequence.

[0054] Candidate edge regions with the same location difference as the same region are written into the candidate region sequence first, then candidate edge regions with the same location difference as adjacent regions are written into the candidate region sequence, and finally candidate edge regions with the same location difference as non-adjacent regions are written into the candidate region sequence. When candidate edge regions are in the same location difference category, the geographical distance between the corresponding candidate edge regions and the source location is compared, and they are written into the candidate region sequence in ascending order of geographical distance to form the initial candidate region sequence.

[0055] S2.3: Based on the historical scheduling records of the edge region, determine the historical success rate, historical content hit rate and historical transmission continuity rate of each candidate edge region, and correct the initial candidate region sequence to form the target candidate region sequence.

[0056] Based on historical scheduling records of edge regions, the total number of historical scheduling records, the number of times the historical successful connection status occurred, the number of times the historical content hit status occurred, and the number of times the historical continuous transmission status occurred are counted for each candidate edge region.

[0057] The proportion of the number of times a historical successful acceptance status occurs to the total number of historical scheduling records is determined as the historical acceptance success rate of the corresponding candidate edge region. The proportion of the number of times the historical content hit status occurs to the total number of historical scheduling records is determined as the historical content hit rate of the corresponding candidate edge area; The proportion of the number of times the historical transmission continuity status occurred to the total number of historical scheduling records is determined as the historical transmission continuity rate of the corresponding candidate edge region.

[0058] While maintaining the original order of positional differences within the same region, adjacent regions, and non-adjacent regions in the initial candidate region sequence, the candidate edge regions belonging to the same positional difference category are reordered. Within the same location difference category, the regions are sorted from high to low according to their historical success rate; if the historical success rates are the same, they are sorted from high to low according to their historical content hit rate; if the historical content hit rate is the same, they are sorted from high to low according to their historical transmission continuity rate, thus forming a sequence of target candidate regions.

[0059] S3: Based on the target candidate region sequence, historical scheduling records of edge regions, and the starting service status of the corresponding CDN nodes, obtain the predicted acceptance information and predicted reliability information within the preset scheduling period.

[0060] S3.1: Based on the target candidate region sequence, obtain the starting service status, historical acceptance status and historical service status of the corresponding CDN node within the preset scheduling period.

[0061] Read the candidate edge regions in the target candidate region sequence, and determine the CDN node corresponding to each candidate edge region according to the order of the target candidate region sequence.

[0062] The access time corresponding to the request to be scheduled is determined as the start time of the preset scheduling period. The total duration of the preset scheduling period is determined as the coverage of the preset scheduling period. The duration of the preset period segment is determined as the period division length. The preset scheduling period is continuously divided according to the duration of the period segment to form multiple period segments divided sequentially according to the duration of the period segment.

[0063] Read the starting service status of each CDN node according to the target candidate region sequence and periodic segments; the starting service status includes the current acceptance status, the current content hit status and the current transmission status. Based on the target candidate region sequence and periodic segments, the historical acceptance status and historical service status of each CDN node are extracted from the historical scheduling records of the edge region; the historical service status includes the historical content hit status and the historical transmission status.

[0064] Record the initial acceptance status, initial content hit status, and initial transmission status of each CDN node at the start of the preset scheduling period to form the initial service status statistics.

[0065] For example, the access time corresponding to the request to be scheduled is determined as the start time of the preset scheduling period, 30 minutes is determined as the total duration of the preset scheduling period, and 5 minutes is determined as the duration of the period segment. When the access time corresponding to the request to be scheduled is 10:00, a preset scheduling period is formed from 10:00 to 10:30, and it is divided into 6 period segments.

[0066] S3.2: Perform periodic acceptance analysis on the target candidate region sequence through the initial service status, historical acceptance status, and historical service status to obtain the predicted acceptance information corresponding to the target candidate region sequence.

[0067] The priority values ​​are set according to the order of each CDN node in the target candidate region sequence; the CDN node with the highest priority value corresponds to the maximum priority value, the CDN node with the lowest priority value corresponds to the minimum priority value, and the CDN nodes in between are set with their priority values ​​decreasing sequentially from the beginning to the end of the order, and the priority values ​​corresponding to adjacent positions decrease by the same difference.

[0068] Set the starting service status value based on the starting service status statistics of each CDN node; The value is 1 when the current initial acceptance state has the acceptance capability, and 0 otherwise; the value is 1 when the initial content hit state is hit, and 0 otherwise; the value is 1 when the initial transmission state is continuous, and 0 otherwise.

[0069] Based on the historical acceptance status statistics and historical service status statistics of each CDN node, the historical acceptance status value is determined according to the proportion of the number of successful occurrences of the historical acceptance status to the total number of historical scheduling records in the edge area within the corresponding period segment.

[0070] The historical content hit status value is determined by the proportion of the number of times the historical content hit status occurs to the total number of historical scheduling records in the edge area within the corresponding periodic segment. The historical transmission status value is determined by the proportion of the number of consecutive occurrences of the historical transmission status to the total number of historical scheduling records in the edge area within the corresponding periodic segment.

[0071] For each periodic segment, the values ​​of priority, initial acceptance status, initial content hit status, initial transmission status, historical acceptance status, historical content hit status, and historical transmission status are averaged to form the predicted acceptance value, expressed as: ; In the formula, For the target candidate region sequence, the first The CDN node is at the _ Predicted acceptance value within each periodic segment For the first The priority value corresponding to each CDN node. For the first The initial acceptance status value corresponding to each CDN node. For the first The initial content hit status value corresponding to each CDN node. For the first The initial transmission status values ​​corresponding to each CDN node. For the first The CDN node is at the _ The historical state values ​​within each periodic segment. For the first The CDN node is at the _ The historical content hit state value within each period segment. For the first The CDN node is at the _ The historical transmission state values ​​within each period segment. This is a sequence number marker for the corresponding CDN node. This is used to mark the sequence number of the corresponding periodic segment within the preset scheduling period.

[0072] Based on the periodic segments, the predicted capacity values ​​of each CDN node are sorted, and the CDN node with the largest predicted capacity value is determined as the predicted capacity CDN node for the corresponding periodic segment. When the predicted acceptance values ​​are the same, the CDN node with the highest order in the target candidate region sequence is determined as the predicted acceptance CDN node for the corresponding period segment.

[0073] Records corresponding to the same predicted CDN node in consecutive periodic segments are merged to form a predicted period.

[0074] The predicted CDN nodes, predicted time periods, target candidate region sequence order, and predicted acceptance values ​​are linked and recorded to form the predicted acceptance information.

[0075] S3.3: Use the predicted acceptance information to verify the corresponding historical acceptance status and historical service status, and output the predicted reliable information corresponding to the target candidate region sequence.

[0076] Based on historical scheduling records in the edge region, historical scheduling records used for predicting the acceptance value and historical scheduling records used for corresponding verification are determined separately. The historical scheduling records used for corresponding verification and the historical scheduling records used for predicting the acceptance value correspond to different historical time periods.

[0077] Based on the predicted CDN nodes and predicted time periods in the predicted acceptance information, according to the periodic segments corresponding to the predicted acceptance time periods, extract the historical acceptance status, historical content hit status and historical transmission status corresponding to the predicted CDN nodes from the corresponding historical scheduling records, and perform corresponding verifications respectively. When the historical reception status is successful, the reception verification is confirmed to have passed; when the historical content hit status is hit, the hit verification is confirmed to have passed; when the historical transmission status is continuous, the transmission verification is confirmed to have passed.

[0078] Based on the predicted acceptance period, the number of successful acceptance verifications, the number of successful hit verifications, and the number of successful transmission verifications are counted. The proportions of successful acceptance verifications to the total number of period segments corresponding to the predicted acceptance period, the number of successful hit verifications to the total number of period segments corresponding to the predicted acceptance period, and the number of successful transmission verifications to the total number of period segments corresponding to the predicted acceptance period are calculated respectively.

[0079] The average of the acceptance verification pass rate, the hit verification pass rate, and the transmission verification pass rate is used to form the predicted confidence value.

[0080] The prediction confidence value is associated with the predicted CDN node, the predicted time period, and the sequence order of the target candidate region to form prediction confidence information.

[0081] S4: Based on the predicted acceptance information and the predicted confidence information, determine the predicted confidence level of the target candidate region and generate pre-occupancy permission information.

[0082] S4.1: Determine the expected acceptance period and corresponding reliable acceptance level of the target candidate region within the preset scheduling cycle by predicting acceptance information and predicting reliable information.

[0083] Based on the target candidate region sequence, predicted acceptance information, and predicted reliability information, the target candidate region, corresponding CDN node, and predicted acceptance time period are matched accordingly. When the predicted CDN node is consistent with the CDN node corresponding to the target candidate region, and the sequence order of the target candidate region is consistent, the predicted acceptance period is determined as the expected acceptance period of the corresponding target candidate region, and the corresponding prediction confidence value is determined as the acceptance confidence value of the target candidate region. When the acceptance credibility value is greater than or equal to the preset first credibility threshold, the credibility acceptance level is determined to be a high credibility acceptance level. When the acceptance credibility value is less than the preset first credibility threshold and greater than or equal to the preset second credibility threshold, the credibility acceptance level is determined to be the medium credibility acceptance level. When the acceptance credibility value is less than the preset second credibility threshold, the credibility acceptance level is determined to be a low credibility acceptance level. The target candidate region, the expected time period for receiving the data, and the reliable level of receiving the data are linked and recorded to form the expected content to be received by the target candidate region within the preset scheduling period.

[0084] It should be noted that the predicted confidence values ​​in the historical scheduling records of the edge area are statistically analyzed, and the predicted confidence values ​​are divided into intervals according to their numerical values, forming high confidence intervals, medium confidence intervals, and low confidence intervals. The percentage of samples that meet the service compliance requirements of the corresponding content business category within each numerical interval is statistically analyzed. The lowest interval boundary value where the sample percentage reaches the high-level service compliance standard is determined as the first confidence threshold, and the lowest interval boundary value where the sample percentage reaches the basic service compliance standard is determined as the second confidence threshold, and the first confidence threshold is greater than the second confidence threshold.

[0085] For example, statistical analysis of historical scheduling records in the edge area over the past 30 days shows that when the prediction confidence value is greater than or equal to 0.80, the proportion of service-compliant samples is no less than 98%; when the prediction confidence value is greater than or equal to 0.60 and less than 0.80, the proportion of service-compliant samples is no less than 90%; 0.80 is determined as the first confidence threshold, and 0.60 is determined as the second confidence threshold.

[0086] S4.2: Utilize the expected acceptance period and credible acceptance level to perform prediction confidence gate determination on the target candidate area, and determine the corresponding gate release status or gate restriction status.

[0087] Read the expected acceptance period and trusted acceptance level corresponding to the target candidate region, and determine the starting period segment in the preset scheduling cycle that corresponds to the current scheduling starting point.

[0088] The expected time period corresponding to the target candidate region is compared with the position of the starting period segment; When the expected takeover period covers the initial period segment, the target candidate area is determined as the target candidate area that can be entered into the gate; when the expected takeover period does not cover the initial period segment, the target candidate area is determined as the gate restricted state.

[0089] When the target candidate region has the basis for immediate acceptance and the level of credible acceptance is high, the target candidate region is determined to be in the gate release state. For a target candidate area that can be determined to enter the gate, when the credible acceptance level is medium credible acceptance level and the expected acceptance period covers two or more consecutive period segments, the target candidate area is determined to be in the gate release state. For a target candidate region that can be entered into the gate, when the trusted acceptance level is low, the target candidate region is determined to be in a gate-restricted state. For a target candidate region that can be entered into the gate, when the trusted acceptance level is low, the target candidate region is determined to be in a gate-restricted state. For a target candidate region that can be entered into the gate, when the credible acceptance level is medium credible acceptance level and the expected acceptance period only covers a single period segment, the target candidate region is determined to be in a gate-restricted state. The target candidate region is associated with the corresponding gate release status or gate restriction status and recorded to form the gate judgment content with prediction confidence.

[0090] S4.3: Mark the target candidate area in the gate release state as the pre-occupancy permission area and form the pre-occupancy permission range; mark the target candidate area in the gate restriction state as the restriction area; and generate pre-occupancy permission information based on the pre-occupancy permission area, the pre-occupancy permission range, and the restriction area.

[0091] Based on the prediction confidence gate determination content, the target candidate areas in the gate release state are marked as release areas, and the corresponding target candidate areas are determined as pre-occupied permission areas; the target candidate areas in the gate restriction state are marked as restriction areas, and the corresponding target candidate areas are determined as restriction areas.

[0092] The pre-occupied permitted areas are summarized according to the sequence order of the target candidate areas to form the pre-occupied permitted range; the restricted areas are summarized according to the sequence order of the target candidate areas to form the restricted range.

[0093] The pre-owned license area, pre-owned license scope, restricted area, restricted scope, expected service period, trusted service level, and corresponding CDN node are associated and recorded to form pre-owned license information.

[0094] S5: Generate a revocable pre-occupancy token based on the pre-occupancy license information, and associate the revocable pre-occupancy token with the CDN node corresponding to the target candidate region sequence to obtain the target CDN node.

[0095] S5.1: Generate revocable pre-occupancy tokens one by one based on the candidate edge regions within the pre-occupied license area, and associate the revocable pre-occupancy tokens with the CDN nodes corresponding to the target candidate region sequence to form a token association relationship.

[0096] Based on the pre-licensed information, extract the pre-licensed area, pre-licensed scope, expected service period, trusted service level, target candidate area ranking, and corresponding CDN node; and identify candidate edge areas within the pre-licensed area according to the target candidate area ranking.

[0097] For each candidate edge region within the pre-reserved license area, a revocable pre-reservation token is generated. The revocable pre-reservation token includes the token number, the target candidate region ranking, the corresponding CDN node, the expected service period, the trusted service level, and a revocation flag. The revocation flag is initially set to an unrevoked flag.

[0098] The target candidate region order, corresponding CDN node, and expected service period are concatenated sequentially to form a token number, which is then written into the corresponding revocable pre-claim token. Each revocable pre-claimed token is associated one by one with the CDN node corresponding to the candidate edge region in the same order in the target candidate region sequence to form a token association relationship.

[0099] S5.2: Perform restriction verification on the token association based on the restricted area, remove the revocable pre-occupied token corresponding to the restricted area, and output the verified revocable pre-occupied token.

[0100] Based on the pre-occupancy license information, the restricted area and restricted range are extracted, and the candidate edge areas and corresponding CDN nodes within the restricted area are identified in order of priority of the target candidate areas.

[0101] The target candidate region priority and corresponding CDN node in the token association relationship are compared with the target candidate region priority and corresponding CDN node in the restricted region. When the target candidate region priority and the corresponding CDN node are consistent, the revocation mark of the corresponding revocable pre-claimed token is rewritten as the revocation mark.

[0102] Remove revocable pre-held tokens marked as revoked from the token association and output the verified revocable pre-held tokens.

[0103] S5.3: Perform sequential matching of the verified revocable pre-occupied tokens based on the target candidate region sequence to generate valid token association relationships.

[0104] The revocable pre-claimed tokens after verification are sorted according to the sequence order of the target candidate regions; and sorted in ascending order of the sequence order value of the target candidate regions to form the sorted content.

[0105] Identify the revocable pre-claim tokens corresponding to the same CDN node based on the order of content; when different target candidate regions correspond to the same CDN node, merge and record the pre-claim license relationships corresponding to different target candidate regions to the same CDN node, and retain the target candidate region with the smallest target candidate region sequence order value among different target candidate regions as the priority order of the corresponding CDN node.

[0106] The reserved priority order, corresponding CDN node and corresponding pre-owned license relationship are associated and recorded to generate a valid token association relationship.

[0107] S5.4: Determine the CDN node with the highest priority in the target candidate region sequence based on the valid token association relationship, and obtain the target CDN node.

[0108] Search based on the priority of valid token associations to determine the valid token association with the lowest priority. Extract the CDN node corresponding to the valid token association with the smallest priority, and determine the corresponding CDN node as the target CDN node. S6: Direct the request traffic corresponding to the request record to be scheduled to the target CDN node for content distribution, obtain actual service feedback, update the historical scheduling record of the edge area based on the actual service feedback, and correct the target candidate area sequence.

[0109] S6.1: Based on the target CDN node, the request traffic corresponding to the request to be scheduled is directed to the target CDN node, the target CDN node is used to perform content distribution on the request traffic, and the content response information corresponding to the request traffic is recorded.

[0110] Based on the target CDN node, determine the content distribution destination corresponding to the request traffic, and map the source address, access path, content identifier and request association information in the request record to be scheduled to the content distribution path of the target CDN node.

[0111] The request traffic is directed to the target CDN node, and the target CDN node responds to the content access request corresponding to the request traffic, performs content retrieval, content location and content output, thus forming the content distribution process.

[0112] Content response information is recorded during the content distribution process; the content response information includes response reception identifier, content hit identifier, transmission process record, and response time record.

[0113] The response reception identifier is used to indicate whether the requested traffic has received a content response, the content hit identifier is used to indicate whether the target CDN node has directly hit the target content, the transmission process record is used to indicate whether the content transmission process is continuous, and the response time record is used to indicate the start and end times of the content response.

[0114] The response reception identifier, content hit identifier, transmission process record, and response time record are associated and recorded to form the content response information corresponding to the request traffic.

[0115] S6.2: Perform status identification on the response reception status, content hit status, and transmission continuity status in the content response information to form the acceptance status, content hit status, and transmission status, and generate actual service feedback.

[0116] The response reception is identified based on the response reception identifier in the content response information. When the response reception identifier indicates that the content response has been received, the reception status is determined to be successful; when the response reception identifier indicates that the content response has not been received, the reception status is determined to be failed.

[0117] Content hit identification is performed based on the content hit identifier in the content response information. When the content hit identifier indicates that the target CDN node directly hits the target content, the content hit status is determined to be hit; when the content hit identifier indicates that the target CDN node does not directly hit the target content, the content hit status is determined to be miss.

[0118] Transmission continuity is identified based on the transmission process record in the content response information. When there is no transmission interruption record within the time range corresponding to the response time record, the transmission status is determined to be continuous; when there is a transmission interruption record within the time range corresponding to the response time record, the transmission status is determined to be interrupted.

[0119] The system associates and combines the acceptance status, content hit status, and transmission status, and associates them with the target CDN node, pending request records, and response time records to generate actual service feedback.

[0120] For example, the response reception identifier indicates that the content response has been received, the content hit identifier indicates that the target CDN node directly hits the target content, and the transmission process record indicates that when there are no interrupted segments, the acceptance status is determined to be successful, the content hit status is determined to be hit, and the transmission status is determined to be continuous.

[0121] S6.3: Based on actual service feedback, write the acceptance status, content hit status, and transmission status of the target CDN node into the historical scheduling record of the edge area.

[0122] Based on actual service feedback, extract records of target CDN nodes, target candidate region ranking, acceptance status, content hit status, transmission status, and response time.

[0123] Write the target CDN node, target candidate region ranking, acceptance status, content hit status, transmission status, and response time into the edge region historical scheduling record in sequence.

[0124] Create a new service record corresponding to the target CDN node in the historical scheduling records of the edge area.

[0125] S6.4: Based on the historical scheduling records of the edge regions after actual service feedback is written, identify the service offset and region failure status of each candidate edge region in the target candidate region sequence.

[0126] Based on actual service feedback, extract the target CDN node and the target candidate region ranking, and determine the candidate edge region corresponding to the target CDN node in the target candidate region sequence; When the priority of the target candidate region corresponding to the target CDN node is consistent with the original priority of the candidate edge region in the target candidate region sequence, the service offset of the candidate edge region is determined to be no offset. When the priority of the target candidate region corresponding to the target CDN node is less than the original priority of the candidate edge region in the target candidate region sequence, the service offset of the candidate edge region is determined as the forward offset. When the priority of the target candidate region corresponding to the target CDN node is greater than the original priority of the candidate edge region in the target candidate region sequence, the service offset of the candidate edge region is determined as a backward offset. For candidate edge regions in the target candidate region sequence that do not correspond to the target CDN node in this round, the service offset situation will be determined as not participating in the offset determination; The number of historical scheduling records extracted from the edge region is determined as a fixed number of records before the start of the current round of scheduling and remains unchanged during the current round of scheduling.

[0127] Using the current record time after the actual service feedback is written into the historical scheduling record of the edge area as the statistical endpoint, the most recent fixed number of historical scheduling records of the same candidate edge area are continuously extracted backward, and the corresponding range is determined as the statistical range. Within the statistical range, the number of failed acceptance records, the number of missed records, and the number of interrupted records are counted.

[0128] The ratio of the sum of failed records, missed records, and interrupted records to a fixed number of statistical records is determined as the regional failure rate. When the failure rate of a region is less than the preset failure rate threshold for the region, the failure status of the region is determined to be no failure. When the failure rate of a region is greater than or equal to the preset failure rate threshold for the region, the region is determined to be in failure. The service offset and region failure status corresponding to each candidate edge region are associated and recorded to form the candidate edge region status correction content.

[0129] It should be noted that the sample values ​​of the regional failure ratio corresponding to the historical scheduling records of each edge area are statistically analyzed, sorted by numerical value, and each regional failure ratio sample value is used as a candidate boundary value. The inter-class variance between the two groups of samples obtained by each candidate boundary value is calculated, and the candidate boundary value with the largest inter-class variance is selected as the regional failure ratio threshold (the example value is 0.15).

[0130] S6.5: Based on the service offset, adjust the sorting position of each candidate edge region in the target candidate region sequence to form the adjusted candidate region sorting order.

[0131] Based on the status correction content of the candidate edge regions, extract the service offset corresponding to each candidate edge region, and keep the candidate edge regions with no failure status participating in the sorting position adjustment.

[0132] When the service offset is no offset or not involved in the offset determination, the original order of the corresponding candidate edge region in the target candidate region sequence remains unchanged. When the service offset is a forward offset, the corresponding candidate edge region will be moved one position to the position with a smaller priority value; When the service offset is a backward offset, the corresponding candidate edge region will be moved one position to the position with a larger priority value.

[0133] When multiple candidate edge regions correspond to the same direction of movement, they are sorted in descending order of the number of records successfully received by the target CDN node; when the number of records successfully received by the target CDN node is the same, they are sorted in descending order of the number of content hit records; when the number of content hit records is the same, they are sorted in descending order of the number of consecutive transmission records, and then written into the new sorting position in sequence.

[0134] The adjusted candidate edge regions are linked and recorded to form the adjusted candidate region sorting order.

[0135] S6.6: Based on the region failure status, restrict and mark the failed candidate edge regions in the adjusted candidate region sorting order to form a corrected target candidate region sequence.

[0136] Based on the status of the candidate edge region, the content of the content correction is used to extract the region failure status as the candidate edge region with failure, and the order of the candidate edge region in the adjusted candidate region sorting order is determined. Candidate edge regions with failure status are marked with restrictions; the restriction marks include failure identifiers and priority restriction identifiers.

[0137] The failure flag indicates that the corresponding candidate edge region will not participate in the next round of target CDN node selection, while the priority restriction flag indicates that the corresponding candidate edge region will remain in the adjusted candidate region ranking order but will not participate in the next round of priority selection.

[0138] Unrestricted candidate edge regions are retained according to the adjusted candidate region sorting order, while restricted candidate edge regions are retained according to their original order.

[0139] The reserved content and the restricted marked content are associated and combined to form a corrected sequence of target candidate regions.

[0140] This embodiment also provides a computer device applicable to the CDN node traffic scheduling method, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the CDN node traffic scheduling method proposed in the above embodiment.

[0141] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0142] This embodiment also provides a storage medium storing a computer program, which, when executed by a processor, implements the CDN node traffic scheduling method proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0143] In summary, this invention improves the stability, consistency, and traceability of target CDN node selection by: performing a prediction confidence gate judgment on target candidate regions based on predicted acceptance information and predicted credibility information, and generating pre-occupancy license information; and by solidifying the expected acceptance period, credibility acceptance level, target candidate region ranking, and corresponding CDN node into the same license judgment path. This provides a clear basis for dividing pre-occupancy license regions and restricted regions, and constrains the judgment offset caused by node switching disturbances, state fluctuations, and time period changes.

[0144] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A CDN node traffic scheduling method, characterized in that, include: Receive content access requests, extract the request association information and content classification identifier corresponding to the content access requests, generate a request record to be scheduled, and retrieve historical scheduling records of edge areas based on the request record to be scheduled to obtain candidate edge area distribution information; Based on the distribution information of candidate edge regions and the records of scheduling requests, the location differences of each candidate edge region are determined, and the candidate edge regions are sorted in combination with the historical scheduling records of the edge regions to form a sequence of target candidate regions; Based on the target candidate region sequence, historical scheduling records of edge regions, and the starting service status of corresponding CDN nodes, predictive acceptance information and predictive reliability information are obtained within a preset scheduling period. Based on the predicted acceptance information and the predicted confidence information, the predicted confidence gate is determined for the target candidate area to generate pre-occupation permission information. A revocable pre-reservation token is generated based on the pre-reservation license information. The revocable pre-reservation token is then associated with the CDN node corresponding to the target candidate region sequence to obtain the target CDN node. The request traffic corresponding to the pending request record is directed to the target CDN node for content distribution, actual service feedback is obtained, and the historical scheduling record of the edge area is updated based on the actual service feedback to correct the target candidate area sequence.

2. The CDN node traffic scheduling method as described in claim 1, characterized in that, The specific steps for obtaining the candidate edge region distribution information are as follows: Receive content access requests, extract source address, access path and content identifier, record access time, determine session context, and form request association information and content classification identifier; Associate the request association information with the content classification identifier to generate a record of requests to be scheduled. Based on the pending scheduling request record, retrieve the historical scheduling records of the edge area, match the historical acceptance path and historical service status in the historical scheduling records of the edge area, determine the distribution location and historical service status of the candidate edge area, and form the candidate edge area distribution information.

3. The CDN node traffic scheduling method as described in claim 1, characterized in that, The specific steps for forming the target candidate region sequence are as follows: The source location is determined based on the source address in the request record to be scheduled, and the distribution location of each candidate edge region in the candidate edge region distribution information is extracted. The distribution location of each candidate edge region is compared with the source location to obtain the location difference of each candidate edge region. The candidate edge regions are initially sorted according to their positional differences to obtain an initial candidate region sequence. Based on the historical scheduling records of the edge regions, the historical success rate, historical content hit rate and historical transmission continuity rate of each candidate edge region are determined, and the initial candidate region sequence is corrected to form the target candidate region sequence.

4. The CDN node traffic scheduling method as described in claim 1, characterized in that, The specific steps for obtaining the prediction acceptance information and prediction reliability information are as follows: Based on the target candidate region sequence, obtain the starting service status, historical acceptance status and historical service status of the corresponding CDN node within the preset scheduling period; By performing periodic acceptance analysis on the target candidate region sequence based on the initial service status, historical acceptance status, and historical service status, the predicted acceptance information corresponding to the target candidate region sequence is obtained. The predicted acceptance information is used to verify the corresponding historical acceptance status and historical service status, and the predicted reliability information corresponding to the target candidate region sequence is output.

5. The CDN node traffic scheduling method as described in claim 1, characterized in that, The specific steps for generating the pre-owned license information are as follows: The expected acceptance period and corresponding reliable acceptance level of the target candidate region within the preset scheduling cycle are determined by predicting acceptance information and predicting reliable acceptance information. Using the expected acceptance period and the credible acceptance level, perform prediction confidence gate determination on the target candidate area to determine the corresponding gate release status or gate restriction status; The target candidate area in the gate release state is marked as the pre-occupancy permission area and a pre-occupancy permission range is formed. The target candidate area in the gate restriction state is marked as the restriction area. Based on the pre-occupancy permission area, the pre-occupancy permission range and the restriction area, pre-occupancy permission information is generated.

6. The CDN node traffic scheduling method as described in claim 1, characterized in that, The steps for generating a revocable pre-reservation token based on the pre-reservation license information and associating the revocable pre-reservation token with the CDN node corresponding to the target candidate region sequence to obtain the target CDN node are as follows: Based on the candidate edge regions within the pre-owned licensed area, a revocable pre-owned token is generated one by one, and the revocable pre-owned token is associated with the CDN node corresponding to the target candidate region sequence to form a token association relationship. Perform restriction verification on the token association relationship based on the restricted area, remove the revocable pre-claimed tokens corresponding to the restricted area, and output the verified revocable pre-claimed tokens; Based on the target candidate region sequence, the verified revocable pre-occupied tokens are matched sequentially to generate valid token association relationships; Based on the valid token association relationship, the CDN node with the highest priority in the target candidate region sequence is determined, and the target CDN node is obtained.

7. The CDN node traffic scheduling method as described in claim 1, characterized in that, The specific steps for obtaining actual service feedback are as follows: Based on the target CDN node, the request traffic corresponding to the request to be scheduled is guided to the target CDN node, the target CDN node is used to perform content distribution on the request traffic, and the content response information corresponding to the request traffic is recorded; The system identifies the status of response reception, content hit, and transmission continuity in the content response information, forming acceptance status, content hit status, and transmission status, and generates actual service feedback.

8. The CDN node traffic scheduling method as described in claim 1, characterized in that, The steps for updating the historical scheduling records of the edge area based on actual service feedback and correcting the target candidate area sequence are as follows: Based on actual service feedback, the acceptance status, content hit status and transmission status of the target CDN node are written into the historical scheduling record of the edge area. Based on the historical scheduling records of edge regions after actual service feedback is written, identify the service offset and region failure status of each candidate edge region in the target candidate region sequence; Based on the service offset, the sorting position of each candidate edge region in the target candidate region sequence is adjusted to form the adjusted candidate region sorting order. Based on the failure status of the regions, the failed candidate edge regions in the adjusted candidate region sorting order are restricted and marked to form a corrected target candidate region sequence.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the CDN node traffic scheduling method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the CDN node traffic scheduling method according to any one of claims 1 to 8.

Images