Digital service network trace link reestablishment method and apparatus

Abstract

Embodiments of the present specification provide a digital service network tracking link reconstruction method and device, wherein the digital service network tracking link reconstruction method comprises: acquiring link tracking data and aggregation metric data of a digital service network, and constructing a global directed acyclic graph according to the link tracking data; updating the global directed acyclic graph into a heterogeneous directed graph using the aggregation metric data, and constructing a heterogeneous directed snapshot sequence according to the heterogeneous directed graph; inputting a heterogeneous directed snapshot in the heterogeneous directed snapshot sequence into a heterogeneous dynamic directed graph model for processing to obtain node features corresponding to nodes included in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence; performing link completion on the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node features to obtain a target heterogeneous directed snapshot sequence, and generating a reconstructed link sequence according to the target heterogeneous directed snapshot sequence.

Description

TECHNICAL FIELD

[0001] Embodiments of the present specification relate to the technical field of digital service network, and particularly relate to a digital service network tracking link reconstruction method and device. BACKGROUND

[0002] With the development of computer and Internet technology, micro-service architecture is widely used in digital service network. The large scale of service nodes and the complexity of calling relationship in digital service network bring severe challenges to operation and management. The calling chain generated by distributed tracking technology can accurately depict the execution logic between each service in digital service network, which has important research significance for improving the governance efficiency of digital service network. There are two main challenges in distributed tracking in digital service network. First, since there is no unified supervisor to control in digital service network, distributed tracking technology is an optional item for most services in digital service network, resulting in link missing problem in tracking chain. Second, distributed tracking technology focuses on the perception of calling chain, and calling data faces the problem of inconsistent naming and inconsistent granularity. These missing problems of calling chain data seriously affect the reliability of downstream analysis tasks based on observable data. Therefore, an effective solution is needed to solve the above problems. SUMMARY

[0003] Therefore, the embodiments of the present specification provide a digital service network tracking link reconstruction method. One or more embodiments of the present specification also relate to a digital service network tracking link reconstruction device, a computing device, a computer readable storage medium and a computer program product to solve the technical defects in the prior art.

[0004] According to a first aspect of the embodiments of the present specification, a digital service network tracking link reconstruction method is provided, comprising:

[0005] obtaining link tracking data and aggregated metric data of a digital service network, and constructing a global directed acyclic graph according to the link tracking data;

[0006] updating the global directed acyclic graph into a heterogeneous directed graph using the aggregated metric data, and constructing a heterogeneous directed snapshot sequence according to the heterogeneous directed graph;

[0007] inputting the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence into a heterogeneous dynamic directed graph model for processing, to obtain node features corresponding to nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence;

[0008] completing links of the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node features, to obtain a target heterogeneous directed snapshot sequence, and generating a reconstructed link sequence according to the target heterogeneous directed snapshot sequence.

[0009] Optionally, the constructing the global directed acyclic graph according to the link tracking data comprises:

[0010] determining a plurality of link tracking sub-data in the link tracking data, and constructing a directed acyclic graph corresponding to each link tracking sub-data;

[0011] performing traversal on the directed acyclic graph corresponding to each link tracking sub-data, and merging the directed acyclic graph corresponding to each link tracking sub-data according to a traversal result to obtain the global directed acyclic graph.

[0012] Optionally, the updating the global directed acyclic graph into a heterogeneous directed graph by using the aggregated metric data comprises:

[0013] parsing the aggregated metric data, and determining a node representation vector corresponding to each global node in the global directed acyclic graph according to a parsing result;

[0014] configuring the node representation vector for the global directed acyclic graph, and obtaining a heterogeneous directed graph according to a configuration result.

[0015] Optionally, the determination of the node representation vector corresponding to any global node in the global directed acyclic graph comprises:

[0016] determining a target metric name, a target timestamp and a target metric value corresponding to a target global node according to the parsing result;

[0017] constructing a target semantic vector according to the target metric name, constructing a target time vector according to the target timestamp, and constructing a target vector according to the target metric value;

[0018] splicing the target semantic vector, the target time vector and the target vector to obtain the node representation vector corresponding to the target global node.

[0019] Optionally, the constructing a heterogeneous directed snapshot sequence according to the heterogeneous directed graph comprises:

[0020] determining a sliding window of a fixed time interval;

[0021] dividing the heterogeneous directed graph based on the sliding window, and obtaining a plurality of heterogeneous directed snapshots and composing a heterogeneous directed snapshot sequence according to a division result.

[0022] Optionally, the determination of a node representation corresponding to a node included in any heterogeneous directed snapshot in the heterogeneous directed snapshot sequence comprises:

[0023] inputting the first heterogeneous directed snapshot into a heterogeneous dynamic directed graph model, wherein the heterogeneous dynamic directed graph model comprises an intra-snapshot encoder and an inter-snapshot encoder;

[0024] performing message passing processing and message aggregation processing on the first heterogeneous directed snapshot by using the intra-snapshot encoder, and determining heterogeneous representations corresponding to nodes included in the first heterogeneous directed snapshot according to a processing result;

[0025] performing time-dependent processing on the heterogeneous representations corresponding to the nodes included in the first heterogeneous directed snapshot by using the inter-snapshot encoder, to obtain node representations corresponding to the nodes included in the first heterogeneous directed snapshot.

[0026] Optionally, the link completion of any one of the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots comprises:

[0027] constructing a target directed edge according to node representations corresponding to nodes included in a second heterogeneous directed snapshot;

[0028] detecting, by using a preset classifier, whether the second heterogeneous directed snapshot comprises the target directed edge;

[0029] if yes, performing link completion on the second heterogeneous directed snapshot based on the target directed edge.

[0030] Optionally, the generating of the sequence of reconstructed links according to the sequence of target heterogeneous directed snapshots comprises:

[0031] merging target heterogeneous directed snapshots included in the sequence of target heterogeneous directed snapshots according to a preset link reconstruction algorithm, to obtain a heterogeneous directed acyclic graph;

[0032] performing traversal on the heterogeneous directed acyclic graph, and obtaining a plurality of connected subgraphs according to a traversal result;

[0033] constructing a sequence of reconstructed links based on the plurality of connected subgraphs.

[0034] According to a second aspect of an embodiment of the present specification, a digital service network tracking link reconstruction apparatus is provided, comprising:

[0035] an acquisition module configured to acquire link tracking data and aggregated measurement data of a digital service network, and construct a global directed acyclic graph according to the link tracking data;

[0036] a construction module configured to update the global directed acyclic graph into a heterogeneous directed graph by using the aggregated measurement data, and construct a sequence of heterogeneous directed snapshots according to the heterogeneous directed graph;

[0037] The input module is configured to input the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots into a heterogeneous dynamic directed graph model for processing, to obtain node representations corresponding to nodes included in the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots.

[0038] The generation module is configured to perform link completion on the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots according to the node features, to obtain a target sequence of heterogeneous directed snapshots, and to generate a reconstructed link sequence according to the target sequence of heterogeneous directed snapshots.

[0039] According to a third aspect of the embodiments of the present specification, a computing device is provided, comprising:

[0040] a memory and a processor;

[0041] The memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions, and the computer executable instructions, when executed by the processor, implement the steps of the digital service network tracking link reconstruction method.

[0042] According to a fourth aspect of the embodiments of the present specification, a computer readable storage medium is provided, which stores computer executable instructions, and the instructions, when executed by a processor, implement the steps of the digital service network tracking link reconstruction method.

[0043] According to a fifth aspect of the embodiments of the present specification, a computer program product is provided, comprising a computer program or instructions, and the computer program or instructions, when executed by a processor, implement the steps of the digital service network tracking link reconstruction method.

[0044] The embodiments provide a digital service network tracking link reconstruction method. In order to improve the accuracy of the call link reconstruction, link tracking data and aggregation measurement data of a digital service network are acquired first. At this time, a global directed acyclic graph can be constructed according to the link tracking data. Then, the global directed acyclic graph can be updated to a heterogeneous directed graph using the aggregation measurement data, and a sequence of heterogeneous directed snapshots can be constructed according to the heterogeneous directed graph. The link relationship at different times can be embodied through the sequence of heterogeneous directed snapshots. Then, the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots can be input into a heterogeneous dynamic directed graph model for processing, to mine the relationship between nodes through the model, and then node representations corresponding to nodes included in the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots can be obtained. Then, the heterogeneous directed snapshots in the sequence of heterogeneous directed snapshots can be subjected to link completion according to the node features, and then a target sequence of heterogeneous directed snapshots can be obtained. Finally, a reconstructed link sequence can be generated according to the target sequence of heterogeneous directed snapshots, so as to ensure the accuracy of the call link reconstruction. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1is a flowchart of a method for digital service network trace link reestablishment according to an embodiment of the present specification;

[0046] Figure 2 is a schematic diagram of snapshot construction in a method for digital service network trace link reestablishment according to an embodiment of the present specification;

[0047] Figure 3 is a schematic diagram of model processing in a method for digital service network trace link reestablishment according to an embodiment of the present specification;

[0048] Figure 4 is a schematic diagram of link reestablishment in a method for digital service network trace link reestablishment according to an embodiment of the present specification;

[0049] Figure 5 is a structural schematic diagram of a device for digital service network trace link reestablishment according to an embodiment of the present specification;

[0050] Figure 6 is a structural block diagram of a computing device according to an embodiment of the present specification. DETAILED DESCRIPTION

[0051] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present specification. However, the present specification can be practiced without the specific details, other than in the examples provided herein. Those of ordinary skill in the art, and others, can readily ascertain combinations and sub-combinations of the elements of the present specification that are suitable for particular applications without departing from the scope of the present specification. Thus, the following description is not intended to limit the present specification, but to present various examples of it. Those skilled in the art will readily recognize from this disclosure that alternative embodiments of the present specification can be employed without departing from the true scope and spirit of the present specification.

[0052] The terminology used in one or more embodiments of the present specification is for the purpose of describing particular embodiments only and is not intended to be limiting of one or more embodiments of the present specification. As used in one or more embodiments of the present specification and the accompanying claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in one or more embodiments of the present specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0053] It will be understood that, although the terms first, second, etc. can be used herein to describe various information, these terms are not intended to denote a temporal sequence. Rather, these terms are used only as distinguish one from another. For example, without departing from the scope of one or more embodiments of the present specification, first can be termed second; and similarly, second can be termed first. The term "if' as used herein, can be interpreted as meaning "when" or "if," depending on the context.

[0054] In addition, it should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in one or more embodiments of the present specification are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation portal for user to choose authorization or refusal.

[0055] In the present specification, a digital service network tracking link reconstruction method is provided, and the present specification also relates to a digital service network tracking link reconstruction device, a computing device, a computer readable storage medium and a computer program product, which are described in detail one by one in the following embodiments.

[0056] In practical application, link reconstruction depends on link completion, and the link completion task is an important task in the field of graph data analysis, which mainly aims to predict the missing links or edges in the existing graph structure to complete the topology of the graph. Calling link completion is an application of link completion technology in distributed tracking of digital service network, aiming to complete the missing microservice call links due to various reasons. Observability in digital service network mainly includes three dimensions: link tracing (Tracing), aggregation metric (Metric) and log (Log), which is the basic requirement for reliable operation of digital service network, and can solve the problem of monitoring the operation of digital service network in a highly distributed environment. Due to the characteristics of large scale of service nodes and complexity of call relationship in digital service network, there are missing calls in digital service network only by using distributed tracing. Therefore, there are call link completion methods based on deep learning algorithm and statistical method mainly based on graph neural network. Among them, the call link completion method based on graph neural network models the tracing link as graph data, in which the service instance is regarded as the node and the call relationship between the nodes is regarded as the edge. The existing call link completion method based on graph neural network is mainly carried out in homogeneous graph, which captures and learns the complex relationship and pattern between service nodes to complete the potential call relationship between service nodes. However, the accuracy is low, so an effective scheme is needed to solve the above problems.

[0057] Therefore, in order to improve the accuracy of the call link reconstruction, the link tracking data and the aggregated metric data of the digital service network are acquired first, at this time, the global directed acyclic graph can be constructed according to the link tracking data; then the global directed acyclic graph is updated to a heterogeneous directed graph by using the aggregated metric data, and a heterogeneous directed snapshot sequence is constructed according to the heterogeneous directed graph; the link relationship at different times can be embodied through the heterogeneous directed snapshot sequence, after that, the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence is input to the heterogeneous dynamic directed graph model for processing, so as to mine the relationship between nodes through the model, and then the node representation corresponding to the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be obtained; after that, the link completion of the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be performed according to the node characteristics, and then the target heterogeneous directed snapshot sequence is obtained, and finally, the reconstructed link sequence is generated according to the target heterogeneous directed snapshot sequence, so as to ensure the accuracy of the call link reconstruction.

[0058] Referring to Figure 1 , Figure 1 A flow chart of a digital service network tracking link reconstruction method provided according to one embodiment of the present specification is shown, which specifically includes the following steps.

[0059] In step S102, the link tracking data and the aggregated metric data of the digital service network are acquired, and a global directed acyclic graph is constructed according to the link tracking data.

[0060] The digital service network tracking link reconstruction method provided by the embodiment is used to reconstruct the call link in the digital service network, and then stable digital service network services such as online education, remote medical treatment, electronic commerce, digital finance, etc. are provided to users, and the digital service network can realize the network system connected and interwoven by digital technology and Internet platform, and provide users with convenient, efficient and personalized service experience.

[0061] In specific implementation, the digital service network trace link reconstruction method provided by the embodiment can be divided into three steps, which are: step 1, context-enhanced heterogeneous directed snapshot sequence construction, which is used to obtain Trace data (link trace data) and Metric data (aggregated metric data) from the digital service network, construct a heterogeneous directed graph with Span as a node and Metric information as a node feature, slice and divide the heterogeneous directed graph to obtain a heterogeneous directed snapshot sequence. Step 2, learning the space-time representation of the Span node, which is used to learn the Metric representation and calling relationship of the Span node in the heterogeneous directed snapshot by constructing a heterogeneous dynamic directed graph neural network model, and obtain the node-level representation of the Span. Step 3, Trace link reconstruction, which is used to judge whether the target directed edge exists and perform link completion by learning the representation of each node, and then perform the Trace reconstruction task on the heterogeneous snapshot sequence to generate a reconstructed link sequence.

[0062] Specifically, the link trace data specifically refers to the Trace data in the digital service network, and the aggregated metric data specifically refers to the Metric data in the digital service network. Correspondingly, the global directed acyclic graph specifically refers to an acyclic graph that matches the digital service network and is constructed in combination with the link trace data, which is used for subsequent link reconstruction.

[0063] Based on this, in order to improve the accuracy of the calling link reconstruction, the link trace data and the aggregated metric data of the digital service network can be obtained first, at this time, the global directed acyclic graph can be constructed according to the link trace data. Then, the global directed acyclic graph can be updated to a heterogeneous directed graph by using the aggregated metric data, and a heterogeneous directed snapshot sequence can be constructed according to the heterogeneous directed graph. The link relationship at different times can be embodied through the heterogeneous directed snapshot sequence. Thereafter, the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be input into the heterogeneous dynamic directed graph model for processing, so as to mine the relationship between nodes through the model, and then the node representation corresponding to the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be obtained. Thereafter, the link of the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be completed according to the node feature, and then the target heterogeneous directed snapshot sequence can be obtained, and finally the reconstructed link sequence can be generated according to the target heterogeneous directed snapshot sequence.

[0064] In one or more optional embodiments in the embodiment, constructing the global directed acyclic graph according to the link trace data includes:

[0065] determining a plurality of link trace sub-data in the link trace data, and constructing a directed acyclic graph corresponding to each link trace sub-data; traversing the directed acyclic graph corresponding to each link trace sub-data, and merging the directed acyclic graph corresponding to each link trace sub-data according to the traversal result to obtain the global directed acyclic graph.

[0066] Specifically, the link tracking sub-data specifically refers to any one Trace data in the link tracking data. Based on this, when constructing the directed acyclic graph, considering that the link tracking data contains multiple link tracking sub-data, a method of respectively constructing directed acyclic graphs and then merging can be used, that is, a global directed acyclic graph can be constructed. Specifically, multiple link tracking sub-data can be determined in the link tracking data, and a directed acyclic graph corresponding to each link tracking sub-data can be constructed. At this time, the directed acyclic graph corresponding to each link tracking sub-data can be traversed, the directed acyclic graph corresponding to each link tracking sub-data is merged according to the traversal result, and then a global directed acyclic graph is obtained for subsequent use.

[0067] In summary, by using the method of respectively constructing directed acyclic graphs and then merging, the global directed acyclic graph finally constructed can be more matched with the digital service network, and then subsequent link reconstruction can be facilitated.

[0068] Step S104, updating the global directed acyclic graph to a heterogeneous directed graph by using the aggregated measurement data, and constructing a heterogeneous directed snapshot sequence according to the heterogeneous directed graph.

[0069] Specifically, the heterogeneous directed graph specifically refers to a directed graph obtained after the aggregated measurement data is configured in the global directed acyclic graph. Correspondingly, the heterogeneous directed snapshot sequence specifically refers to a sequence composed of heterogeneous directed snapshots obtained by snapshot processing the heterogeneous directed graph according to a set time interval. Since the heterogeneous directed graph changes over time, each time node can correspond to different structures, so snapshot processing can obtain multiple heterogeneous directed snapshots with different structures, thereby forming a sequence for subsequent link reconstruction.

[0070] Based on this, after obtaining the global directed acyclic graph, the global directed acyclic graph can be updated to a heterogeneous directed graph by using the aggregated measurement data, so as to configure the attribute information corresponding to the nodes in the heterogeneous directed graph. Then, a heterogeneous directed snapshot sequence can be constructed according to the heterogeneous directed graph, so that subsequent link reconstruction processing can be completed in combination with the link tracking data and the aggregated measurement data.

[0071] In one or more optional embodiments of the present embodiment, the updating of the global directed acyclic graph to a heterogeneous directed graph by using the aggregated measurement data comprises:

[0072] parsing the aggregated measurement data, determining a node representation vector corresponding to each global node in the global directed acyclic graph according to the parsing result, configuring the node representation vector for the global directed acyclic graph, and obtaining a heterogeneous directed graph according to the configuration result.

[0073] Specifically, the global node specifically refers to a node contained in the global directed acyclic graph. Correspondingly, the node representation vector specifically refers to a representation obtained by vectorizing data associated with the global node in the aggregated measurement data. Based on this, in order to ensure that the heterogeneous directed graph fully reflects the relevant attributes of the current digital service network, the aggregated measurement data can be parsed to determine the node representation vector corresponding to each global node in the global directed acyclic graph according to the parsing result; thereafter, the node representation vector can be configured for the global directed acyclic graph, so as to obtain the heterogeneous directed graph according to the configuration result for subsequent use.

[0074] In summary, by configuring the node representation vector for the directed graph, the directed graph can be attributed, thereby providing a basis for subsequent link reconstruction.

[0075] In one or more optional embodiments of the present embodiment, the determination of the node representation vector corresponding to any global node in the global directed acyclic graph comprises:

[0076] According to the parsing result, a target measurement name, a target timestamp, and a target measurement value corresponding to a target global node are determined; a target semantic vector is constructed according to the target measurement name, a target time vector is constructed according to the target timestamp, and a target vector is constructed according to the target measurement value; the target semantic vector, the target time vector, and the target vector are spliced to obtain the node representation vector corresponding to the target global node.

[0077] Specifically, the target measurement name specifically refers to the name information corresponding to the name field in the aggregated measurement data, and the target timestamp and the target measurement value specifically refer to the information corresponding to the first element and the second element of each column element in the values value list in the aggregated measurement data. Correspondingly, the target semantic vector specifically refers to a vector expression generated after semantic information extraction of the target measurement name. The target time vector specifically refers to a vector expression obtained after vector construction of the target timestamp, and the target vector specifically refers to a vector expression directly constructed from the target measurement value.

[0078] Based on this, for the node representation vector corresponding to any global node in the global directed acyclic graph, the target measurement name, the target timestamp, and the target measurement value corresponding to the target global node can be determined according to the parsing result; thereafter, the target semantic vector can be constructed according to the target measurement name, the target time vector can be constructed according to the target timestamp, and the target vector can be constructed according to the target measurement value; on this basis, the target semantic vector, the target time vector, and the target vector can be spliced to obtain the node representation vector corresponding to the target global node, and so on, until the node representation vectors corresponding to all global nodes are constructed, and subsequent processing can be performed.

[0079] In summary, by combining multi-dimensional information to construct node representation vectors, the attribute information corresponding to each global node can be more accurately reflected, thereby facilitating subsequent use.

[0080] In one or more optional embodiments of the present embodiment, constructing a heterogeneous directed snapshot sequence according to the heterogeneous directed graph comprises:

[0081] determining a fixed time interval sliding window; dividing the heterogeneous directed graph based on the sliding window, obtaining a plurality of heterogeneous directed snapshots according to the division result and composing a heterogeneous directed snapshot sequence.

[0082] Specifically, the fixed time interval can be set according to actual needs, and the present embodiment does not make any limitation here, such as 1 ms, 1 s or 1 min, etc. Based on this, after obtaining the heterogeneous directed graph, a fixed time interval sliding window can be determined; then the heterogeneous directed graph can be divided based on the sliding window, so that a plurality of heterogeneous directed snapshots can be obtained according to the division result and composed into a heterogeneous directed snapshot sequence.

[0083] In actual application, as shown in the schematic diagram, Figure 2 The service instances in the digital service network have heterogeneity, so the call graph network generated by the Trace data should also present the characteristics of the heterogeneous graph. At the same time, the calling relationship of the service instances will change over time, making the Trace data have significant dynamics. The calling relationship between services not only reflects their association, but also reveals the interaction logic between services. When the service instance is called, the corresponding metric indicators (Metric data) will change, reflecting the network running state and containing rich context information. In order to effectively capture the heterogeneity and dynamics of the Trace data, reveal the association relationship between service instances, and make full use of the Metric data to enrich the context information of the service instances, a context-enhanced heterogeneous directed snapshot sequence construction module based on the Metric data can be used, and the snapshot sequence construction process is as follows:

[0084] (1) Analyze link tracking data and generate a global directed acyclic graph. Specifically, each Span in Trace and its corresponding operation name, IP address, call start time, duration, call relationship can be extracted. Then, all service call relationships are analyzed to form a directed acyclic graph HDAG. After obtaining the HDAG of each Trace, to find the full dependency relationship of the same node in different call chains, an algorithm can be used to merge all HDAGs. The HDAG merging algorithm specifically refers to traversing the edges of the HDAGs to be merged and adding them to the completed merged HDAG. It is judged whether the updated HDAG has a loop in the time range of the newly added edge. If a loop appears, it is considered that there is a circular call in the time period, and an exception is thrown; otherwise, the information of the start and end points of the edge is merged into the node corresponding to the HDAG. Thus, the merged global directed acyclic graph HDAG is obtained.

[0085] (2) Analyze aggregation metric data and complete node representation injection enhancement. Specifically, the collected Metric data is parsed, and the “__name__” field, the first element and the second element of each list element in the “values” value list in the Metric data are extracted to form a three-tuple form of (metric name, timestamp, metric value). The three elements are converted into vectors using three methods respectively. The metric name can use BERT to extract its semantic information and generate a semantic vector. The timestamp can be converted into a time vector using random Fourier features. The metric value can directly generate a vector using its original value. Thereafter, the three vectors are spliced to serve as the representation vector of the current node. Further, through the parsing of the Metric data, the corresponding relationship between the Metric and the node on the global directed acyclic graph HDAG can be found through the specific field in the Metric data. For example, to determine the service corresponding to the metric in the current Span, first find the node_id value in the tags data in the current Span to determine the service instance corresponding to the sidecar container in the currently running pod, and then find the instance value and kubernetes_pod_name value in the Metric data to match the node_id value. Finally, according to the startTime and duration of the Span, the Metric value corresponding to the time stamp in the Metric data values list is found, and the corresponding relationship between a Span and a Metric can be completed. After determining the corresponding relationship, the node representation vector information is injected into the global directed acyclic graph HDAG to complete the node representation mapping and injection, and after all is completed, the snapshot sequence construction can be performed.

[0086] (3) Heterogeneous directed snapshot sequence division. Specifically, after obtaining the node information-rich heterogeneous directed graph, the heterogeneous directed graph can be divided into a heterogeneous directed snapshot sequence using a fixed time interval sliding window . Thus facilitating subsequent use.

[0087] Step S106, input the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence to the heterogeneous dynamic directed graph model for processing, and obtain the node representation corresponding to the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence.

[0088] Specifically, the heterogeneous dynamic directed graph model specifically refers to processing the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence, and outputting the node representation corresponding to each node. It can combine the dependency relationship between the nodes in the snapshot and the dependency relationship between the snapshots to complete the prediction of the node representation corresponding to any node. Correspondingly, the node representation specifically refers to the vector expression corresponding to any node contained in any heterogeneous directed snapshot.

[0089] Based on this, after obtaining the heterogeneous directed snapshot sequence, the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be input to the heterogeneous dynamic directed graph model for processing, realizing the output of the node representation corresponding to the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence by the model. Thus facilitating subsequent link completion operation.

[0090] In one or more optional embodiments in the present embodiment, the determination of the node representation corresponding to the nodes contained in any heterogeneous directed snapshot in the heterogeneous directed snapshot sequence includes:

[0091] input the first heterogeneous directed snapshot to the heterogeneous dynamic directed graph model, wherein the heterogeneous dynamic directed graph model includes an intra-snapshot encoder and an inter-snapshot encoder; use the intra-snapshot encoder to perform message passing processing and message aggregation processing on the first heterogeneous directed snapshot, and determine the heterogeneous representation corresponding to the nodes contained in the first heterogeneous directed snapshot according to the processing result; use the inter-snapshot encoder to perform time-dependent processing on the heterogeneous representation corresponding to the nodes contained in the first heterogeneous directed snapshot, and obtain the node representation corresponding to the nodes contained in the first heterogeneous directed snapshot.

[0092] Specifically, the intra-snapshot encoder specifically refers to an encoder that encodes in combination with the attribute information corresponding to the nodes in the heterogeneous directed snapshot. The inter-snapshot encoder specifically refers to an encoder that encodes in combination with the time-dependent information between the heterogeneous directed snapshots.

[0093] Based on this, for the node representation corresponding to a node contained in any heterogeneous directed snapshot in the heterogeneous directed snapshot sequence, the first heterogeneous directed snapshot can be first input into a heterogeneous dynamic directed graph model, which includes an intra-snapshot encoder and an inter-snapshot encoder. Then, the intra-snapshot encoder can be used to perform message passing and message aggregation processing on the first heterogeneous directed snapshot, determining the heterogeneous representation corresponding to the node contained in the first heterogeneous directed snapshot based on the processing results. Next, the inter-snapshot encoder is used to perform time-dependent processing on the heterogeneous representation corresponding to the node contained in the first heterogeneous directed snapshot, thereby obtaining the node representation corresponding to the node contained in the first heterogeneous directed snapshot. This process is repeated until the node representations corresponding to all nodes are obtained, which can then be used for subsequent link completion operations.

[0094] In practical applications, see Figure 3 The diagram illustrates that in a digital service network, service instance nodes in a call chain possess different attributes and functions. Furthermore, the dynamic nature of system operation results in significant temporal dependencies within the call chain, making the order and timing patterns of calls crucial for characterizing system behavior. Ignoring the heterogeneous representations or temporal dependencies between nodes makes it difficult to effectively model the node context information, thus hindering the accurate description of the system's operational logic and dynamic characteristics. Therefore, a heterogeneous dynamic directed graph neural network can be employed to learn the heterogeneous representation information between nodes and the temporal dependencies of the call chain, thereby enriching node representations. The structure and specific process of the heterogeneous dynamic directed graph neural network are as follows:

[0095] (1) Intra-snapshot heterogeneous attention encoder. Specifically, for each snapshot, the heterogeneous representation of the Span node is learned through two steps: message passing and message aggregation. Message passing: For the edge e=(u,v) on the k-th layer, the multi-head message aggregation from node u to node v can be calculated according to the following formulas (1) and (2):

[0096] (1)

[0097] (2)

[0098] in, For nodes In the Layer node representation, Indicates node type linear mapping layer, The number of long positions.

[0099] Message aggregation: using Compute each source node on the snapshot The importance of this is thus recognized, leading to the aggregation of neighbor node information to the target node. The formula (3) can be obtained as follows:

[0100] (3)

[0101] wherein, denotes the set of all neighbor nodes of the node .

[0102] (2) Snapshot inter-Transformer encoder. Specifically, considering the time dependency of the past and current call chains, the node embedding sequence obtained by the intra-snapshot encoder is input into the Transformer to extract the time dependency information . Wherein, is the number of nodes in the intra-snapshot. At this time, the node representation introduces the time dependency of the call chain. Thus, the subsequent link completion and reconstruction of the link can be completed based on the node representation.

[0103] Step S108, link completion is performed on the heterogeneous directed snapshots in the heterogeneous directed snapshot sequence according to the node features, a target heterogeneous directed snapshot sequence is obtained, and a reconstructed link sequence is generated according to the target heterogeneous directed snapshot sequence.

[0104] Specifically, the link completion specifically refers to the completion of the missing edges of the heterogeneous directed snapshots, which is used to ensure that the completed snapshots are more in line with the actual situation. Correspondingly, the target heterogeneous directed snapshot sequence specifically refers to a sequence composed of new snapshots obtained after the link completion of the heterogeneous directed snapshots in the heterogeneous directed snapshot sequence. Correspondingly, the reconstructed link sequence specifically refers to a sequence composed of new links obtained after the reconstruction of the link tracking data related to the digital service network.

[0105] Based on this, when performing link reconstruction, the link completion can be performed on the heterogeneous directed snapshots in the heterogeneous directed snapshot sequence according to the node features, the target heterogeneous directed snapshot sequence is obtained according to the completion result, and then the reconstructed link sequence can be generated according to the target heterogeneous directed snapshot sequence, thereby facilitating the use of downstream businesses.

[0106] In one or more optional embodiments in the present embodiment, the link completion of any one of the heterogeneous directed snapshots in the heterogeneous directed snapshot sequence includes:

[0107] constructing a target directed edge according to the node representation corresponding to the nodes included in the second heterogeneous directed snapshot; detecting whether the second heterogeneous directed snapshot includes the target directed edge by using a preset classifier; if yes, performing link completion on the second heterogeneous directed snapshot based on the target directed edge.

[0108] Specifically, the second heterogeneous directed snapshot specifically refers to any one of the heterogeneous directed snapshot sequence. Correspondingly, the target directed edge specifically refers to the possible existing edge added for the snapshot. Correspondingly, the classifier specifically refers to the classifier for judging the target directed edge.

[0109] Based on this, when completing the link for any one of the heterogeneous directed snapshot sequence, the target directed edge can be constructed according to the node representation corresponding to the node pair included in the second heterogeneous directed snapshot. At this time, the preset classifier can be used to detect whether the second heterogeneous directed snapshot includes the target directed edge. If yes, it means that the currently constructed target directed edge is the one that the snapshot should have, and therefore the second heterogeneous directed snapshot can be link-completed based on the target directed edge. If no, it means that the edge does not exist, and therefore the heterogeneous directed snapshot can be directly added to the target heterogeneous directed snapshot sequence.

[0110] In summary, by detecting each edge to be added, it can be ensured that the snapshots included in the target heterogeneous directed snapshot sequence are more consistent with the application scenario, thereby facilitating subsequent link reconstruction.

[0111] In one or more optional embodiments in the present embodiment, generating a reconstructed link sequence according to the target heterogeneous directed snapshot sequence comprises:

[0112] According to a preset link reconstruction algorithm, the target heterogeneous directed snapshots included in the target heterogeneous directed snapshot sequence are merged to obtain a heterogeneous directed acyclic graph; the heterogeneous directed acyclic graph is traversed, and a plurality of connected subgraphs are obtained according to the traversal result; and a reconstructed link sequence is constructed based on the plurality of connected subgraphs.

[0113] Specifically, the link reconstruction algorithm specifically refers to an algorithm for reconstructing a link according to the target heterogeneous directed snapshot sequence, which is inverse to the above-mentioned algorithm for constructing a sequence. Correspondingly, the heterogeneous directed acyclic graph specifically refers to a directed acyclic graph constructed in combination with the target heterogeneous directed snapshots included in the target heterogeneous directed snapshot sequence.

[0114] Based on this, when performing link reconstruction, the target heterogeneous directed snapshots included in the target heterogeneous directed snapshot sequence can be merged according to a preset link reconstruction algorithm, and then a heterogeneous directed acyclic graph is obtained. Thereafter, the heterogeneous directed acyclic graph can be traversed, and a plurality of connected subgraphs are obtained according to the traversal result. Finally, a reconstructed link sequence can be constructed based on the plurality of connected subgraphs.

[0115] In actual application, for example, Figure 4As shown in the schematic diagram, after obtaining the snapshot sequence with rich node representation information, in order to realize the link reconstruction task, it is necessary to detect the missing situation of the link and complete the edges judged to be missing, and then complete the missing edges to the corresponding heterogeneous snapshots, and then reconstruct the snapshot sequence into a complete Trace sequence. In order to realize the above function, a link completion classifier and a Trace reconstruction algorithm can be used to realize the link reconstruction task, as follows:

[0116] (1) Link completion. Specifically, after obtaining the representation of each node, the starting point and the ending point of the target directed edge can be spliced, and the spliced features obtained can be input into the link completion classifier to judge whether the target directed edge exists, which can be realized by the following formula (4):

[0117] (4)

[0118] Wherein, , are the representations of the starting node and the ending node of the target directed edge.

[0119] In the training process, a global random negative sample sampling strategy can be used, and the loss function is the cross entropy loss function, as shown in the following formula (5):

[0120] (5)

[0121] Wherein, is the number of edges, is the true label of the sample , is the model judgment result.

[0122] (2) Reconstructing the Trace call chain. Specifically, first, the missing edges are added back to the heterogeneous snapshots to obtain a new heterogeneous snapshot sequence For the new snapshot sequence, the Trace reconstruction algorithm is used to reconstruct the snapshot sequence into a complete Trace sequence. The reconstruction algorithm process is as follows: first, the same HDAG merging algorithm is used to merge the heterogeneous snapshot sequence into a heterogeneous directed acyclic graph DAG'. For DAG', traverse each node without a parent node, traverse the time period information of each node, and obtain the maximum connected subgraph DAG'T of DAG' under the time period information. The nodes of DAG'T are converted into spans, and the connection relationship is converted into a call relationship to obtain . The final sequence is the reconstructed call chain sequence.

[0123] The embodiment provides a digital service network tracking link reconstruction method. In order to improve the accuracy of call link reconstruction, link tracking data and aggregated measurement data of a digital service network are acquired first, at this time, a global directed acyclic graph can be constructed according to the link tracking data. Then, the global directed acyclic graph can be updated into a heterogeneous directed graph by using the aggregated measurement data, and a heterogeneous directed snapshot sequence can be constructed according to the heterogeneous directed graph. The link relationship at different times can be embodied through the heterogeneous directed snapshot sequence. Then, the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be input into a heterogeneous dynamic directed graph model for processing, so as to mine the relationship between nodes through the model, and then the node representation corresponding to the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be obtained. Then, the link of the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be completed according to the node characteristics, and then the target heterogeneous directed snapshot sequence can be obtained. Finally, the reconstructed link sequence can be generated according to the target heterogeneous directed snapshot sequence, so as to ensure the accuracy of call link reconstruction.

[0124] Corresponding to the method embodiment, the specification also provides a digital service network tracking link reconstruction device embodiment, Figure 5 The structure of a digital service network tracking link reconstruction device provided by one embodiment of the specification is shown. As shown in the figure, Figure 5 The device comprises:

[0125] The acquisition module 502 is configured to acquire link tracking data and aggregated measurement data of a digital service network, and construct a global directed acyclic graph according to the link tracking data.

[0126] The construction module 504 is configured to update the global directed acyclic graph into a heterogeneous directed graph by using the aggregated measurement data, and construct a heterogeneous directed snapshot sequence according to the heterogeneous directed graph.

[0127] The input module 506 is configured to input the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence into a heterogeneous dynamic directed graph model for processing, and obtain the node representation corresponding to the nodes contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence.

[0128] The generation module 508 is configured to complete the link of the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node characteristics, obtain a target heterogeneous directed snapshot sequence, and generate a reconstructed link sequence according to the target heterogeneous directed snapshot sequence.

[0129] In an optional embodiment, the acquisition module 502 is further configured to:

[0130] A plurality of link tracking sub-data are determined in the link tracking data, and a directed acyclic graph corresponding to each link tracking sub-data is constructed; each directed acyclic graph corresponding to each link tracking sub-data is traversed, and each directed acyclic graph corresponding to each link tracking sub-data is merged according to a traversal result to obtain a global directed acyclic graph.

[0131] In an optional embodiment, the construction module 504 is further configured to:

[0132] The aggregation metric data is parsed, and a node representation vector corresponding to each global node in the global directed acyclic graph is determined according to a parsing result; the node representation vector is configured for the global directed acyclic graph, and a heterogeneous directed graph is obtained according to a configuration result.

[0133] In an optional embodiment, the determination of the node representation vector corresponding to any global node in the global directed acyclic graph comprises:

[0134] A target metric name, a target timestamp, and a target metric value corresponding to a target global node are determined according to a parsing result; a target semantic vector is constructed according to the target metric name, a target time vector is constructed according to the target timestamp, and a target vector is constructed according to the target metric value; the target semantic vector, the target time vector, and the target vector are spliced to obtain the node representation vector corresponding to the target global node.

[0135] In an optional embodiment, the construction module 504 is further configured to:

[0136] A sliding window of a fixed time interval is determined; the heterogeneous directed graph is divided based on the sliding window, and a plurality of heterogeneous directed snapshots are obtained according to a division result to form a heterogeneous directed snapshot sequence.

[0137] In an optional embodiment, the determination of the node representation corresponding to the node included in any heterogeneous directed snapshot in the heterogeneous directed snapshot sequence comprises:

[0138] The first heterogeneous directed snapshot is input into a heterogeneous dynamic directed graph model, wherein the heterogeneous dynamic directed graph model includes an intra-snapshot encoder and an inter-snapshot encoder; the intra-snapshot encoder is used for message passing processing and message aggregation processing on the first heterogeneous directed snapshot, and a heterogeneous representation corresponding to the node included in the first heterogeneous directed snapshot is determined according to a processing result; the inter-snapshot encoder is used for time-dependent processing on the heterogeneous representation corresponding to the node included in the first heterogeneous directed snapshot, and a node representation corresponding to the node included in the first heterogeneous directed snapshot is obtained.

[0139] In an optional embodiment, the link completion of any heterogeneous directed snapshot in the heterogeneous directed snapshot sequence comprises:

[0140] According to the node representation corresponding to the nodes included in the second heterogeneous directed snapshot, a target directed edge is constructed; a preset classifier is used to detect whether the second heterogeneous directed snapshot includes the target directed edge; if yes, link completion is performed on the second heterogeneous directed snapshot based on the target directed edge.

[0141] In an optional embodiment, the generation module 508 is further configured to:

[0142] According to a preset link reconstruction algorithm, the target heterogeneous directed snapshots included in the target heterogeneous directed snapshot sequence are merged to obtain a heterogeneous directed acyclic graph; the heterogeneous directed acyclic graph is traversed, and a plurality of connected subgraphs are obtained according to a traversal result; and a reconstruction link sequence is constructed based on the plurality of connected subgraphs.

[0143] The embodiment provides a digital service network tracking link reconstruction device. In order to improve the accuracy of the call link reconstruction, link tracking data and aggregation measurement data of the digital service network can be acquired first. At this time, a global directed acyclic graph can be constructed according to the link tracking data. Then, the global directed acyclic graph can be updated into a heterogeneous directed graph by using the aggregation measurement data, and a heterogeneous directed snapshot sequence can be constructed according to the heterogeneous directed graph. The link relationship at different times can be embodied through the heterogeneous directed snapshot sequence. Then, the heterogeneous directed snapshots in the heterogeneous directed snapshot sequence can be input into a heterogeneous dynamic directed graph model for processing, so as to mine the relationship between nodes through the model. Accordingly, the node representation corresponding to the nodes included in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be obtained. Then, the link of the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence can be completed according to the node representation, and a target heterogeneous directed snapshot sequence can be obtained. Finally, a reconstruction link sequence can be generated according to the target heterogeneous directed snapshot sequence, so as to ensure the accuracy of the call link reconstruction.

[0144] The above is a schematic scheme of the digital service network tracking link reconstruction device of the embodiment. It should be noted that the technical scheme of the digital service network tracking link reconstruction device and the technical scheme of the digital service network tracking link reconstruction method described above belong to the same concept. The technical scheme of the digital service network tracking link reconstruction device, which is not described in detail, can be referred to the description of the technical scheme of the digital service network tracking link reconstruction method.

[0145] Figure 6 A structural block diagram of a computing device 600 is shown according to an embodiment of the present specification. The components of the computing device 600 include but are not limited to a memory 610 and a processor 620. The processor 620 is connected with the memory 610 through a bus 630, and a database 650 is used to save data.

[0146] The computing device 600 also includes an access device 640 that enables the computing device 600 to communicate via one or more networks 660. Examples of such networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or combinations of such networks, such as the Internet. The access device 640 can include one or more of any type of network interface (for example, a network interface card (NIC)) such as a wired or wireless network interface, such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Worldwide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a Near Field Communication (NFC).

[0147] In one embodiment of the present specification, the above-mentioned components of the computing device 600 and other components not shown in the Figure 6 may be connected to each other, such as through a bus. It should be understood that Figure 6 The computing device structure diagram shown is only for the purpose of example, and is not a limitation on the scope of the present specification. Those skilled in the art can add or replace other components as needed.

[0148] The computing device 600 can be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (for example, a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, and the like), a mobile phone (for example, a smartphone), a wearable computing device (for example, a smart watch, smart glasses, and the like), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 600 can also be a mobile or stationary server.

[0149] The processor 620 is configured to execute computer-executable instructions, which, when executed by the processor, implement the steps of the above-mentioned digital service network tracking link reconstruction method.

[0150] The above is a schematic scheme of the computing device of the embodiment. It should be noted that the technical scheme of the computing device and the technical scheme of the digital service network tracking link reestablishment method described above belong to the same concept, and the details of the technical scheme of the computing device that are not described in detail can be seen from the description of the technical scheme of the digital service network tracking link reestablishment method.

[0151] An embodiment of the present specification also provides a computer readable storage medium storing computer executable instructions, which, when executed by a processor, implement the steps of the digital service network tracking link reestablishment method described above.

[0152] The above is a schematic scheme of the computer readable storage medium of the embodiment. It should be noted that the technical scheme of the storage medium and the technical scheme of the digital service network tracking link reestablishment method described above belong to the same concept, and the details of the technical scheme of the storage medium that are not described in detail can be seen from the description of the technical scheme of the digital service network tracking link reestablishment method.

[0153] An embodiment of the present specification also provides a computer program product comprising a computer program or instructions, which, when executed by a processor, implement the steps of the digital service network tracking link reestablishment method described above.

[0154] The above is a schematic scheme of the computer program product of the embodiment. It should be noted that the technical scheme of the computer program product and the technical scheme of the digital service network tracking link reestablishment method described above belong to the same concept, and the details of the technical scheme of the computer program product that are not described in detail can be seen from the description of the technical scheme of the digital service network tracking link reestablishment method.

[0155] The above describes specific embodiments of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the acts or steps recited in the claims can be performed in an order other than that described in the embodiments and still achieve desirable results. In addition, the processes depicted in the figures do not necessarily require the particular order shown or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing can be advantageous or possible.

[0156] The computer readable medium can include any entity or apparatus capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, software distribution medium, etc. It should be noted that the computer readable medium can include appropriate additions or subtractions according to the requirements of patent practice. For example, according to the patent practice in some regions, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0157] It should be noted that, for the foregoing method embodiments, in order to facilitate description, each is described as a combination of a series of acts, but those skilled in the art should appreciate that the embodiments of the present specification are not limited by the order of the described acts, because according to the embodiments of the present specification, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should appreciate that the embodiments described in the specification are all preferred embodiments, and the acts and modules involved are not necessarily essential to the embodiments of the present specification.

[0158] In the above embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0159] The preferred embodiments of the present specification disclosed above are only used to help explain the present specification. The alternative embodiments do not describe all the details and do not limit the invention to the specific embodiments described. Obviously, according to the content of the embodiments of the present specification, many modifications and changes can be made. The present specification selects and describes these embodiments in order to better explain the principles and practical applications of the embodiments of the present specification, so that those skilled in the art can well understand and use the present specification.

Claims

1. A method of digital service network trace link reestablishment, the method comprising: The method comprises the following steps: acquiring link tracking data and aggregated metric data of a digital service network, and constructing a global directed acyclic graph according to the link tracking data; updating the global directed acyclic graph into a heterogeneous directed graph by using the aggregated metric data, and constructing a heterogeneous directed snapshot sequence according to the heterogeneous directed graph; inputting a heterogeneous directed snapshot in the heterogeneous directed snapshot sequence into a heterogeneous dynamic directed graph model, wherein the heterogeneous dynamic directed graph model comprises an intra-snapshot encoder and an inter-snapshot encoder, the intra-snapshot encoder is used for message passing processing and message aggregation processing on each heterogeneous directed snapshot, the heterogeneous representation corresponding to a node included in each heterogeneous directed snapshot is determined according to a processing result, the inter-snapshot encoder is used for time-dependent processing on the heterogeneous representation corresponding to the node included in each heterogeneous directed snapshot, and a node representation corresponding to the node included in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence is obtained; link completion is performed on the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node representation, a target heterogeneous directed snapshot sequence is obtained, and a reconstructed link sequence is generated according to the target heterogeneous directed snapshot sequence, wherein the link completion performed on the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node representation comprises: constructing a target directed edge according to the node representation corresponding to the node included in the heterogeneous directed snapshot; and when the target directed edge is detected in any one of the heterogeneous directed snapshots by using a preset classifier, performing link completion on the heterogeneous directed snapshot based on the target directed edge.

2. The method of claim 1, wherein, The global directed acyclic graph is constructed according to the link tracking data, which comprises the following steps: determining a plurality of link tracking sub-data in the link tracking data, and constructing a directed acyclic graph corresponding to each link tracking sub-data; traversing the directed acyclic graph corresponding to each link tracking sub-data, and merging the directed acyclic graphs corresponding to each link tracking sub-data according to a traversal result to obtain a global directed acyclic graph.

3. The method of claim 1, wherein, The global directed acyclic graph is updated into a heterogeneous directed graph by using the aggregated metric data, which comprises the following steps: parsing the aggregated metric data, and determining a node representation vector corresponding to each global node in the global directed acyclic graph according to a parsing result; configuring the node representation vector for the global directed acyclic graph, and obtaining a heterogeneous directed graph according to a configuration result.

4. The method of claim 3, wherein, The node representation vector corresponding to each global node in the global directed acyclic graph is determined according to the parsing result, which comprises the following steps: determining a target metric name, a target timestamp and a target metric value corresponding to a target global node according to the parsing result; constructing a target semantic vector according to the target metric name, a target time vector according to the target timestamp, and a target vector according to the target metric value; splicing the target semantic vector, the target time vector and the target vector to obtain the node representation vector corresponding to the target global node.

5. The method of claim 1, wherein, The heterogeneous directed snapshot sequence is constructed according to the heterogeneous directed graph, which comprises the following steps: determining a sliding window with a fixed time interval; The heterogeneous directed graph is divided based on the sliding window, and a plurality of heterogeneous directed snapshots are obtained according to a division result and combined to form a heterogeneous directed snapshot sequence.

6. The method of claim 1, wherein, The generating a reconstructed link sequence according to the target heterogeneous directed snapshot sequence comprises: merging target heterogeneous directed snapshots contained in the target heterogeneous directed snapshot sequence according to a preset link reconstruction algorithm to obtain a heterogeneous directed acyclic graph; performing traversal on the heterogeneous directed acyclic graph to obtain a plurality of connected subgraphs according to a traversal result; constructing a reconstructed link sequence based on the plurality of connected subgraphs.

7. A digital service network trace link reestablishment apparatus characterized by comprising: comprises: an acquisition module configured to acquire link tracking data and aggregated metric data of a digital service network, and construct a global directed acyclic graph according to the link tracking data; a construction module configured to update the global directed acyclic graph to a heterogeneous directed graph by using the aggregated metric data, and construct a heterogeneous directed snapshot sequence according to the heterogeneous directed graph; an input module configured to input a heterogeneous directed snapshot in the heterogeneous directed snapshot sequence to a heterogeneous dynamic directed graph model, wherein the heterogeneous dynamic directed graph model contains an intra-snapshot encoder and an inter-snapshot encoder, the intra-snapshot encoder is used to perform message passing processing and message aggregation processing on each heterogeneous directed snapshot, and a heterogeneous representation corresponding to a node contained in each heterogeneous directed snapshot is determined according to a processing result, the inter-snapshot encoder is used to perform time-dependent processing on the heterogeneous representation corresponding to the node contained in each heterogeneous directed snapshot to obtain a node representation corresponding to the node contained in the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence; a generation module configured to perform link completion on the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node representation to obtain a target heterogeneous directed snapshot sequence, and generate a reconstructed link sequence according to the target heterogeneous directed snapshot sequence, wherein the performing link completion on the heterogeneous directed snapshot in the heterogeneous directed snapshot sequence according to the node representation comprises: constructing a target directed edge according to the node representation corresponding to the node contained in the heterogeneous directed snapshot; and based on the target directed edge, performing link completion on the heterogeneous directed snapshot in a case where the heterogeneous directed snapshot contains the target directed edge by using a preset classifier.

8. A computing device, comprising: comprises: 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, and the computer executable instructions, when executed by the processor, realize the steps of the method in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer executable instructions, when executed by the processor, realize the steps of the method in any one of claims 1 to 6.

10. A computer program product, characterised in that, The computer program or instructions, when executed by the processor, realize the steps of the method in any one of claims 1 to 6.

Images