Unified identification generation and management method for multi-source heterogeneous data

Abstract

The application relates to the technical field of data management, and particularly discloses a unified identification generation and management method for multi-source heterogeneous data, which comprises the following steps: extracting a target identification in a unified identification mapping table, synchronously collecting corresponding multi-dimensional features of the target identification, such as a silent duration, a merging conflict frequency, a query frequency, a dependency depth and a source station connectivity state, forming a complete sign snapshot with a time-stamped dynamic index and a topological association parameter, performing time sequence alignment and multi-dimensional mapping on multi-source historical logs and an association graph, and constructing a sign health degree model capable of quantifying identification activity and dependency strength. The core of the model is to inversely deduce a precise health score of the identification by using a dynamic ratio of a decay index and a topological support index. The method breaks the limitation of traditional fixed threshold judgment, realizes an intelligent degradation strategy based on multi-source fusion dynamic evaluation, and provides a reliable scientific basis for identification life cycle management in a distributed system.

Description

Technical Field

[0001] This invention belongs to the field of data management technology and relates to a unified identifier generation and management method for multi-source heterogeneous data. Background Technology

[0002] In a multi-source heterogeneous data environment, unified identifier mapping serves as the core hub for cross-system entity association, and its accuracy and consistency directly determine the completeness and reliability of query results. Meanwhile, the continuous activity and high-quality supply of the underlying data sources are fundamental to ensuring the freshness and effectiveness of the information mapped by the identifiers, playing an irreplaceable role in the stable operation of the data service chain.

[0003] It is worth noting that the vitality of an identifier is highly dependent on the dynamic state of its corresponding data source. Once the source system becomes inactive due to maintenance, failure, or shutdown, the data associated with that identifier will gradually become distorted over time. If such zombie identifiers continue to participate in resolution and topology referencing, they will quietly erode the accuracy of data products. Especially for identifiers that are frequently queried or deeply nested, even minor deterioration in their health status can be amplified along the dependency chain, making multidimensional activity monitoring of identifiers particularly significant.

[0004] However, current management of such identifiers is clearly inadequate. Traditional monitoring methods rely excessively on fixed silent duration thresholds, ignoring individual differences in recovery patterns and quality fluctuations across different data sources. Judging solely by the last update time is like focusing on surface values ​​while neglecting deeper health issues; it fails to distinguish between temporary delays and permanent cessation of updates, and cannot detect hidden degradation caused by failed data integrity checks or declining source network connectivity. Furthermore, existing mechanisms lack integrated analysis of multiple factors such as downstream citation intensity, source activity, and quality events, resulting in reactive and delayed health warnings. A large number of potentially problematic identifiers remain on active links, continuously producing incomplete or even erroneous results undetected, exposing significant shortcomings in the dynamic perception and proactive degradation capabilities of identifier lifecycle management. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides a unified identifier generation and management method for multi-source heterogeneous data to solve the above-mentioned technical problems.

[0006] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows: This invention provides a unified identifier generation and management method for multi-source heterogeneous data, the method comprising: Extract the target identifier from the unified identifier mapping table; Obtain the last update time of the data source corresponding to the target identifier, calculate the time difference between the current system time and the last update time, and obtain the silence duration; Retrieve the multi-source historical merge logs of the target identifier, and use the cumulative number of attribute alignment failures as the merge conflict count; Capture the number of parsing requests for the target identifier within a preset time window, and calculate the average number of parsing requests per unit time as the query frequency; Traverse the unified identifier association graph and extract the longest hierarchical link value that is referenced by other identifiers with the target identifier as the endpoint as the dependency depth; Detect the current network heartbeat value of the underlying physical data source corresponding to the target identifier, and convert the current network heartbeat value into the source station connectivity state; The attenuation index is calculated based on the silence duration and the number of merging conflicts; Calculate the topology support index based on the query frequency, the dependency depth, and the source site connectivity state; The health status of the target identifier is calculated by dividing the topological support index by the sum of the attenuation index and the preset base value. When the health status of the vital signs is greater than or equal to a preset health threshold, a normal maintenance instruction is generated; When the physical health status is less than the preset health threshold, compare the topological support index with the preset dependency threshold. When the topology support index is greater than or equal to the preset dependency threshold, a shadow route degradation instruction is generated; When the topology support index is less than the preset dependency threshold and the dependency depth is greater than the preset minimum depth, a tombstone blocking instruction is generated. When the topology support index is less than the preset dependency threshold and the dependency depth is equal to the preset minimum depth, a physical erase command is generated.

[0007] As described above, the unified identifier generation and management method for multi-source heterogeneous data provided by the present invention has at least the following beneficial effects: This invention integrates the silence duration and the number of merge conflicts due to attribute alignment failures into a decay index, and aggregates query frequency, dependency depth, and origin server connectivity into a topology support index. The health status is calculated using the dynamic ratio of the two, which effectively avoids the rigid misjudgment caused by the traditional solution's reliance on a fixed silence threshold. This makes the identification activity assessment more closely match the actual recovery capability and quality fluctuations of the data source, and significantly improves the accuracy of capturing latent deterioration identifications.

[0008] This invention utilizes a dual assessment of health indicators and topology support index to implement shadow routing degradation for identifiers below the health threshold but with strong topology support, preventing error propagation while preserving recoverability. Identifiers with weak topology support and shallow dependency depth are directly physically erased, completely releasing resources from the index. This tiered elimination mechanism not only ensures the query quality of frequently dependent identifiers but also automatically cleans up zombie entries in the data swamp without manual intervention, significantly enhancing the self-healing capability and service stability of the distributed identifier system. Attached Figure Description

[0009] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. 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.

[0010] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention. Detailed Implementation

[0011] The present invention will be further described in detail below with reference to specific embodiments. The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined by the claims, all of which should fall within the protection scope of the present invention.

[0012] Please see Figure 1 As shown, a unified identifier generation and management method for multi-source heterogeneous data includes: Extract the target identifier from the unified identifier mapping table; Obtain the last update time of the data source corresponding to the target identifier, calculate the time difference between the current system time and the last update time, and obtain the silence duration; Preferably, the step of obtaining the last update time of the data source corresponding to the target identifier, calculating the time difference between the current system time and the last update time, and obtaining the silence duration includes: Parse the source merging trajectory column recorded in the unified identifier mapping table; Extract in batches the set of local modification timestamps of each heterogeneous sub-identifier belonging to different source data tables from the source merged trajectory column; Compare the values ​​in the set of local modification timestamps, and select the timestamp with the largest value as the last update time value; A synchronization request is sent to the system's central clock unit to obtain a real-time stable timestamp as the current system time; Solve for the absolute difference sequence formed by subtracting the last update time from the current system time, and output the absolute difference sequence as the silence duration.

[0013] In the daily operation and data flow of the unified identifier system, each unified identifier typically corresponds to data mapping relationships of multiple different business source systems at the underlying level. The underlying backbone process of the system will read the system cache or persistent storage medium according to the preset batch processing cycle, extract the primary key sequence in the unified identifier mapping table, and use the primary key field currently being processed as the target identifier for subsequent evaluation.

[0014] Because the constituent attributes of a single target identifier are scattered across data tables from various heterogeneous sources, the freshness of its data cannot be determined solely by a single data source. The system's underlying parser reads the source merging trajectory column field corresponding to the data row containing the target identifier. This field is structurally represented as a structured log text recording the history of source data aggregation, which includes associated subkeys and metadata information from the HIS system, LIS system, and other external sources. The system performs string matching or object traversal operations to retrieve the time parameters recorded within, extracting the last attribute change time for each heterogeneous sub-identifier belonging to different source data tables, and summarizing them to generate a set of local modification timestamps consisting of multiple absolute time records.

[0015] After obtaining the aforementioned set, the system needs to determine the freshest data anchor point for the target identifier within the overall network view. Since the underlying timestamps are incrementing in seconds or milliseconds starting from a predetermined epoch, their absolute magnitude directly reflects the chronological order of events. The processor unit compares and filters the data sequence within the local modification timestamp set, removing smaller, older timestamp values ​​and selecting the largest one. This maximum value represents the most recent data entry or modification action across all heterogeneous data sources associated with the target identifier, and the system uses this timestamp as the last update time value for the target identifier.

[0016] To prevent time drift or tampering errors in the local physical hardware clock of the current computing node, the system does not directly call the local time of the computing server. Instead, it initiates a synchronization request to the system's central clock unit through the internal network communication bus. The central clock unit maintains a globally synchronized time sequence, and the system receives the standard timeline feedback packets returned by it, using the parsed values ​​as the current system time.

[0017] After obtaining the data from the two key time points mentioned above, the system calls the arithmetic logic operation module to perform an arithmetic difference operation. Let's assume the obtained current system time value is denoted as... The last update time value that has been selected is denoted as The computational logic involves calculating the algebraic difference between the two. (System computation) minus The result generates a sequence of absolute differences representing the time span. Considering that the timestamp differences stored at the computer's underlying level are typically large integers in the millisecond range, the system further divides this absolute difference sequence by a unit time conversion factor to convert its dimensions to hours or days. The converted value is output and defined as the silent duration parameter, which physically represents the length of the actual gap / vacuum period experienced by the target identifier since the last capture of an external heterogeneous data update.

[0018] Retrieve the multi-source historical merge logs of the target identifier, and use the cumulative number of attribute alignment failures as the merge conflict count; Capture the number of parsing requests for the target identifier within a preset time window, and calculate the average number of parsing requests per unit time as the query frequency; Traverse the unified identifier association graph and extract the longest hierarchical link value that is referenced by other identifiers with the target identifier as the endpoint as the dependency depth; Preferably, the step of traversing the unified identifier association graph and extracting the longest hierarchical link value referenced by other identifiers with the target identifier as the endpoint as the dependency depth includes: In the unified identifier association graph, the target identifier is set as the starting vertex of the search, and a breadth-first backtracking search based on directed edges is performed; Establish an initial depth meter and set the initial measurement value to zero when the search is triggered; Each time a directed edge is used to address the preceding associated node that directly calls the target identifier at a layer, the initial measurement value is accumulated by a preset cross-layer increment to generate the current branch depth value, and this process is gradually advanced until the edge terminal node that is no longer called by other nodes is reached. Collect the current branch depth value generated by all branches that are successfully traced back to the edge terminal node; Compare the current branch depth values ​​of all collections, and select the largest scalar value to assign to the dependency depth.

[0019] In the data processing flow, the system needs to quantitatively evaluate the quality loss during the merging of multi-source data flows towards a unified identifier. The underlying data synchronization component of the system calls the log files stored in the persistent database to retrieve and extract merge event entries containing the target identifier's identification code. For the extracted entries, the system scans the status feedback field in the message body and uses a filtering function to filter out records carrying error codes for data format mismatch, field length exceeding limits, or primary key conflicts. These filtered abnormal record entries are summed and counted, and the integer result is directly used as the merge conflict count. This value directly reflects the frequency of obstruction when external heterogeneous sources attempt to update data to this identifier.

[0020] Building upon this, the system connects to the access log of the underlying routing gateway through a traffic monitoring component. It extracts a fixed-duration span from the current time point back in time, using this as a preset time window. The system then counts the total number of commands from all downstream application subsystems that initiate data parsing or reading actions towards the target identifier within this time span, using this total number as the parsing request volume. Subsequently, the system performs an arithmetic division operation, dividing this parsing request volume by the number of standard time units contained within the preset time window. For example, dividing the total number of calls by the corresponding span of hours calculates the average request occurrence rate per unit time. This quotient is output as the query frequency, used to characterize the read frequency of data for this identifier by downstream businesses.

[0021] To further evaluate the structural weight of the target identifier in the global data network topology, the system needs to parse the internally maintained unified identifier association graph. This graph resides in memory as an adjacency list data structure, where each identifier is a graph node, and the data call relationships between identifiers are directed edges, with the arrows pointing to the underlying data being called. The system extracts the memory address pointer corresponding to the target identifier in the graph and sets it as the starting vertex for the backtracking algorithm. Next, the processor allocates an integer register space in memory as a depth meter and assigns it an initial value of zero.

[0022] The system then initiates a breadth-first backtracking search process based on directed edges. The search unfolds hierarchically, first addressing along the reverse direction of the directed edges to find all first-level preceding related nodes that directly use the target identifier as a data source. Each time the search process advances outward by a relation level, the system adds a cross-level increment of one to the initial measurement value, thus generating a current branch depth value representing the current level. The search process continues to initiate reverse addressing towards outermost levels, starting from the first-level preceding related nodes, and so on in a cyclical manner. When addressing a node, if it is found that there are no in-degree reference edges pointing to it in the adjacency list, the system determines that the node is an edge terminal node not called by other nodes, and the unidirectional branch backtracking search terminates. At this time, the accumulated current branch depth value of this branch is stored in a temporary array.

[0023] Due to the complex mesh-like call structure, a complete graph traversal generates multiple backtracking branch paths of varying lengths. After all branches reach the edge terminal nodes, the system collects and traverses the temporary array storing the depths of each branch, performing a scalar comparison operation. The system selects the scalar with the largest value from all current branch depth values ​​and assigns this largest scalar to the dependency depth parameter. By performing this graph structure search and cumulative comparison operation, the system completes the dimensionality reduction processing of the multidimensional referencing mesh relationship, transforming the topological association form into specific integer scalar parameters.

[0024] Detecting the current network heartbeat value of the underlying physical data source corresponding to the target identifier, and converting the current network heartbeat value into a source-server connectivity state, includes: Send a network diagnostic control message protocol echo request packet to the registered network protocol address of the underlying physical data source bound to the target identifier; Start the timing receiving program, and summarize the network latency in milliseconds and the transmission packet loss rate recorded in the returned packets as the current network heartbeat value; When the network latency in milliseconds is less than the preset latency defense value and the transmission packet loss rate is equal to the preset lower limit of zero, the source station connectivity state is assigned a preset normal weight constant. When the network latency in milliseconds is greater than or equal to the preset latency defense value, and the packet loss rate is greater than the preset lower limit of zero, the source station connectivity state is assigned a preset degradation constant, and the conversion calculation to the source station connectivity state is completed.

[0025] During the network communication link state assessment phase, the system's main control component needs to quantitatively determine the current physical responsiveness of the target data source. The processor parses the target identifier's registration configuration file, reading the registered network protocol address of the underlying physical data source it is bound to. This address typically appears as a standard Internet Protocol endpoint identifier pointing to the source database server. After confirming the address, the processor calls the operating system's underlying network protocol stack interface to construct a set of fixed-length network control message protocol echo request packets. The system, through the network device driver, continuously sends this set of request packets to the registered network protocol address. At the moment the underlying driver module pushes each request packet to the physical network card, it reads the system clock to obtain the corresponding packet sending timestamp and stores it in a memory queue.

[0026] Along with the sending action, the system synchronously starts a timing reception program, instructing the network card to listen for echo response data packets from the data source endpoint. Whenever the receive queue captures a matching response packet, the system immediately generates a receive timestamp. The processor performs algebraic subtraction on the receive and send timestamps of the same sequence number data packet in the memory processing area to calculate the round-trip time of that packet in the current physical link. By integrating all the time-consuming data within this probe task and calculating the algebraic average, the system obtains the network latency in milliseconds for that address link at this moment. The system calculates the total number of data packets actually sent and the total number of successfully received response packets in this probe command, and uses the absolute difference between the total number sent and the total number received, divided by the initial total number of sent packets, to obtain the packet loss rate for this process. The network latency in milliseconds and the packet loss rate, obtained through the above statistical calculations and stored in memory, together constitute the current network heartbeat value parameter describing the physical connectivity quality.

[0027] To convert the acquired physical network parameters into mathematical factors that can be directly used in subsequent topology connectivity calculations, the system needs to perform hard thresholding and discrete mapping on the heartbeat parameter. The processor reads the operation rule configuration file, extracts the pre-set latency protection value used to define the boundary of network lag-free operation, and loads the lower limit of zero, representing no data loss. The arithmetic logic unit retrieves the previously generated current network heartbeat value and performs a dual-condition parallel check. When it is determined that the acquired network latency in milliseconds is less than the preset latency protection value, and the recorded packet loss rate is equal to the preset lower limit of zero, the system confirms that the communication channel to the source is in a high-availability state. The processor then performs an assignment operation on the variable segment representing the source station's connectivity state in the memory address, writing a preset normal weight constant into it. Such constants are usually assigned a value of one in engineering integer operations to ensure that the relevant topology weights are not reduced in subsequent multiplication calculation instructions.

[0028] Under the opposite boundary conditions, when the arithmetic logic unit's judgment result shows that the network latency in milliseconds is greater than or equal to the preset latency defense value, and the detected packet loss rate is greater than the preset lower limit of zero, the system's underlying layer determines that there is severe network congestion or backlog in the source system's processing queue for the current data capture channel targeting this underlying physical data source. To prevent continuous request congestion in upper-layer services, the processor overwrites the source station connectivity state value in the memory variable with a preset degradation constant. This degradation constant is generally preset to a positive floating-point value less than one, used to proactively reduce the credibility of the data source corresponding to this identifier at the mathematical level. At this point, the variable overwriting action ends, the system terminates the network-level protocol probing, completes the conversion calculation process from the physical quality indicators of the communication link to the source station connectivity state logic constant, and outputs stable parameter values.

[0029] The attenuation index is calculated based on the silence duration and the number of merging conflicts; Preferably, the step of calculating the attenuation index based on the silence duration and the number of merging conflicts includes: Extract the preset conflict penalty coefficient; Multiply the number of merging conflicts by the preset conflict penalty coefficient to obtain the conflict penalty multiplier; Adding a preset reference unit of one to the conflict penalty multiplier yields the amplification factor; The attenuation index is calculated by multiplying the silence duration by the amplification factor.

[0030] In assessing the degree of quality degradation of unified identifier data, the system needs to comprehensively consider the erosion effect of time-dimension gaps and the frequency of anomalies during data access on the overall usability of the identifier. The processor first sends an addressing instruction to the system's constant configuration management module to read the preset conflict penalty coefficient pre-loaded in the memory configuration file. This coefficient is typically set to a positive decimal floating-point constant, primarily used at the system level to measure the negative deterioration weight of each underlying data alignment failure on the overall health of the identifier.

[0031] After obtaining the penalty coefficient, the system's underlying arithmetic logic unit retrieves the number of merged conflicts previously statistically recorded in the status register. The processor performs a conventional algebraic multiplication operation on this integer frequency parameter and the aforementioned floating-point preset conflict penalty coefficient, generating a value reflecting the overall scale of abnormal backlog impact, which the system defines as the conflict penalty multiplier. Considering that underlying data interactions may have a normal intake state with zero conflicts within a specific period, the calculated result of this conflict penalty multiplier is zero. If this zero-value parameter is directly used as a factor in subsequent multiplication operations, the final calculated attenuation index will be forced to zero, thus completely masking the natural data aging attribute brought about by the passage of time. To prevent such operational truncation, the processor calls a preset base unit of one in the arithmetic register and performs an algebraic addition operation on it and the just-calculated conflict penalty multiplier. Through this linear translation process with a raised base, the system transforms the original absolute penalty amount, calculated from zero, into an amplification factor that increases from one.

[0032] After the amplification factor is constructed in memory, the system calls the computation process to retrieve the silent duration parameter generated in the previous steps from the parameter cache. At this point, the processor multiplies the silent duration, representing the physical time span, with the amplification factor, which serves as a scaling parameter, in a multiplier. In this data transformation logic, since the amplification factor is always greater than or equal to one, it mathematically exacerbates the time decay effect. When a target identifier frequently experiences source rejection or field errors during historical merging and aggregation, its amplification factor value will show a significant increase, resulting in a much larger numerical inflation after the same silent duration is multiplied. The final product obtained from the multiplication operation is output by the processor and defined as the decay exponent. The system then writes this decay exponent into the corresponding byte of the target identifier's current memory state snapshot, retaining it as a core quantitative indicator representing the overall degradation degree of the identifier.

[0033] Calculate the topology support index based on the query frequency, the dependency depth, and the source site connectivity state; Preferably, the step of calculating the topology support index based on the query frequency, the dependency depth, and the origin server connectivity includes: The basic call load is obtained by adding the query frequency to the dependency depth; Read the values ​​in the source station's connectivity state; The topology support index is calculated by multiplying the basic call load by the value in the source station connectivity state.

[0034] When evaluating the structural weight and data supply capacity of a target identifier within the overall system network architecture, the processor needs to fuse the read activity of downstream services with the dependency levels in the graph topology. The system memory controller issues read instructions to the state buffer, extracting the query frequency and dependency depth parameters generated in previous calculations. Query frequency numerically reflects the number of direct read accesses per unit time, while dependency depth represents the maximum hierarchical span of reverse calls in the association graph. To comprehensively quantify the overall downstream service pressure currently borne by the identifier, the system's underlying arithmetic logic unit retrieves these two values ​​and performs algebraic addition. In this calculation logic, both query frequency and dependency depth are stripped of their original physical dimensions and treated as dimensionless scalar factors in the merging process. The direct numerical sum of the two is stored in a temporary register and defined as the basic call payload. This payload value reflects the theoretical maximum data supply expectation that the system network topology holds for the target identifier.

[0035] After deriving the theoretical basic call load, the system's operational logic must incorporate the actual supply capacity of the underlying physical data source as a constraint. The processor executes a memory read instruction, addresses the variable memory block allocated in the preceding network link state diagnostic step, and reads the current recorded value of the source station's connectivity state. According to the preceding network heartbeat detection logic, this value is a pre-mapped floating-point constant, whose actual value can be either a normal weight constant that maintains the original weights, or a degradation constant representing network congestion and packet loss.

[0036] After the above parameters are prepared, the processor synchronously loads the basic call payload temporarily stored in the register, along with the source station connectivity value just extracted, into the multiplier to perform algebraic multiplication. The underlying business logic of this multiplication operation is to use the real-time link communication quality of the underlying physical network card to calculate the theoretical topology payload. When the underlying physical data source network is unobstructed, the source station connectivity value is usually one, and the basic call payload is preserved without loss; however, when the physical link deteriorates, the source station connectivity value exhibits a degradation coefficient of less than one, and the multiplication operation will proportionally reduce the value of the basic call payload. This calculation mechanism objectively reflects, mathematically, that when the physical channel is blocked, even if there is a large theoretical call demand downstream, the actual topology support capability that the identifier node can provide will be significantly reduced. The final product result calculated by the multiplier is output as the topology support index. The system then updates the core attribute data table of the target identifier with this index as a quantitative judgment basis for subsequently determining whether the identifier data needs to be forcibly refreshed or cached.

[0037] The health status of the target identifier is calculated by dividing the topological support index by the sum of the attenuation index and the preset base value. Preferably, performing route connection state recovery based on the health recalculation action includes: Extract the current system time at the moment the health recalculation action is triggered; The duration of silence at this moment is recalculated based on the current system time; Based on the re-extracted silence duration and the source station connectivity state equal to the preset network connectivity benchmark value, the latest health status is obtained through exponential recombination. When the latest health status is greater than or equal to the preset health threshold, the memory routing state parameters are restored to the original read / write state, and the interception action for the cross-network penetration verification request is canceled, thus completing the routing connection state restoration.

[0038] After summarizing the various runtime parameters of the target identifier, the system needs to output a final quantitative indicator that comprehensively reflects the current overall availability status of the identifier. The processor retrieves the topology support index and attenuation index generated by the previous process from the temporary memory processing area. To prevent the attenuation index from approaching zero and causing a division-by-zero overflow error in the underlying system under certain extreme data scenarios, the system pre-sets a preset base value in the constant configuration table, which is usually defined as the real number one. The arithmetic logic unit performs algebraic addition on the read attenuation index and the preset base value to construct a base number as the denominator of the operation. Subsequently, the processor uses the topology support index as the numerator and divides it by the base number to perform a floating-point division operation. This division calculation logic mathematically establishes an inverse attenuation relationship, that is, the stronger the network support capability of the target identifier, the higher the calculated quotient; and the more severe the attenuation over time, the lower the quotient. The final calculated quotient is output by the system and defined as the health status of the target identifier, serving as the core judgment basis for subsequent routing status management.

[0039] When the system detects that network jitter has subsided or receives a wake-up command from an external system, the main control program triggers a health recalculation for the target identifier. In response to this trigger, the processor's clock management module immediately captures the current system time at the moment of triggering. This is typically achieved by calling the operating system kernel's timestamp interface to obtain a long integer sequence representing the current absolute time. The processor retrieves the timestamp of the last status update recorded in the persistent log for this target identifier, subtracts this last update timestamp from the current system time, and calculates the new time span from the last update pause until the present moment. The system assigns this span as the newly calculated silence duration. This step ensures that the decay calculation system can incorporate the latest elapsed time cost in real time.

[0040] After obtaining the latest time parameters, the system executes the parameter variable replacement and exponent reorganization logic. For network physical communication status parameters, the system assumes this recalculation is based on the premise that the network link is already connected. Therefore, the processor directly overwrites the source station connectivity value in memory with a preset network connectivity baseline value. This baseline value is generally set to one to represent that the physical link is currently free of packet drop obstacles. The processor substitutes the updated source station connectivity value into the preceding multiplication operation node to regenerate the supporting parameters, and simultaneously substitutes the newly extracted current silence duration into the multiplication operation to generate a new attenuation exponent. By reusing the established addition and division calculation flow, the processor performs algebraic reorganization and solution on the regenerated underlying exponents, ultimately outputting a calculated quotient value that reflects the latest physical and temporal state of the target identifier. This quotient value is recorded in the system as the latest health status indicator.

[0041] After obtaining the latest health status, the system's logic comparison component retrieves the pre-defined business security operation baseline value, i.e., the preset health threshold, from the memory configuration file. The arithmetic logic unit compares the latest health status with this preset health threshold. When the determination result shows that the latest health status is greater than or equal to the preset health threshold, the system confirms that after network channel recovery, even with the added extended silence time decay, the overall usability of the identifier has returned to the safe operating range. At this time, the processor addresses the memory control block responsible for routing and forwarding the data request, and overwrites the memory routing state parameters that were previously marked as blocked or isolated due to substandard health status, restoring them to their original read-write state. Simultaneously, the underlying network gateway component of the system issues a rule cleanup command, canceling the access interception action previously set for external systems initiating cross-network penetration verification requests to this target identifier. The system restores the network communication pipeline that was temporarily suspended due to the anomaly, thus completing the entire routing connection state recovery process.

[0042] When the health status of the vital signs is greater than or equal to a preset health threshold, a normal maintenance instruction is generated; When the physical health status is less than the preset health threshold, the topological support index is compared with the preset dependency threshold.

[0043] When the topology support index is greater than or equal to the preset dependency threshold, a shadow route degradation instruction is generated; Preferably, performing shadow state polling transformation based on the shadow routing degradation instruction includes: According to the shadow route degradation instruction, the memory route status parameter of the target identifier is changed to shadow read-only state; The system intercepts the cross-network penetration verification request initiated by the target identifier to the underlying physical data source, and returns a historical cache snapshot of the target identifier while blocking the addressing request to the source end; Initiate a detection loop with a preset maximum number of rounds; In each round of the detection loop, the current network heartbeat value is re-detected at a preset time interval and the value of the source station connectivity state is updated; When the updated source station connectivity value equals the preset network connectivity baseline value, a health recalculation action is triggered to exit the detection loop; When the preset upper limit round is reached and the updated source station connectivity value is less than the preset network connectivity benchmark value, the topology support index is set to zero, and the shadow state polling conversion is completed.

[0044] After obtaining the current health status of the target identifier, the underlying logical comparator of the system retrieves a preset health threshold pre-written in the system environment variable configuration table. This threshold represents the minimum performance limit that the system architecture can tolerate to ensure normal business data flow. The processor compares the calculated health status with this preset health threshold. When the comparison result shows that the health status is greater than or equal to the preset health threshold, it indicates that the data quality and network supply capacity of the current target identifier are within a safe operating range. The system's main control module then issues a normal maintenance command to the routing control center to maintain the current data access and distribution channels of the identifier in an unobstructed normal operating state.

[0045] If the comparison result of the logic comparator shows that the health status is less than the preset health threshold, it indicates that the data supply capacity of the identifier has substantially declined. At this point, the system cannot simply cut off the data link directly, as it needs to prevent downstream core services from experiencing cascading failures due to data interruption. Therefore, the system extracts the previously calculated topology support index and retrieves the preset dependency threshold from memory for a second comparison. This preset dependency threshold quantifies the minimum tolerance of downstream services for the target identifier's data dependency. When the topology support index is greater than or equal to the preset dependency threshold, it means that although the health of the identifier's source end is poor, the downstream service network still has a strong demand for data access. To achieve a balance between ensuring uninterrupted downstream services and preventing overload of the upstream source end, the system's main control module generates a shadow route degradation instruction, preparing to execute a degradation compensation mechanism.

[0046] Based on the shadow routing degradation instruction transmitted via the system's internal bus, the processor first addresses the memory-mapped block of the target identifier in the routing control table and overwrites its original regular read-write memory routing status parameters to a shadow read-only state. This state change fundamentally locks out any new data writes or forced synchronization permissions initiated against this identifier. Subsequently, the system's network gateway component synchronously updates the access control list. When the gateway receives a cross-network penetration verification request from an external or downstream system seeking the underlying physical data source for this target identifier, the gateway recognizes that the identifier has been marked as a shadow read-only state. At this point, the gateway's underlying protocol stack directly discards or intercepts the outbound data packet, blocking its transmission to the source via the physical network card. Simultaneously, the gateway's redirection module points the request's read pointer to the system's local persistent storage area, extracts the historical cache snapshot of the target identifier saved during its most recent healthy state, encapsulates the snapshot data according to the original request protocol format, and returns it as a response message to the caller. Through this interception and response mechanism, the system can still provide read-only fallback data services even during physical source station anomalies.

[0047] While maintaining the shadow proxy state, the system must continuously probe the recovery status of the underlying source station. The processor's task scheduling module creates a separate thread in the background and initializes a pre-set upper limit count value in the thread's register, thus initiating the detection loop. This upper limit count is usually set to a fixed integer to prevent the system from endlessly consuming resources to probe a permanently failed physical node. In each iteration of the detection loop, the system clock component forces the thread into a sleep state until it is awakened after a preset time interval. After waking up, the underlying network interface controller sends a lightweight network heartbeat probe packet to the physical source station where the target identifier is located. Based on the round-trip latency and packet loss rate of the probe packet, the system recalculates the current network heartbeat value and overwrites the mapped latest value at the address of the source station's connectivity variable in memory.

[0048] After a single state update, the logic comparator checks the updated source connectivity value against the system's preset network connectivity benchmark. If the updated source connectivity value equals the benchmark, it indicates that the underlying physical network congestion or fault has been eliminated and the link has returned to normal. At this point, the scheduling module sends an interrupt signal to terminate the current detection loop early and triggers a health recalculation to the main computing module, allowing the system to reassess and restore routing connections based on the recovered network status. Conversely, if the loop iteration count register continuously decreases until it reaches zero (reaching the preset maximum number of iterations), and the logic comparator confirms that the last updated source connectivity value is still less than the preset network connectivity benchmark, the system determines that the underlying physical data source is in a deep downtime or long-term unreachable state. To prevent this failed node from continuing to occupy scheduling weight in the topology network, the processor forcibly executes a memory assignment instruction, directly clearing the topology support index corresponding to the target identifier to zero. This zeroing operation completely severs the theoretical data support for the downstream architecture from the mathematical logic of the node. Based on this, the system completes the shadow state polling transformation for the failed node, which will trigger the corresponding node removal or alarm process.

[0049] When the topology support index is less than the preset dependency threshold and the dependency depth is greater than the preset minimum depth, a tombstone blocking instruction is generated. Preferably, the empty shell fast fuse operation is performed based on the tombstone blocking command, including: Based on the tombstone blocking instruction, retrieve the storage addresses of all heterogeneous attribute data payloads mounted on the target identifier in the storage database; The address segment erase operation instruction is issued to clear the heterogeneous attribute data payload and set the memory bytes occupied by the data structure containing the target identifier to zero; The unique identifier field of the target identifier is retained in the global index table; Modify the system interaction identification number of the target identifier to a data extinction code; When the system bus receives a retrieval application layer instruction that parses the character segment corresponding to the unique identifier, it directly extracts the data extinction code as an outbound signal source and sends it back to the retrieval party, terminating the join table retrieval process for the storage database and completing the empty shell fast circuit breaker operation.

[0050] During system operation, the logic comparator of the main control program continuously extracts the current topology support index and the dependency depth obtained from previous statistics for the target identifier. The processor retrieves the preset dependency threshold and the preset minimum depth from the constant area of ​​the system configuration file. The arithmetic logic unit performs a dual conditional judgment on these two sets of parameters. When the topology support index is less than the preset dependency threshold, it indicates that the data supply value of the target identifier node to downstream services has decayed to a negligible level; at the same time, the system checks whether the dependency depth is greater than the preset minimum depth. This minimum value is set to identify the node's position in the call chain. If the dependency depth is greater than this value, it means that the identifier node is at a deeper level in the call tree. If left unchecked, each time the downstream system initiates an address lookup for it, it will need to traverse multiple intermediate nodes, resulting in extremely high invalid traversal overhead. When both conditions are met simultaneously through logical AND operation, the system determines that the node is a high-overhead and valueless redundant node, and the main control module then generates a tombstone blocking instruction to initiate the forced resource reclamation and isolation process for the node.

[0051] In response to a tombstone blocking command transmitted via the system bus, the storage management component first extracts the metadata information of the target identifier from memory. By traversing the pointer mapping relationships in the metadata, the storage controller retrieves the specific physical or logical storage addresses of all heterogeneous attribute data payloads attached to the target identifier from the underlying storage database. These payloads typically occupy a large amount of storage space, such as attached text records or status logs. After confirming the address range, the storage controller issues an address segment erase operation command to the underlying file system or database engine. Upon receiving the command, the database engine releases the file occupancy lock of the aforementioned address blocks, marking them as free and writable, thereby clearing the heterogeneous attribute data payloads. Simultaneously, the memory controller performs an underlying memory overwrite operation, setting all contiguous memory bytes occupied by the data structure corresponding to the target identifier in the runtime memory to zero. By clearing persistent data and emptying runtime memory, the system strips away all substantial data content of the identifier, degrading it to an empty shell state.

[0052] After physically clearing the payload data, the system must guard against an extreme scenario: downstream services might unknowingly send frequent requests to the cleared node, causing queries to penetrate directly to the physical database layer and trigger invalid disk I / O operations. To prevent this penetration effect, the system's index management module deliberately bypasses the node's key-value pair record in the global index table during routine cleanup. Specifically, the system retains the unique identifier field of the target identifier as the query key in the global index table, which is based on a hash or tree structure. While retaining the key value, the memory controller addresses the value storage area corresponding to the key, erases the system interaction identifier that originally pointed to the memory address of the actual data structure, and replaces it with a data extinction code extracted from a dictionary pre-defined in the system kernel. This extinction code is a static constant identifier with a specific byte length, specifically used to declare to the system that the corresponding data has been permanently logically removed.

[0053] After the above modifications, the target identifier is essentially transformed into a system tombstone that blocks subsequent queries. In subsequent business cycles, when the network gateway receives a retrieval application layer instruction initiated by an external application service via the system bus, and this instruction aims to parse data corresponding to the aforementioned unique identifier character field, the gateway's routing resolution component will first query the global index table in memory. At this time, the index query will successfully hit the deliberately reserved identifier character field. When the processor reads its corresponding value, it will directly obtain the overwritten data extinction code, rather than a normal memory pointer. Once the circuit breaker component at the system's underlying level recognizes that the outbound data type is an extinction code, it will immediately trigger the interception logic. The circuit breaker directly encapsulates the data extinction code in a standard response message body and sends it back to the original retrieval party as an outbound signal source, informing it that the data has been invalidated. At the same time as sending this response, the circuit breaker sends a termination signal to the database connection pool, forcibly stopping the join table retrieval process that was originally intended to be initiated to the underlying storage database. By using this set of execution logic that truncates and directly returns to the destroyed state, the system completely avoids the unnecessary consumption of the underlying database for invalid data, thus completing the empty shell fast circuit breaker operation.

[0054] When the topology support index is less than the preset dependency threshold and the dependency depth is equal to the preset minimum depth, a physical erase command is generated.

[0055] Preferably, the physical destruction of the memory root node based on the physical erase instruction includes: Deconstruct the physical erase command to extract the controlled target identifier sequence it carries; Retrieve the physical root node memory block address pointer that is aligned with the sequence of the controlled target identifier in the global inverted hierarchical index table; Issue a cut-off command to sever the routing chain between all adjacent active nodes and the memory block address pointer of the physical root node; The kernel space manager is invoked to overwrite and initialize the hard storage address corresponding to the physical root node memory block address pointer to an unallocated, available, empty address, thus completing the physical destruction of the memory root node.

[0056] The main control program's logic processing unit continuously interfaces with the node status parameters returned from the system bus, extracting the target identifier's current topology support index and dependency depth. The arithmetic logic unit reads the preset dependency threshold and preset minimum depth from the basic configuration table. The system executes the conditional branch's evaluation logic, first determining whether the topology support index is less than the preset dependency threshold to confirm whether the node has lost its substantial data supply value. Next, the system compares the dependency depth with the preset minimum depth. When both values ​​are exactly equal, it means the node is at the top of the entire mesh call chain, i.e., the root node directly exposed as the business entry point. In the previous scenario, the disappearance of deep nodes would cause requests to penetrate the call chain ineffectively. Currently, there are no extended dependency call paths attached below this root node, and external requests will terminate and return directly at the entry point after failure. Based on this physical topology characteristic, the system determines that there is no need to reserve a placeholder shell for intercepting requests and can perform a complete recycling action. After the logic AND gate confirms that the above two conditions are met simultaneously, the system kernel's scheduling module generates a physical erase instruction, preparing to completely destroy the node at the underlying hardware level.

[0057] Upon receiving the physical erase instruction forwarded from the bus, the memory management subsystem begins the destructuring process. The processor, following the system's internal communication protocol, reads data at a specific offset in the instruction message payload area and extracts a string of bytecode representing the unique identity of the node to be processed—the controlled target identifier sequence. After obtaining this sequence, the memory controller accesses the global inverted index table maintained by the system kernel. Unlike a forward lookup table, this inverted index table uses the hash digest of the identifier as the key to reverse-map the underlying physical storage location. The system substitutes the extracted identifier sequence into a hash function for calculation, using the calculated hash value to match the corresponding entry in the index table, and then reads the physical root node memory block address pointer bound to that entry. This pointer data not only contains the starting address but also the memory block allocation span, allowing the system to precisely pinpoint the absolute physical location of the target node in running memory.

[0058] Before directly clearing the physical memory, the system must first sever the node's network communication layer connections to prevent other normally operating nodes from continuing to send data to that address, thus causing memory access violations or dangling pointer errors. The routing control component traverses the topology connection mapping table in memory, retrieving all adjacent active nodes that contain the physical root node's memory block address pointer in the next-hop addressing cache. The routing component then generates and sends status update messages in the form of cut-off instructions to these active nodes. Upon receiving the message, the active nodes forcibly refresh their local routing table, removing the routing entry pointing to that physical address, thereby logically severing all routing chains pointing to that address pointer to isolate the target region. After completing the topology isolation, the processor enters kernel mode via an interrupt call and wakes up the kernel space manager. The kernel space manager takes over control of the memory region, sending a low-level overwrite signal to the corresponding memory control unit based on the locked physical root node memory block address pointer, using a continuous stream of empty bytes to overwrite the original storage structure's level state to erase the service data. The space manager then locates the region in the operating system's memory allocation bitmap and flips the corresponding status flag from a high occupied state to a low available state. Through this coordinated operation of physical overwriting and status initialization, the system releases the address region back into the free memory pool, thereby completing the physical destruction of the entire memory root node.

[0059] It should be noted that the interval and threshold sizes are set for ease of comparison. The size of the threshold depends on the amount of sample data and the base number set by those skilled in the art for each set of sample data, as long as it does not affect the proportional relationship between the parameter and the quantized value. Furthermore, the above formulas are all dimensionless calculations, and the formulas are derived from software simulations using a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0060] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0061] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

[0062] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A unified identifier generation and management method for multi-source heterogeneous data, characterized in that, include: Extract the target identifier from the unified identifier mapping table; Obtain the last update time of the data source corresponding to the target identifier, calculate the time difference between the current system time and the last update time, and obtain the silence duration; Retrieve the multi-source historical merge logs of the target identifier, and use the cumulative number of attribute alignment failures as the merge conflict count; Capture the number of parsing requests for the target identifier within a preset time window, and calculate the average number of parsing requests per unit time as the query frequency; Traverse the unified identifier association graph and extract the longest hierarchical link value that is referenced by other identifiers with the target identifier as the endpoint as the dependency depth; Detect the current network heartbeat value of the underlying physical data source corresponding to the target identifier, and convert the current network heartbeat value into the source station connectivity state; The attenuation index is calculated based on the silence duration and the number of merging conflicts; Calculate the topology support index based on the query frequency, the dependency depth, and the source site connectivity state; The health status of the target identifier is calculated by dividing the topological support index by the sum of the attenuation index and the preset base value. When the health status of the vital signs is greater than or equal to a preset health threshold, a normal maintenance instruction is generated; When the physical health status is less than the preset health threshold, compare the topological support index with the preset dependency threshold. When the topology support index is greater than or equal to the preset dependency threshold, a shadow route degradation instruction is generated; When the topology support index is less than the preset dependency threshold and the dependency depth is greater than the preset minimum depth, a tombstone blocking instruction is generated. When the topology support index is less than the preset dependency threshold and the dependency depth is equal to the preset minimum depth, a physical erase command is generated.

2. The method according to claim 1, characterized in that, The calculation of the attenuation index based on the silence duration and the number of merging conflicts includes: Extract the preset conflict penalty coefficient; Multiply the number of merging conflicts by the preset conflict penalty coefficient to obtain the conflict penalty multiplier; Adding a preset reference unit of one to the conflict penalty multiplier yields the amplification factor; The attenuation index is calculated by multiplying the silence duration by the amplification factor.

3. The method according to claim 1, characterized in that, The calculation of the topology support index based on the query frequency, the dependency depth, and the origin server connectivity includes: The basic call load is obtained by adding the query frequency to the dependency depth; Read the values ​​from the source station's connectivity state; The topology support index is calculated by multiplying the basic call load by the value in the source station connectivity state.

4. The method according to claim 1, characterized in that, Based on the shadow routing degradation instruction, shadow state polling transformation is performed, including: According to the shadow route degradation instruction, the memory route status parameter of the target identifier is changed to shadow read-only state; The system intercepts the cross-network penetration verification request initiated by the target identifier to the underlying physical data source, and returns a historical cache snapshot of the target identifier while blocking the addressing request to the source end; Initiate a detection loop with a preset maximum number of rounds; In each round of the detection loop, the current network heartbeat value is re-detected at a preset time interval and the value of the source station connectivity state is updated; When the updated source station connectivity value equals the preset network connectivity baseline value, a health recalculation action is triggered to exit the detection loop; When the preset upper limit round is reached and the updated source station connectivity value is less than the preset network connectivity benchmark value, the topology support index is set to zero, and the shadow state polling conversion is completed.

5. The method according to claim 4, characterized in that, Based on the health recalculation action, route connectivity state recovery is performed, including: Extract the current system time at the moment the health recalculation action is triggered; The duration of silence at this moment is recalculated based on the current system time; Based on the re-extracted silence duration and the source station connectivity state equal to the preset network connectivity benchmark value, the latest health status is obtained through exponential recombination. When the latest health status is greater than or equal to the preset health threshold, the memory routing state parameters are restored to the original read / write state, and the interception action for the cross-network penetration verification request is canceled, thus completing the routing connection state restoration.

6. The method according to claim 1, characterized in that, Executing a fast fuse operation based on the tombstone blocking command includes: Based on the tombstone blocking instruction, retrieve the storage addresses of all heterogeneous attribute data payloads mounted on the target identifier in the storage database; The address segment erase operation instruction is issued to clear the heterogeneous attribute data payload and set the memory bytes occupied by the data structure containing the target identifier to zero; The unique identifier field of the target identifier is retained in the global index table; Modify the system interaction identification number of the target identifier to a data extinction code; When the system bus receives a retrieval application layer instruction that parses the character segment corresponding to the unique identifier, it directly extracts the data extinction code as an outbound signal source and sends it back to the retrieval party, terminating the join table retrieval process for the storage database and completing the empty shell fast circuit breaker operation.

7. The method according to claim 1, characterized in that, The step of obtaining the last update time of the data source corresponding to the target identifier, calculating the time difference between the current system time and the last update time, and obtaining the silence duration includes: Parse the source merging trajectory column recorded in the unified identifier mapping table; Extract in batches the set of local modification timestamps of each heterogeneous sub-identifier belonging to different source data tables from the source merged trajectory column; Compare the values ​​in the set of local modification timestamps, and select the timestamp with the largest value as the last update time value; A synchronization request is sent to the system's central clock unit to obtain a real-time stable timestamp as the current system time; Solve for the absolute difference sequence formed by subtracting the last update time from the current system time, and output the absolute difference sequence as the silence duration.

8. The method according to claim 1, characterized in that, The process of traversing the unified identifier association graph and extracting the longest hierarchical link value referenced by other identifiers with the target identifier as the endpoint as the dependency depth includes: In the unified identifier association graph, the target identifier is set as the starting vertex of the search, and a breadth-first backtracking search based on directed edges is performed; Establish an initial depth meter and set the initial measurement value to zero when the search is triggered; Each time a directed edge is used to address the preceding associated node that directly calls the target identifier at a layer, the initial measurement value is accumulated by a preset cross-layer increment to generate the current branch depth value, and this process is gradually advanced until the edge terminal node that is no longer called by other nodes is reached. Collect the current branch depth value generated by all branches that are successfully traced back to the edge terminal node; Compare the current branch depth values ​​of all collections, and select the largest scalar value to assign to the dependency depth.

9. The method according to claim 1, characterized in that, The step of detecting the current network heartbeat value of the underlying physical data source corresponding to the target identifier and converting the current network heartbeat value into the source station connectivity state includes: Send a network diagnostic control message protocol echo request packet to the registered network protocol address of the underlying physical data source bound to the target identifier; Start the timing receiving program, and summarize the network latency in milliseconds and the transmission packet loss rate recorded in the returned packets as the current network heartbeat value; When the network latency in milliseconds is less than the preset latency defense value and the transmission packet loss rate is equal to the preset lower limit of zero, the source station connectivity state is assigned a preset normal weight constant. When the network latency in milliseconds is greater than or equal to the preset latency defense value, and the packet loss rate is greater than the preset lower limit of zero, the source station connectivity state is assigned a preset degradation constant, and the conversion calculation to the source station connectivity state is completed.

10. The method according to claim 1, characterized in that, The physical destruction of the memory root node is performed based on the physical erase instruction, including: Deconstruct the physical erase command to extract the controlled target identifier sequence it carries; Retrieve the physical root node memory block address pointer that is aligned with the sequence of the controlled target identifier in the global inverted hierarchical index table; Issue a cut-off command to sever the routing chain between all adjacent active nodes and the memory block address pointer of the physical root node; The kernel space manager is invoked to overwrite and initialize the hard storage address corresponding to the physical root node memory block address pointer to an unallocated, available, empty address, thus completing the physical destruction of the memory root node.

Images