Cooperative digital monitoring backup method and system

Abstract

The invention relates to the technical field of data visualization and integration, in particular to a collaborative digital monitoring backup method and system, which comprises the following steps: counting channel faults to generate a task number set, matching healthy nodes to generate migration contrast, extracting number difference and injecting into a scheduling queue, comparing task variables to complete processing synchronization, and generating a collaborative backup record. According to the method, continuous fault statistics is carried out on monitoring node channel response log data, invalid channels are marked, abnormal tasks are recognized and positioned in time, and in combination with mapping matching between fault tasks and healthy nodes, task migration has the dynamic sensing and optimal distribution capacity; compared with the prior art, the task response sequence number difference value is compared before migration, the number segment is extracted, data gap supplementation and continuous maintenance are effectively achieved, the data integrity and availability during fault recovery are improved through scheduling queue filling and task table labeling, and the timeliness and consistency of data synchronous updating are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data visualization and integration, in particular to a collaborative digital monitoring backup method and system. BACKGROUND

[0002] The technical field of collaborative fault-tolerant backup relates to the technical system for realizing data fault tolerance and redundant backup in a multi-node system or distributed environment, including fault detection mechanism, fault location mechanism, fault handling strategy, and coordinated execution of backup and recovery process. Usually through the way of inter-node collaborative operation, real-time perception and response control of data state are realized, and fault-tolerant scheduling mechanism is adopted to distribute backup tasks according to the strategy in multiple computing nodes, so as to guarantee the continuous availability of data in the case of hardware damage or system abnormality. In the digital management of agriculture and animal husbandry, with the increasing dependence on remote sensing equipment and video monitoring terminals in the aspects of livestock breeding, intelligent temperature control, water irrigation, environmental monitoring, etc., collaborative digital monitoring backup systems are widely deployed in pasture management servers, edge computing gateways, and greenhouse control centers, etc. multiple nodes, for guaranteeing the continuous recording and high availability access of environmental parameters and monitoring image data. In such scenarios, data usually contains high-frequency change of temperature and humidity indicators, video image sequences, and livestock behavior monitoring records, requiring the system to timely handle abnormal situations such as channel failure and node disconnection, and complete automatic transfer and redundant backup of data, to ensure that critical data can still be effectively stored and recovered in the case of network fluctuations or device failures. Among them, the traditional collaborative digital monitoring backup method refers to controlling the generation and storage of backup data through pre-device backup strategy, and monitoring the state of multiple data sources through a timing polling mechanism. For how to realize high-reliability storage and synchronous update of monitoring data in a digital monitoring system, an independent node type backup method is usually adopted to generate data copies in combination with a statically configured storage path, and a one-way check logic is relied on to complete data consistency checking, and a timestamp-based update control method is used to manage data synchronization process.

[0003] In the existing digital monitoring backup process, a preset static backup strategy is mainly adopted and a timing polling mechanism is used to monitor the state of each data source, resulting in a lag in response when facing sudden channel failures or task interruptions, and the optimal correspondence between abnormal tasks and healthy nodes cannot be dynamically identified. Especially in agricultural and animal husbandry production, if livestock positioning tag transmission is interrupted, greenhouse environmental sensing node fails, remote pasture camera channel is interrupted for a long time, etc., if the system only relies on one-way check logic for consistency detection, some faulty tasks may not be repaired in time after data transmission interruption. For example, after a video monitoring channel is invalid for a long time, the related data copies cannot be transferred to other nodes, which may eventually lead to a gap in monitoring data synchronization, missing of key period information in livestock behavior analysis, and execution error of water and fertilizer irrigation control instructions, reducing the continuous operation ability of the system in agricultural and animal husbandry scenarios. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art and propose a cooperative digital monitoring backup method.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: a cooperative digital monitoring backup method, comprising the following steps: S1: acquiring monitoring node channel response log data, counting the number of channel failure events, judging whether it is continuously more than the heartbeat loss threshold and marking the corresponding channel as a failed migration state, and generating a continuous failure task number set; S2: based on the continuous failure task number set, acquiring the task running state data of all monitoring nodes, selecting the node number with task quantity in the load balanced quantile value and channel failure number zero as the target node number, one-to-one mapping and pairing the corresponding node number with each failure task number, and generating a failure task migration node comparison table; S3: according to the failure task migration node comparison table, detecting the latest response sequence number of the same task number under two nodes, calculating the response number sequence difference value, if the difference value exceeds the transaction log sequence number tolerance, extracting the number segment in the response number difference value range, recording and generating a fault task sequence number segment set; S4: based on the fault task sequence number segment set, extracting the corresponding response number segment in the normal node, injecting into the task migration node task scheduling ready queue, and marking the segment in the migration node task table The identification of injection is generated, and the task data injection record is generated.

[0006] As a further scheme of the present application, the continuous failure task number set includes abnormal channel binding relationship, task failure mark state, task number identification information, the failure task migration node comparison table includes target node number, failure task number mapping relationship, node scheduling load index, the fault task sequence number segment set includes fault number interval, task response sequence difference value, response sequence missing identification, and the task data injection record includes response number segment sequence structure, injection queue position identification, and task segment injection state.

[0007] As a further scheme of the present application, the acquisition step of the continuous failure task number set is specifically: S111: acquiring each monitoring node channel response log data, detecting channel number, response event type and time interval, screening the same type of failure event in unit time, and counting the number of occurrences to obtain a failure frequency value set; S112: According to the fault frequency value set, it is judged whether there is a channel number continuously exceeding the heartbeat loss threshold value, if the continuous number is greater than the heartbeat loss threshold value, the corresponding channel number is marked as an abnormal state, and a continuous failure channel number set is generated; S113: Based on the continuous failure channel number set, the task number bound to each number is extracted, and comprehensive judgment is carried out in combination with the number duration, fault type characteristics and event occurrence density information, all task numbers meeting the conditions are screened, and a continuous fault task number set is obtained.

[0008] As a further scheme of the application, the acquisition step of the fault task migration node correspondence table is specifically: S211: Based on the continuous fault task number set, the task running state data of all monitoring nodes is acquired, the number of active tasks corresponding to the node number and the cumulative number of channel faults are detected, and the mapping between the node number and the state data is carried out, to obtain a node running state parameter set; S212: According to the node running state parameter set, the node number whose task number is in the load balancing fraction value and the cumulative number of channel faults is zero is screened, the load balancing fraction value reference is combined, the node scheduling adaptation value set is calculated and obtained, all node numbers are sorted, the smallest node number set is screened, and the target node number set is obtained; S213: Based on the target node number set and the continuous fault task number set, one-to-one mapping pairing is performed in sequence, each fault task number is sequentially mapped to the target node number, the pairing data table structure is established, the corresponding information of the task number and the target node number is recorded, and the fault task migration node correspondence table is obtained.

[0009] As a further scheme of the application, the acquisition step of the fault task migration node correspondence table is specifically: S311: Based on the fault task migration node correspondence table, the response sequence number of the same task number in the source node and the target node is detected, the maximum response number corresponding to the node and the maximum response number of the target node are obtained, the difference value of the response sequence numbers of the two nodes is calculated, and a response sequence number difference value set is obtained; S312: According to the response sequence number difference value set, threshold value judgment operation is carried out on each item number difference value in the set, all task numbers with a difference value greater than the threshold value are screened and processed in combination with the transaction log sequence number tolerance threshold value, the task number with obvious response difference is extracted, and a fault response task number set is generated; S313: Based on the fault response task number set, difference range segment extraction operation is carried out on the source node and the target node response sequence number corresponding to each task number in sequence, the task number and its number segment structure are archived according to the task number dimension, and a fault task sequence number segment set is obtained.

[0010] As a further scheme of the present application, the task data injection record obtaining step is specifically: S411: Based on the fault task sequence number segment set, obtain the task number corresponding to each number segment, the start value and the end value of the number range, extract the response log record content of each task number in the normal node, perform a section matching operation within the range of each number segment, locate the response sequence segment in the normal node that completely matches the task number and the number interval, and obtain a normal node response number segment segment set; S412: Based on the normal node response number segment segment set, read the response record entries corresponding to each task number in turn, arrange them in the order of response number, perform an injection operation on the segment, insert the corresponding number segment segment in the node scheduling ready queue according to the migration node address corresponding to the task number, and obtain a task queue injection state information set; S413: According to the task queue injection state information set, write into the migration node task table piece by piece, perform a marking operation on the record item corresponding to the task number, sort it into a tracking record detail table according to the task number, and obtain a task data injection record.

[0011] As a further scheme of the present application, the method further comprises: S5: According to the task data injection record, perform integrity comparison on the task variable and the processing phase identifier in the migration node, if there is missing or inconsistent content, perform corresponding variable item covering replacement operation, and register the task number set and the processing state that complete processing synchronization, and generate a digital monitoring cooperative backup record; The digital monitoring cooperative backup record comprises a variable item covering identifier, a task processing consistency state, and a task synchronization completion number set.

[0012] As a further scheme of the present application, the digital monitoring cooperative backup record obtaining step is specifically: S511: According to the task data injection record, read the structure record of the task migration node task table, for each task number, establish a comparison relationship between the number segment content in the injection record and the variable record field of the task table, identify whether there is an abnormal situation such as content missing or number segment mismatch in the task variable field, if there is a field not covered or pointing to an error in the comparison result, mark the task number corresponding variable item, and integrate all abnormal task numbers and corresponding processing states, obtain a variable consistency covering identifier set; S512: Based on the variable consistency coverage identification set, the replacement operation of the corresponding variable item is performed according to the task number list, the number segment content is extracted from the task injection record to perform variable value coverage, the processing stage identification record associated with the variable field is updated, whether the field has completed the synchronization processing is marked, the state information of all field coverage operations is summarized, and the variable processing completion mark set is obtained; S513: According to the variable processing completion mark set, all variable fields are replaced, the task number of the updated completed stage identification is written into the migration node task operation registration table, all field registration records are integrated, and are combined into a parseable data structure segment according to a standard format, and a digital monitoring cooperative backup record is established.

[0013] A cooperative digital monitoring backup system comprises: The fault detection module is used for performing S1: obtaining monitoring node channel response log data, counting the number of channel fault events, judging whether the number of consecutive faults exceeds the heartbeat loss threshold, marking the corresponding channel as a failed migration state, and generating a continuous fault task number set; The migration matching module is used for performing S2: based on the continuous fault task number set, obtaining the task running state data of all monitoring nodes, selecting the node number with the task number in the load balanced quantile value and the channel fault number being zero as the target node number, one-to-one mapping and pairing the corresponding node number with each fault task number, and generating a fault task migration node correspondence table; The sequence comparison module is used for performing S3: according to the fault task migration node correspondence table, detecting the latest response sequence number of the same task number in two nodes, calculating the response number sequence difference value, and if the difference value exceeds the transaction log sequence number tolerance, extracting the number segment in the response number difference value range, and recording to generate a fault task sequence number segment set; The fragment injection module is used for performing S4: based on the fault task sequence number segment set, extracting the corresponding response number segment in the normal node, injecting into the task migration node task scheduling ready queue, and marking the fragment injection identification in the migration node task table, and generating a task data injection record; The synchronization verification module is used for performing S5: according to the task data injection record, performing integrity comparison on the task variables and processing stage identification in the migration node, if there is missing or inconsistent content, performing corresponding variable item coverage replacement operation, and registering the task number set and processing state of completed processing synchronization, and generating a digital monitoring cooperative backup record.

[0014] Compared with the prior art, the advantages and positive effects of the present application are that: In the application, the abnormal task is identified and located in time by continuously counting the failure of the channel response log data of the monitoring node and marking the failed channel, the mapping and matching between the failure task and the healthy node are completed by combining the node load balancing state and the channel health condition, the task migration has dynamic perception and optimal allocation ability, the data gap is effectively filled and the continuity is maintained by comparing the response sequence number difference of the task before migration and extracting the number segment, the scheduling queue filling and task table labeling are completed through the injection operation of the response number segment, the data integrity and availability during failure recovery are improved, the monitoring task has stable operation ability in abnormal environment, and the timeliness and consistency of data synchronization update are enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 is the main step flowchart of the application; Figure 2 is the continuous failure task number set acquisition flowchart of the application; Figure 3 is the failure task migration node comparison table acquisition flowchart of the application; Figure 4 is the fault task sequence number segment set acquisition flowchart of the application; Figure 5 is the task data injection record acquisition flowchart of the application; Figure 6 is the digital monitoring cooperative backup record acquisition flowchart of the application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical scheme and advantages of the application clearer and more apparent, the application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the application and do not limit the application.

[0017] In the description of the application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the application and simplify the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the application. In addition, in the description of the application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0018] Please refer to Figure 1 A cooperative digital monitoring backup method, comprising the following steps: S1: Obtain each monitoring node channel response log data, detect channel number, fault response event type and time interval, count the number of same type fault events in unit time for the same channel number, judge whether it is continuously more than the heartbeat loss threshold (3 times of continuous heartbeat loss according to IEC61508 standard), if the condition is met, extract the task number bound to the corresponding channel and mark it as invalid to be migrated state, generate a continuous fault task number set; S2: Based on the continuous fault task number set, obtain the task running state data of all monitoring nodes, detect node number, active task number and channel fault cumulative number, select the node number with task number at the load balancing quantile value (25% quantile according to ISO / IEC30134-2) and zero channel fault number as the target node number, and one-to-one map each fault task number to the corresponding node number, generate a fault task migration node correspondence table; S3: According to the fault task migration node correspondence table, detect the latest response sequence number of the same task number in two nodes, calculate the sequence difference value of the response number of the two nodes, if the difference value exceeds the transaction log sequence number tolerance (≤10 LSN according to ACID characteristics), extract the number segment in the response number difference value range, and record it as a fault task state section, generate a fault task sequence number segment set; S4: Based on the fault task sequence number segment set, extract the corresponding response number segment in the normal node, detect the task number, arrange it in the task scheduling ready queue of the task migration node according to the response number segment sequence (in line with the embedded real-time system standard), and mark the segment in the migration node task table. The injection record is generated; S5: According to the task data injection record, compare the task variables and processing stage identifiers in the migration node for integrity, if there is missing or inconsistent content, perform the corresponding variable item replacement operation, and register the task number set and processing state that complete processing synchronization, generate digital monitoring cooperative backup record.

[0019] The continuous fault task number set includes abnormal channel binding relationship, task invalidation mark state, task number identification information, the fault task migration node correspondence table includes target node number, fault task number mapping relationship, node scheduling load index, the fault task sequence number segment set includes fault number interval, task response sequence difference value, response sequence missing identifier, the task data injection record includes response number segment sequence structure, injection queue position identifier, task segment injection state, and the digital monitoring cooperative backup record includes variable item coverage identifier, task processing consistency state, and task synchronization completion number set.

[0020] Please refer to Figure 2 , S1 step is: S111: Obtain each monitoring node channel response log data, detect channel number, response event type and time interval, filter the same type of fault event in unit time, and count the number of occurrences to obtain the fault frequency value set; To obtain the channel response log data of each monitoring node, each node deployed by the monitoring system needs to be accessed one by one, and the original channel response log recorded in the set period is extracted. By setting the channel number rule, the number item in the log is extracted, such as channel numbers A1, A2, B1, etc. Further, the event type parameter item contained in the log is identified, such as "fault response", "state loss", "response timeout", etc. And according to the event timestamp, the time interval data in the continuous log record is extracted. For the events occurring in unit time, according to the set time window such as 10 minutes, the time period window is divided, all the same type of event records in the time window are extracted, and grouped according to the channel number. The number of fault response events in each group is counted, and a mapping table between channel number and fault frequency is constructed. For example, channel A1 records 5 "state loss" events in 10 minutes, and channel A2 records 2 "response timeout". Through this process, the initial frequency statistics are formed, as shown in Table 1.

[0021] Table 1 Channel fault frequency statistics table

[0022] As shown in Table 1, the collected data needs to clearly indicate the fault type and number of each channel in the same time window. This statistical step avoids ambiguous processing and ensures uniform samples, forming a preliminary cluster with statistical significance. In practical applications, it can be deployed in an industrial control system, such as a PLC node in a water treatment site. The node ID corresponds to the channel number. The system synchronizes the log data every 10 minutes. The status response records of A1 and A2 channels in a certain period are extracted from the log and classified. For the judgment criteria in the "filtering" process, the time window is divided into 10 minutes to distinguish the event categories under each channel number and limit their attribution. For the data items appearing in the "statistics" operation, condition filtering is performed relying on the type field in the event log. In the summary stage, a one-to-one corresponding data structure table is established as the input data set for subsequent processing. Finally, the fault frequency value set is obtained.

[0023] S112: According to the fault frequency value set, it is judged whether there is a channel number that continuously exceeds the heartbeat loss threshold. If the number of continuous times is greater than the heartbeat loss threshold, the corresponding channel number is marked as an abnormal state, and a continuous failure channel number set is generated; According to the fault frequency value set, it is judged item by item whether the corresponding fault times are continuously more than the set heartbeat loss threshold value indexed by the channel number. The threshold value is set to 3 times of continuous occurrence. According to the IEC61508 standard, when 3 times of continuous heartbeat loss events occur in a channel, it is considered to exceed the allowable range. In the judgment process, the time stamp needs to be sorted and the continuity of the event needs to be checked. For example, if A1 channel occurs heartbeat loss events at t1, t2, t3, and the time interval between the three is not more than the maximum allowed interval Δt_max=120 seconds, it is judged as a continuous event, and the marking operation is performed to form the abnormal channel number record. If a channel appears an interruptive fault record, such as t1, t3, t6, it is not counted in the continuous sequence. The judgment operation needs to set auxiliary state variables to record the continuous number, which is dynamically updated when traversing the log. The example judgment process is as follows: A1 channel, heartbeat loss record occurrence time is [08:01:05, 08:02:30, 08:04:10], and the interval between any two is less than 120 seconds; The cumulative continuous number reaches 3 times, which exceeds the threshold value, and the flag bit is set to 1, and the abnormal marker pool is added; If there is a breakpoint, for example, there is no event between 08:01:05 and 08:04:10, then reset the count.

[0024] The threshold value setting reference is based on the device hardware update frequency and network transmission redundancy design. In industrial applications, it is usually set to 2-5 times as the tolerable threshold value. To improve the fault tolerance capability, the upper limit value is 3 times. The specific value of the threshold value is determined by the system historical operation and maintenance data. The system background can automatically learn the reasonable interval according to the device log, and in the example, 3 times is used as the judgment reference, and whether the continuous event condition is met is judged through samples, and finally the continuous failure channel number set is formed.

[0025] S113: Based on the continuous failure channel number set, extract the task number bound to each number, combine the number duration, fault type characteristics, and event occurrence density information for comprehensive judgment, filter all task numbers that meet the conditions, and obtain the continuous fault task number set; Based on the set of continuous failure channel numbers, the binding task number of each abnormal channel number in the task system is retrieved, the mapping relationship is called from the historical task allocation table, for example, A1 channel binds task T05, B2 channel binds task T08, after extracting the task number, the task event table is further called to record the number duration, fault type characteristic value and event intensity data, the number duration is obtained by calculating the start and end time fields in the event table, for example, task T05 occurs abnormality from 08:00:00 to 08:10:00, the number duration is 10 minutes, the fault type characteristic value is set by the fault type corresponding weight, as follows: heartbeat loss = 3.0, state switching abnormality = 2.5, response delay = 2.0, the event intensity is calculated by the number of occurrences in a unit of time, if T05 task occurs 6 abnormal events in 10 minutes, the intensity is 0.6 times / minute, the actual data is as follows: Table 2 Task Number Abnormal Feature Data Table

[0026] Referring to Table 2, the comprehensive abnormal features of the corresponding task can be obtained according to the data items, which are further used for task abnormal state recognition, combined with factors such as fault duration, fault level and abnormal intensity for manual or rule screening, if the feature combination corresponding to a task number meets the preset comprehensive judgment rule, it is included in the task list to be migrated, the screening rule is obtained by training the system historical task data and equipment working condition evaluation model, or a static threshold combination can be set by an expert, for example, number duration > 8 minutes, fault type characteristic value ≥ 2.5, event intensity > 0.5 times / minute, and 3 items are met, then it is judged as a continuous failure task, in the example, T05 meets the conditions, then it is included in the final result, and the set of continuous failure task numbers is obtained.

[0027] Please refer to Figure 3 , S2 step is: S211: Based on the set of continuous failure task numbers, the task running state data of all monitoring nodes is obtained, the number of active tasks and the cumulative number of channel failures corresponding to the node number are detected, and the mapping between the node number and the state data is performed, to obtain the set of node running state parameters; Based on the continuous fault task number set, the task running state of all monitoring nodes is extracted, and the task allocation table and channel event record table of the current node are read from the system registered monitoring node list one by one. First, the number of each node is detected and its task queue state field is read, the number of tasks in the state of "running" is extracted as the current active task number of the node. For example, the running task entry of node N01 is T01 to T12, a total of 12, and the active task number is 12. Secondly, the records related to "fault response" in the node channel response log are detected, classified by channel number, and the cumulative fault times are counted. In the example, N01 channels A1, A3 and A4 each appear once, a total of 3 times. At the same time, the resource adjustability index is obtained from the system resource monitoring module. The resource adjustability is a weighted and normalized summary value of multiple sub-indices. For example, according to the CPU occupancy rate 0.6, the memory utilization rate 0.7 and the IO load 0.8, the weighted proportion is 0.75. Finally, the node number, task number, fault number and resource adjustability four field set are constructed to form the following structure: , the resource adjustability of the node is 0.75. Table 3 Node running state parameter table

[0028] As shown in Table 3, node N03 has the lowest active task number and 0 times of fault, and the resource adjustability is 0.93. The table is used as the basis for scheduling adaptation analysis to obtain the node running state parameter set.

[0029] S212: According to the node running state parameter set, the node number whose task number is in the load balancing quantile value and the cumulative number of channel faults is zero is screened, combined with the load balancing quantile value reference (set to the 25th percentile task number according to ISO / IEC30134-2 standard), and the formula is used: ; The node scheduling adaptation value set is obtained by operation, the node numbers are sorted, the smallest node number set is screened, and the target node number set is obtained. Wherein, represents the node scheduling adaptation value, represents the active task number of the node , represents the 25th percentile value of the active task number of all nodes, represents the average resource adjustability of the node , is the cumulative number of channel faults of the node , is the actual scheduling response delay of the node , The upper limit of the permissible response of the scheduling set for the system. The two sub-items reflect the matching degree between the load-fault suppression factor and the response capability. According to the node operation status parameter set, the scheduling adaptability analysis of each node operation data is performed. First, the 25% quantile value is calculated based on the task quantity field to measure whether the node is at the quartile level under load. The active task quantity field is arranged in ascending order as [4, 8, 9, 12] to calculate the quantile position. ,thus , that is, the quartile value under load is 5; each node is screened and judged, and the nodes with active tasks less than or equal to 5 and channel failure times of 0 meet the basic conditions. In the example, N03 meets this constraint; then the parameters are further introduced for the nodes that meet the requirements Resource scalability, Scheduling response delay and The system sets the maximum allowable response delay and uses the formula to calculate the node scheduling adaptability value : Taking N03 as an example, the parameters are: 、 、 、 、 、 , substituting into: ; Then calculate node N02, 、 、 、 、 : ; The scheduling scores of all candidate nodes are summarized as follows: Table 4 Node scheduling adaptability calculation table

[0030] As shown in Table 4, the node N03 with the lowest scheduling fitness value is selected into the target set, and the target node number set is obtained.

[0031] The node scheduling adaptation value is used to measure whether each monitoring node is suitable as a migration target of a fault task under the current system running state, and is a comprehensive scoring index, reflecting the adaptation degree of the node in multiple dimensions such as task load level, resource scheduling ability, channel stability and scheduling response time, and the smaller the value is, the lighter the current running load of the node is, the higher the resource redundancy is, the less the channel fault interference is, and the closer the scheduling response delay is to the system tolerance, and the higher the task undertaking ability is, so that the value can be used as a criterion for priority selection and ordering of the node in the task migration scheduling process, and is helpful to realize the reasonable transfer of the fault task in the system and the balanced distribution of the node load.

[0032] The formula is based on multi-dimensional evaluation of node scheduling adaptability, adopts the form of sum of two independent indexes to reflect the comprehensive judgment mechanism of node load state and response ability, the first term In the formula, represents the deviation degree between the number of active tasks of the current node and the load quantile reference, and is used to measure the deviation degree of the node relative to the ideal load state, represents the resource adjustability of the node, and reflects the resource redundancy available for scheduling of the node at present, so that the product of the two is used to measure the suppression strength of the node load on the available resources for scheduling, and the product value is used as the numerator to constitute the scheduling load influence term; the denominator part represents the adjustment factor of the number of channel faults, and the square root operation avoids the sharp fluctuation of the denominator caused by too small number of faults, and prevents the denominator from being zero, and by adding 1, the square root value is always greater than zero, and the absolute value of the whole is taken to prevent the deviation direction from affecting the score trend, and to ensure positive comparability; the second term represents the absolute value of the difference between the actual scheduling response delay of the node and the maximum allowed response time set by the system, and is used to measure the deviation of the node in response timeliness, and reflects whether the node can meet the current scheduling timeliness requirement, and the two indexes do not cover each other, so that the two parts are combined by addition, and the finally obtained represents the comprehensive scheduling adaptation degree, and the smaller the value is, the more suitable the node is to undertake task migration.

[0033] S213: Based on the target node number set and the continuous fault task number set, a one-to-one mapping pairing is performed in sequence, each fault task number is sequentially mapped with a target node number, a pairing data table structure is established, the corresponding information of the task number and the target node number is recorded, and a fault task migration node comparison table is obtained; Based on the target node number set and the continuous failure task number set, perform node task mapping operation, set the continuous failure task number set as [T05, T08], the target node number set as [N03], establish the pairing structure of task number and node number, if the number of tasks exceeds the number of nodes, use the rotation allocation strategy, each task number is sequentially assigned to the current task load lowest node, the node load is dynamically updated according to the number of active tasks, in this example, two tasks are allocated to node N03; When generating the pairing relationship table, record the task number, corresponding node number and mapping timestamp, for example: Table 5 Fault task migration node comparison table

[0034] By establishing the task node corresponding table as shown in table 5, the execution path of the migrated task can be tracked, and the fault task migration node comparison table can be obtained.

[0035] Please refer to Figure 4 , S3 step is: S311: Based on the fault task migration node comparison table, detect the response sequence number of the same task number in the source node and the target node, obtain the maximum response number corresponding to the node and the maximum response number of the target node, and perform the calculation operation of the difference value of the response number of the two nodes, to obtain the response sequence number difference set; Based on the fault task migration node comparison table, extract each task number and its corresponding source node number and target node number one by one, read the task running log in the source node and the target node respectively, extract the response sequence number content corresponding to the task number field in the log record, and perform numerical conversion processing on the response sequence number field of the same task number in the two nodes, convert the field value with LSN prefix to integer number form uniformly, after executing the sequence number value analysis, extract the maximum response number as the latest response number of the task in the node, then calculate the difference value of the maximum response number of the same task number in the source node and the target node, the result is recorded in integer form, in the example, the maximum response number of task T105 in node N01 is LSN_1087, converted to integer is 1087, the maximum response number in node N03 is LSN_1094, converted to integer is 1094, the difference value is 1094 minus 1087 equal to 7, form mapping record with task number and its response number difference value, after collecting the maximum response number of each task number and its number difference value, generate the following data structure: Table 6 Node task maximum response number and difference table

[0036] As shown in table 6, four groups of task numbers are listed in the maximum response number and difference value of the two nodes, and the response sequence number difference set is obtained.

[0037] S312: According to the response sequence number difference set, the threshold judgment operation is performed on each item number difference in the set, the task number with a difference value greater than the threshold value is screened and processed in combination with the transaction log sequence number tolerance threshold, the task number with a prominent response difference is extracted, and a fault response task number set is generated; According to the response sequence number difference set, the number difference of each task number record is compared with the transaction log tolerance threshold, the tolerance threshold is set to 10, which is the upper limit value of log consistency check, the difference value is read and compared with the threshold value, if the number difference value is greater than 10, the task number is marked as abnormal response, all task numbers exceeding the threshold value are screened and assembled into an abnormal task set, in the example, the task T202 number difference is 15, which is greater than the threshold value 10, and is determined as an abnormal task, the task T218 difference is 10, which is equal to the threshold value and is not included in the abnormal set, T105 and T301 difference is 7, which is lower than the threshold value and is not included in the abnormal set, after screening, only T202 meets the abnormal condition, then the response log records of the task number under the source node and the target node are extracted, the sequence number content in each response record is arranged in time stamp order, the number values in the same time stamp or adjacent time period are compared, the difference value increment rate is calculated, if the rate reaches 5 number units between two consecutive records, it is confirmed as a response interruption behavior, the task number corresponding to the abnormal response number is recorded in the abnormal task set, after executing the above process, the following results are obtained: Table 7 Response number difference judgment and abnormal marking table

[0038] As shown in table 7, only task T202 meets the difference limit and the response number exists interruption, and the fault response task number set is obtained.

[0039] S313: Based on the fault response task number set, the source node and the target node response sequence number corresponding to each task number is extracted in difference range segment, the task number and its number segment structure are archived according to the task number dimension, and the fault task sequence number segment set is obtained; Based on the fault response task number set, all task numbers are subjected to number segment extraction operation, the maximum response number of task number T202 in source node N02 is LSN_3050, and the maximum response number in target node N05 is LSN_3065, the number value is converted into integer 3050 and 3065, the minimum value plus 1 is the starting value of the number segment, and the maximum value minus 1 is the end value of the number segment, that is, the number segment range is LSN_3051 to LSN_3064, all numbers between them are listed in sequence to form the interrupt number segment sequence, and the number segment is bound with the task number T202, and the corresponding node number and number range are recorded, and the number segment detail table structure is generated: Table 8 Fault task number segment extraction table

[0040] As shown in Table 8, the number segment length is 14 sequence number units, the number segment is continuously formed into the task response interrupt range, and the fault task sequence number segment set is obtained.

[0041] Please refer to Figure 5 , S4 step is: S411: Based on the fault task sequence number segment set, the task number, the starting value and the end value of the number range corresponding to each number segment are obtained, the response log record content of each task number in the normal node is extracted, the segment matching operation is performed within the range of each number segment, the response sequence segment in the normal node which is completely matched with the task number and the number interval is located, and the normal node response number segment piece set is obtained; Based on the fault task sequence number segment set, the start and end response number range corresponding to each group of task numbers is extracted, first, the number segment structure is analyzed, the starting number and the ending number value of each segment are determined, then the task number is used as the index to retrieve the normal node response log data, the response record within the number segment corresponding interval is extracted, all records are judged according to the number field interval, only the response record entries falling into the number range are retained, and the out-of-segment numbers are eliminated, to form the in-segment effective record set, then the response records within the number segment range are classified uniformly with the task number as the primary key, in the example, if the task T512 corresponds to the number segment LSN_6200 to LSN_6207, 6 entries with record numbers LSN_6201 to LSN_6206 are extracted after querying the normal node log, it is confirmed that they are in-segment effective data, and they are associated with T512 and archived, the mapping structure of the task number and the response number segment matching data is obtained, after all task number corresponding number segment response records are summarized, the following structure table is formed: Table 9 Normal node response number segment extraction table

[0042] As shown in Table 9, task numbers T512, T208 and T319 extract valid response records in the corresponding number segments in the normal node respectively, and obtain the normal node response number segment fragment set.

[0043] S412: Based on the normal node response number segment fragment set, read the response record entries corresponding to each task number in turn, arrange them in order of response number, perform an injection operation on the fragment, insert the corresponding number segment fragment into the node scheduling ready queue according to the migration node address corresponding to the task number, and obtain the task queue injection state information set; Based on the normal node response number segment fragment set, read the response records in the corresponding number segment of each task number, perform ascending processing on the number field, and ensure the time sequence consistency of the response fragment before entering the migration node scheduling ready queue. Then, write the ready queue structure of the task number in the migration node in turn according to the sorting result of the number segment. During the injection process, perform capacity state checking on the migration node task queue, read the free capacity information of the current queue, judge whether there is an overflow risk in task injection, if there is available capacity in the current ready queue, perform sequential writing operation, and embed the response records into the end of the queue or the specified insertion point one by one. If task number T208 injects records LSN_4306 to LSN_4310 for a total of 5 records, and the remaining space of the queue is 10, then fill them in order directly without compression adjustment. If there are other concurrent injection of task numbers, perform injection processing according to the priority of task number. All successfully injected records are attached with record injection position index and occupied space segment identifier. Complete the queue structure update, and list the injection state information corresponding to each task number after the update in the following table: Table 10 Task Response Record Injection Status Table

[0044] Referring to Table 10, task numbers T512, T208 and T319 complete response number segment injection in the task migration node ready queue, and obtain the task queue injection state information set.

[0045] S413: According to the task queue injection state information set, write the migration node task table one by one, perform a marking operation on the record items corresponding to the task number, sort them into a tracking record detail table according to the task number, and obtain the task data injection record. According to the task queue injection state information set, the task number information of each successfully injected task is obtained, the task table structure of the task migration node is called, the update operation is performed on the record item corresponding to the task number, the injection flag bit in the task state field is set to 1, indicating that the number segment fragment of the task number is injected into the ready queue, then the injection record log line is constructed for each task number, the log structure includes the task number, the task migration node number, the start number and the end number of the number segment, the injection position index interval and the system timestamp of the injection completion, and the structured log set is classified and archived according to the task number, and after the structure is unified, it is written into the task injection operation record table to complete the entire injection information recording process, and the following structure table is generated: Table 11 Task Number Injection Record Table

[0046] As shown in Table 11, the injection information record of each task number has been completed into the warehouse, and the task data injection record is obtained.

[0047] Please refer to Figure 6 , S5 step is: S511: According to the task data injection record, read the structure record of the task table of the task migration node, for each task number, establish a comparison relationship according to the content of the number segment in the injection record and the variable record field of the task table, identify whether there is an abnormal situation such as content missing or number segment mismatch in the task variable field, if there is a field that is not covered or points to an error in the comparison result, mark the corresponding variable item of the task number, and integrate all abnormal task numbers and corresponding processing states to obtain the variable consistency coverage identification set; According to the task data injection record, the task number, number segment range, injection node number and other fields in the record are extracted, and then the task table structure content of each task number in the corresponding migration node is called in turn, the task variable item value and processing stage identifier field are searched, and item-by-item analysis is performed. First, the start number and end number of the number segment are taken as the interval limits, the number segment parameter value marked by the field identifier bit in the task table structure is read and compared, the mapping result is matched at the field level, and if any boundary number of the number segment is missing in the task variable field or there is a cross-segment index value, it is determined that the content is inconsistent and is recorded. At the same time, the structure consistency check of the auxiliary fields such as the number of occupied bytes, offset start value and field check bit of the task table variable item is performed, if the injection record of task number T216 is number segment LSN_5201 to LSN_5207, and the start number of the variable field in the task table is LSN_5203 and the end number is LSN_5205, the offset is marked as 32 bytes, the task variable item is considered as a number segment missing record, and is added to the abnormal comparison list, and its processing state is marked as 0. In this process, it is further judged whether there is a stage identifier inconsistent with the injection task stage according to the processing identifier field, if the task stage in the record is "3" and the task table field identifier shows "2", it is considered as a stage identifier inconsistency item, which also needs to be added to the marker list. Finally, all abnormal variable items and their task numbers are sorted by number and combined into a structure marker set, and the variable consistency coverage identifier set can be obtained.

[0048] S512: Based on the variable consistency coverage identifier set, the replacement operation of the corresponding variable item is performed according to the task number list, the number segment content is extracted from the task injection record to cover the variable value, the processing stage identifier record associated with the variable field is updated, whether the field has completed the synchronization processing is marked, the state information of all field coverage operations is summarized, and the variable processing completion marker set is obtained. Based on the variable consistency coverage identification set, the task number of each record item is subjected to field replacement operation in the task table. The variable value block is read according to the number segment content in the task data injection record, and is written as the source value of the coverage in the corresponding field position in the task table. At the same time, the corresponding processing stage identification bit field is called to replace the processing stage value marked in the injection record. In the operation, the original field occupation length, replacement field length, and updated offset position information need to be extracted and archived. For example, if the task number T216 has a variable field length of 64 bytes before replacement and 72 bytes after replacement, the offset index needs to be reset and the structure index table needs to be updated. After the field coverage is completed, the integrity of the replaced variable structure needs to be judged. The field check bit value is read and compared with the structure hash digest. If the consistency check passes, the variable item replacement is successful, and the updated field offset start value and occupation length are recorded. The processing status is marked as completed, forming a variable coverage completion item. The status information of all completed fields is finally combined into a variable processing record item summary table. The field content is shown in the following table: Table 12 Task variable replacement completion field table

[0049] As shown in Table 12, the variable fields of task numbers T216, T189, and T307 have been replaced and the stage status is synchronized. The variable processing completion marker set is obtained.

[0050] S513: According to the variable processing completion marker set, write the task numbers whose variable field replacement and stage identification update have been completed into the migration node task operation registration table. Integrate all field registration records, combine them into a parseable data structure segment according to the standard format, and establish a digital monitoring cooperative backup record. According to the variable processing completion marker set, the task numbers marked as completed are registered in the migration node task operation registration table one by one. Extract the field processing completion timestamp, field replacement number, stage synchronization result, processing node identification, and other information corresponding to each task number to construct a structured registration record item. For example, task number T216 completes field update in node N07 and records the timestamp 2024-07-03 10:28:44. The field synchronization state is 1, and the field replacement number is 3 items. The registration line content is T216-N07-1-3-2024-07-03 10:28:44. After the registration record is completed, all task number processing information is summarized and integrated into a data structure block, which is transferred to the backup data area of the digital monitoring system. This structure record contains key meta-information fields such as field name, content description, update state, and processing node, which meet the structure field standard for parsing and reading. Finally, the write operation of the processing result block is completed, and the digital monitoring cooperative backup record is established.

[0051] A cooperative digital monitoring backup system comprises: The fault detection module is configured to perform S1: obtaining monitoring node channel response log data, counting the number of channel fault events, determining whether the number of consecutive faults exceeds the heartbeat loss threshold, marking the corresponding channel as a failed migration state, and generating a continuous fault task number set. The migration matching module is configured to perform S2: based on the continuous fault task number set, obtaining task running state data of all monitoring nodes, selecting a node number with a task quantity in a load balanced quantile value and a channel fault number of zero as a target node number, one-to-one mapping and pairing the corresponding node number with each fault task number, and generating a fault task migration node reference table. The sequence comparison module is configured to perform S3: according to the fault task migration node reference table, detecting the latest response sequence number of the same task number in two nodes, calculating the response number sequence difference value, if the difference value exceeds the transaction log sequence number tolerance, extracting the number segment in the response number difference value range, recording and generating a fault task sequence number segment set. The fragment injection module is configured to perform S4: based on the fault task sequence number segment set, extracting the corresponding response number segment in the normal node, injecting it into the task migration node task scheduling ready queue, and marking the fragment injection identification in the migration node task table, and generating a task data injection record. The synchronization verification module is configured to perform S5: according to the task data injection record, the task variables and processing phase identification in the migration node are compared for integrity, if there is missing or inconsistent content, the corresponding variable item is replaced, and the task number set and processing state of the completed processing synchronization are registered, and a digital monitoring cooperative backup record is generated.

[0052] The above is only the preferred embodiment of the present application, not other forms of the present application, any skilled in the art may use the above disclosed technical content to change or modify as equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification of the above embodiments without departing from the technical solution content of the present application, according to the technical essence of the present application, still belongs to the protection scope of the present application technical solution.

Claims

1. A method of cooperative digital monitoring backup, characterized by, The method comprises the following steps: S1: acquiring monitoring node channel response log data, counting the number of channel failure events, judging whether the number of consecutive times exceeds the heartbeat loss threshold, marking the corresponding channel as a failed state to be migrated, and generating a continuous failure task number set; S2: based on the continuous failure task number set, acquiring task running state data of all monitoring nodes, selecting node numbers with task numbers in load balanced quantile values and zero channel failure times as target node numbers, one-to-one mapping and pairing the corresponding node numbers with each failure task number, and generating a failure task migration node comparison table; S3: according to the failure task migration node comparison table, detecting the latest response sequence number of the same task number under two nodes, calculating the response number sequence difference value, if the difference value exceeds the transaction log sequence number tolerance, extracting the number segment in the response number difference value range, recording and generating a fault task sequence number segment set; S4: based on the fault task sequence number segment set, extracting the corresponding response number segment in the normal node, injecting it into the task migration node task scheduling ready queue, and marking the segment in the migration node task table as injected, and generating a task data injection record.

2. The method of claim 1, wherein, The continuous failure task number set includes abnormal channel binding relationship, task failure marker state, and task number identification information, the failure task migration node comparison table includes target node number, failure task number mapping relationship, and node scheduling load index, the fault task sequence number segment set includes fault number interval, task response sequence difference value, and response sequence missing identifier, and the task data injection record includes response number segment sequence structure, injection queue position identifier, and task segment injection state.

3. The method of claim 1, wherein, The acquisition step of the continuous failure task number set is specifically: S111: acquiring channel response log data of each monitoring node, detecting channel number, response event type and time interval, screening the same type of failure events in a unit of time, and counting the number of occurrences to obtain a failure frequency value set; S112: according to the failure frequency value set, judging whether there is a channel number that continuously exceeds the heartbeat loss threshold, if the number of consecutive times is greater than the heartbeat loss threshold, marking the corresponding channel number as an abnormal state, and generating a continuous failure channel number set; S113: based on the continuous failure channel number set, extracting the task number bound to each number, and comprehensively judging the number duration, failure type characteristics and event occurrence density information to screen all task numbers meeting the conditions and obtain the continuous failure task number set.

4. The method of claim 1, wherein, The acquisition step of the failure task migration node comparison table is specifically: S211: based on the continuous failure task number set, acquiring task running state data of all monitoring nodes, detecting the number of active tasks corresponding to the node number and the cumulative number of channel failures, and mapping between the node number and the state data to obtain a node running state parameter set; S212: According to the node running state parameter set, screening the node number in which the task quantity is in the load balancing quantile value and the channel fault cumulative number is zero, combining the load balancing quantile value reference, calculating to obtain the node scheduling adaptation degree value set, sorting all node numbers, screening the smallest node number set, and obtaining the target node number set; S213: Based on the target node number set and the continuous fault task number set, performing one-to-one mapping pairing in order, establishing mapping between each fault task number and the target node number in turn, establishing a pairing data table structure, recording the corresponding information of the task number and the target node number, and obtaining a fault task migration node correspondence table.

5. The method of claim 1, wherein, The acquisition step of the fault task sequence number segment set is specifically: S311: Based on the fault task migration node correspondence table, detecting the response sequence number of the same task number in the source node and the target node, obtaining the maximum response number corresponding to the node and the maximum response number of the target node, performing calculation operation of the difference value of the two node response numbers, and obtaining a response sequence number difference value set; S312: According to the response sequence number difference value set, performing threshold value judgment operation on each number difference value in the set, combining the transaction log sequence number tolerance threshold, screening all task numbers with a difference value greater than the threshold, extracting task numbers with obvious response difference, and generating a fault response task number set; S313: Based on the fault response task number set, performing difference range segment extraction operation on the source node and the target node response sequence number corresponding to each task number in turn, archiving the task number and its number segment structure according to the task number dimension, and obtaining a fault task sequence number segment set.

6. The method of claim 1, wherein, The acquisition step of the task data injection record is specifically: S411: Based on the fault task sequence number segment set, obtaining the task number, the start value and the end value of the number range corresponding to each number segment, extracting the response log record content of each task number in the normal node, performing section matching operation within each number segment range, positioning the response sequence segment in the normal node which completely matches the task number and the number interval, and obtaining a normal node response number segment segment set; S412: Based on the normal node response number segment segment set, reading the response record entries corresponding to each task number in turn, arranging them in order according to the response number, performing injection operation on the segment, inserting the corresponding number segment segment into the node scheduling ready queue according to the migration node address corresponding to the task number, and obtaining a task queue injection state information set; S413: According to the task queue injection state information set, writing into the migration node task table piece by piece, performing marking operation on the record items corresponding to the task number, classifying and arranging into a tracking record detail table according to the task number, and obtaining a task data injection record.

7. The method of claim 1, wherein, The method further comprises: S5: According to the task data injection record, the task variable and the processing stage identifier in the migration node are compared for integrity, if there is missing or inconsistent content, the corresponding variable item is replaced, and the task number set and the processing state of the completed processing synchronization are registered, and the digital monitoring cooperative backup record is generated; The digital monitoring cooperative backup record includes variable item coverage identifier, task processing consistency state, and task synchronization completion number set.

8. The method of claim 7, wherein, The acquisition step of the digital monitoring cooperative backup record is specifically: S511: According to the task data injection record, read the structure record of the task migration node task table, for each task number, compare the content of the injection record number segment with the task table variable record field, identify whether there is content missing or number segment mismatch abnormality, if the comparison result exists field not covered or pointing to error, mark the corresponding variable item of the task number, and integrate all abnormal task numbers and corresponding processing states to obtain the variable consistency coverage identifier set; S512: Based on the variable consistency coverage identifier set, according to the task number list, the corresponding variable item is replaced, the number segment content is extracted from the task injection record to cover the variable value, the processing stage identifier record associated with the variable field is updated, whether the field has completed the synchronization processing is marked, the state information of all field coverage operations is summarized, and the variable processing completion mark set is obtained; S513: According to the variable processing completion mark set, write all variable field replacement and stage identifier update completed task numbers into the migration node task operation registration table, integrate all field registration records, combine them into a parseable data structure segment according to the standard format, and establish the digital monitoring cooperative backup record.

9. A cooperative digital monitoring backup system, characterized by The system is used to realize the cooperative digital monitoring backup method of any one of claims 1-8, comprising: The fault detection module is used to execute S1: obtaining the monitoring node channel response log data, counting the number of channel fault events, judging whether it is continuously more than the heartbeat loss threshold and marking the corresponding channel as invalid to be migrated state, and generating a continuous fault task number set; The migration matching module is used to execute S2: based on the continuous fault task number set, obtaining the task running state data of all monitoring nodes, selecting the node number with the task number in the load balanced quantile value and the channel fault number as zero as the target node number, one by one mapping and pairing the corresponding node number and each fault task number, and generating a fault task migration node correspondence table; The sequence comparison module is used to execute S3: according to the fault task migration node correspondence table, detecting the latest response sequence number of the same task number in two nodes, calculating the response number sequence difference, if the difference exceeds the transaction log sequence number tolerance, extracting the number segment in the response number difference range, and recording the generated fault task sequence number segment set; The fragment injection module is configured to perform S4: based on the fault task sequence number segment set, extract corresponding response number segments in the normal node, inject the number segments into a task migration node task scheduling ready queue, mark a fragment injection identification in a migration node task table, and generate a task data injection record; The synchronization check module is configured to perform S5: according to the task data injection record, perform integrity comparison on a task variable and a processing stage identification in the migration node, perform a corresponding variable item covering replacement operation if there is missing or inconsistent content, register a task number set and a processing state of completed processing synchronization, and generate a digital monitoring cooperative backup record.

Images