Computer data storage method and system based on Internet of Things

Abstract

The invention relates to the technical field of data storage, in particular to a computer data storage method and system based on the Internet of Things, and the method comprises the following steps: obtaining a page block write-in operation stream set, extracting repeated numbers and operation sequences, adjusting a buffer sequence structure, aggregating attribution page block paths, and constructing a write-in distribution and node association relationship. And obtaining a data node write-in organization structure. According to the method, a priority write-in sequence is extracted through a page block number and an operation time sequence, a page block arrangement structure in a buffer area is reconstructed, page blocks with associated characteristics are aggregated into a continuous path in combination with a logic attribution relation, a centralized write-in process in a logic block is established, and a write-in distribution state is extracted through a corresponding relation between a node identifier and the page block number. A data organization structure with attribution boundaries, path connectivity and scheduling order features is generated, the local intensive degree, the structure connection capability and cooperation and coordination among cross-node paths are enhanced, and the organization of write-in paths and the consistency of data distribution structures are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data storage, in particular to a computer data storage method and system based on Internet of Things. BACKGROUND

[0002] The technical field of data storage involves technical solutions for saving, organizing and calling electronic information, and its core matters include data writing mechanism, information retrieval structure, storage path management, redundancy check control and storage medium structure design. This technical field usually realizes effective persistence and rapid access of data by constructing database table structure, file directory system and cache strategy, and mainly relies on information classification index rules, storage path mapping methods and storage capacity allocation strategies to complete the collection and distribution of data between different levels and different storage media. The traditional computer data storage method refers to a method of using a single central server to centrally write and read data uploaded by a computing terminal through an SQL relational database or a file system. This method is aimed at archiving and storing data collected remotely in a multi-terminal environment. It usually sends device status information, running logs or user input data generated by the terminal to the central node by setting database field structure and using data transmission protocols such as TCPIP. The central node completes data insertion according to the data table field, and performs subsequent data extraction and update operations according to the set query statement.

[0003] The prior art relies on a central server to uniformly receive data uploaded by terminals when performing storage operations, and uses field mapping rules to perform insertion and query processing. In dense access scenarios, it cannot identify and classify write operations, resulting in mixed operation states in data streams, scattered page block scheduling order, and inability to guarantee the continuity of cache allocation. In terms of structure ownership and path distribution, no numbering level mapping relationship is established, resulting in ambiguous page block ownership within logical blocks, random scheduling behavior, and difficulty in realizing data aggregation management and structured write control, which leads to problems such as scattered data distribution, reduced processing efficiency and chaotic node path scheduling. SUMMARY

[0004] To solve the technical problems existing in the prior art, the embodiments of the present application provide a computer data storage method based on Internet of Things. To achieve the above-mentioned purpose, the present application adopts the following technical scheme, a computer data storage method based on Internet of Things, comprising the following steps: S1: Obtain the write request sequence in the data cache of the Internet of Things edge node, filter the same instructions according to the page block number, extract the operation content according to the time sequence, and obtain the page block write operation stream set by the page block write process; S2: based on the page block write operation flow set, extract the repeatedly occurring page block number, compare the operation number and the adjacent time interval change order, concentrate and write the repeated number into the output sequence, and obtain the target page block operation identification list; S3: based on the target page block operation identification list, extract the corresponding order index in the write scheduling buffer pool, change the number position using the buffer mapping table, and adjust the page block into the current write execution area specified paragraph to obtain the reconstructed page block buffer order table; S4: based on the reconstructed page block buffer order table, extract the number corresponding logical block attribution, compare the attribution number to filter the page block number consistent data path, and continuously write the flow to obtain the logically concentrated page block list; S5: based on the logically concentrated page block list, extract the corresponding relationship between the page block number and the Internet of Things node identification, extract the write path and order information, associate the logical block attribution range and the actual scheduling behavior, and obtain the data node write organization structure.

[0005] As a further scheme of the application, the page block write operation flow set includes a page block number sequence, a time interval sequence, and a write number change feature, the target page block operation identification list includes a repeated page block number, a time interval comparison value, and a behavior state type, the page block buffer order table includes a page block scheduling index, a buffer mapping relationship, and an execution queue position, the logically concentrated page block list includes a page block attribution logical block number, a page block grouping structure, and a concentrated write sequence, and the data node write organization structure includes a node identity mapping relationship, an actual write path, and a logical block association state.

[0006] As a further scheme of the application, the write output sequence refers to a page block write instruction sequence formed by concentrating and outputting the page block operations with repeated numbers in time sequence after identifying the page block operation behavior. The scheduling buffer pool refers to a buffer structure for temporarily storing and sorting page block write tasks, and the scheduling order of the page blocks in the execution queue is controlled through an index and a mapping relationship.

[0007] As a further scheme of the application, the logical block attribution range refers to the logical block number range corresponding to each page block in the solid-state storage structure, and the attribution logical storage area of the page block is judged by comparing the consistency of the page block number and the logical block number. The actual scheduling behavior refers to the execution order of the page block in the write path and the mapping state of the logical block attribution.

[0008] As a further scheme of the application, the specific steps of S1 are: S101: Obtain the write request sequence in the Internet of Things edge node data cache, extract the page block number and timestamp corresponding to each request, extract the write instructions with the same page block number, sequentially arrange the corresponding timestamps in the order of time sequence, and obtain a page block time sequence action set; S102: Based on the action sequence under the number in the page block time sequence action set, extract the time value and action number between the continuous actions, extract each write sequence in the operation set area, and output the write fragment of the action connected page block, to obtain a continuous operation fragment list; S103: Based on the corresponding relationship between each action fragment in the continuous operation fragment list and the page block number, the action fragment is attributed to the corresponding number, and the action content under each number is output separately, to obtain a page block write operation stream set.

[0009] As a further scheme of the application, the specific steps of S2 are: S201: Based on the page block write operation stream set, extract the continuous operation item corresponding to each page block number, analyze according to the time sequence of operation instructions, filter the number sequence that continuously appears in the time sequence, and obtain a page block continuous number set; S202: Based on the page block continuous number set, compare the time interval and action number value change state between adjacent operations, split the number paragraph with reduced interval and dense action distribution, extract the number sequence with consistent time difference amplitude before and after the operation, and obtain an operation behavior trend group; S203: Based on the operation behavior trend group, analyze the coincidence degree of action time distribution trend and instruction number appearance frequency, identify the number set that appears operation continuity in the same page block number, and obtain a target page block operation identification list.

[0010] As a further scheme of the application, the specific steps of S3 are: S301: Based on the page block number in the target page block operation identification list, extract the corresponding order index in the write scheduling buffer pool, call the current scheduling state table to locate the index position of the page block number, verify the order value of each number in the buffer area, and obtain a page block order index sequence; S302: Call the order value in the page block order index sequence, extract the mapping segment corresponding to the number in the buffer mapping table, refer to the current mapping relationship of the number in the queue, transfer the corresponding position of the number in the original queue to the execution segment, and obtain a buffer mapping position set; S303: Based on the number segment mapping result in the buffer mapping position set, synchronize the page block number sequence in the current scheduling queue, organize the new sequence after number mapping to the end of the adjustable execution queue, and obtain a reconstructed page block buffer order table.

[0011] As a further scheme of the present application, the specific steps of S4 are: S401: Based on the page block number in the reconstructed page block buffer order table, retrieve the logical block belonging number corresponding to each number in the solid state storage distribution structure, extract the page block number item by item according to the pointing relationship of logical block belonging, merge the same belonging items into the corresponding list of the same logical block, and obtain the page block logical belonging sequence set; S402: Based on the page block logical belonging sequence set, filter the page block items with consistent logical block belonging numbers, compare the order connection between page block numbers, identify the page block groups that are continuously associated in logical distribution, and obtain the logical block page block association set; S403: Based on the logical block page block association set, merge the page blocks with consistent belonging numbers into the same logical block channel under the write path, connect the action connection order between page blocks, extract the sequence in the page block set under each logical block, and obtain the logical centralized page block list.

[0012] As a further scheme of the present application, the specific steps of S5 are: S501: Based on the logical block number corresponding to each group of page blocks in the logical centralized page block list, call the Internet of Things node identity in the node configuration parameter set, locate the index mapping path between the page block number and the node identifier, and compare the matching items between the number and the identifier to obtain the page block node mapping sequence; S502: Based on the page block node mapping sequence, extract the physical storage path set corresponding to the page block number, filter the write path content in the mapping path, and compare the paths according to the order sequence of the page block number to obtain the page block write path chain group; S503: Call the node pointing field of the path item in the page block write path chain group, and perform bit retrieval with the page block logical block number field, verify the mapping structure connection state between the belonging number and the scheduling order of the page block appearing in the path chain, and obtain the data node write organization structure.

[0013] A computer data storage system based on Internet of Things, comprising: The request extraction module obtains the write request sequence in the Internet of Things edge node data cache, filters the same instructions according to the page block number, extracts the operation content according to the time sequence, and obtains the page block write operation flow set; The behavior recognition module is based on the page block write operation flow set, repeatedly appears the page block number, compares the operation number and the adjacent time interval change order, merges and repeatedly writes the output sequence in the action set, and obtains the target page block operation identifier list; The order control module is based on the target page block operation identifier list, extracts the corresponding order index in the write scheduling buffer pool, changes the number position using the buffer mapping table, adjusts the page block into the current write execution area specified paragraph, and obtains the reconstructed page block buffer order table. The logical belonging module extracts the logical block number corresponding to each number based on the reconstructed page block buffer order table, compares the belonging number to screen the page block number consistent data path, and continuously writes the flow to obtain the logical centralized page block list; The structure mapping module extracts the corresponding relationship between the page block number and the Internet of Things node identifier based on the logical centralized page block list, extracts the write path and sequence information, associates the logical block belonging range and the actual scheduling behavior, and obtains the data node write organization structure.

[0014] Compared with the prior art, the advantages and positive effects of the present application are: In the present application, the priority write sequence is extracted through the page block number and the operation timing, the page block arrangement structure in the buffer area is reconstructed, the page blocks with associated characteristics are aggregated into a continuous path by combining the logical belonging relationship, the centralized write flow in the logical block is established, the write distribution state is extracted through the corresponding relationship between the node identifier and the page block number, and the data organization structure with the belonging boundary, path connectivity and scheduling order characteristics is generated, thereby enhancing the local intensive degree, structure connection ability and cooperation and coordination between cross-node paths, and improving the organization of the write path and the consistency of the data distribution structure. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0016] Figure 1 The step flowchart of the present application is shown in the figure. Figure 2 The S1 refinement diagram of the present application is shown in the figure. Figure 3 The S2 refinement diagram of the present application is shown in the figure. Figure 4 The S3 refinement diagram of the present application is shown in the figure. Figure 5 The S4 refinement diagram of the present application is shown in the figure. Figure 6 The S5 refinement diagram of the present application is shown in the figure. Figure 7 The system module diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0017] The technical solutions in the present application will be described below with reference to the drawings.

[0018] In the embodiments of the present application, the words such as "exemplary", "for example", etc. are used to represent an example, illustration, or description. Any embodiment or design scheme described as "exemplary" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "exemplary" is intended to present the concept in a specific manner. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0019] In the embodiments of the present application, "image" and "picture" can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent. "Of", "corresponding" and "relevant" can be used interchangeably at times. It should be pointed out that when the distinction is not emphasized, the meanings expressed are consistent.

[0020] In the embodiments of the present application, sometimes the subscript such as W1 can be written in the form of non-subscript such as W1. When the distinction is not emphasized, the meanings expressed are consistent.

[0021] In order to make the technical problems, technical schemes and advantages to be solved by the present application clearer, specific embodiments will be described in detail below with reference to the drawings.

[0022] Please refer to Figure 1 The embodiments of the present application provide a computer data storage method based on Internet of Things, comprising the following steps: S1: Obtain the write request sequence in the data cache of the Internet of Things edge node, extract the write instruction with the same number according to the page block number, arrange the instructions according to the time sequence, describe the continuous write of the page block according to the time interval and the number change of the adjacent operations, and obtain the page block write operation flow set; S2: Based on the number and sequence continuity of the operations corresponding to the page block in the page block write operation flow set, extract the page block number that appears repeatedly, compare the time interval and the number distribution between the adjacent operations, identify the current write behavior state according to the change of the operation distribution in the time period, and obtain the target page block operation identification list; S3: Based on the page block number in the target page block operation identification list, extract the corresponding order index in the write scheduling buffer pool, change the arrangement sequence of the number in the current scheduling queue through the buffer mapping table, and adjust the corresponding position in the front execution area to obtain the reconstructed page block buffer order table; S4: Based on the page block number in the reconstructed page block buffer order table, extract the logical block belonging number corresponding to each number in the solid state storage distribution structure, extract the page blocks with the same belonging number, and associate the page blocks to the centralized write sequence under the unified logical block to obtain the logical centralized page block list; S5: Based on the logical block number corresponding to each group of page blocks in the logical centralized page block list, the identity of the Internet of Things node is mapped and called to extract the actual write distribution path corresponding to the page block from the differentiated node, compare the logical attribution source and scheduling order state between the page blocks, output the organization structure of the associated logical block of the page block under the node, and obtain the data node write organization structure.

[0023] The page block write operation flow set includes a page block number sequence, a time interval sequence, and a write quantity change feature. The target page block operation identifier list includes a repeated page block number, a time interval comparison value, and a behavior state type. The page block buffering order table includes a page block scheduling index, a buffering mapping relationship, and an execution queue position. The logical centralized page block list includes a page block attribution logical block number, a page block grouping structure, and a centralized write sequence. The data node write organization structure includes a node identity mapping relationship, an actual write path, and a logical block association state.

[0024] Please refer to Figure 2 , the specific steps of S1 are as follows: S101: Obtain the write request sequence in the Internet of Things edge node data cache, extract the page block number and timestamp corresponding to each request, extract the write instruction with the same page block number, and sequentially arrange the corresponding timestamp in the number corresponding structure in order, to obtain the page block time sequence action set; First, the data cache of the edge node is continuously scanned, the current write log sequence is called through the cache monitoring interface, all write operation requests are extracted as a linear sequence, each request needs to record the corresponding target page block number and the timestamp triggering the request at the time of extraction, the page block number can be obtained by mapping the target address in the request instruction, for example, the request address is 0xAB20C, the mapped page block number is 37, the timestamp is read through the local cache timestamp register and bound to the request, for example, the request triggers the register value to be 157324, then the timestamp is the bound timestamp of the write instruction, after completing the numbering and time recording of all requests, all requests are traversed in turn, whether the page block numbers are consistent is compared to identify the instruction set with the same page block number, if there are multiple request numbers that are 37, all instructions under the number are classified into the same action sequence set, then the timestamps in the set are sorted, the sorting method adopts ascending operation, that is, the size relationship of two timestamp values is compared in order and rearranged from small to large, for example, the timestamps in the set are 157324, 157322, 157325 and 157321, the order after sorting is 157321, 157322, 157324 and 157325, the stable sorting mechanism needs to be called in the sorting process to ensure that the same timestamp request maintains the original sequence position, after completing the sorting, the instructions corresponding to the sorted timestamps are arranged in the time-ordered action set under the corresponding page block number in turn, and the page block time-ordered action set is obtained.

[0025] S102: Based on the action sequence under the number in the page block time-ordered action set, the time value and the action number between the continuous actions are extracted, each write sequence is extracted according to the operation set area, the write segment of the action connected page block is classified, and the continuous operation segment list is obtained. First, the time interval values between adjacent actions and the cumulative number of actions under the corresponding number are extracted one by one. When extracting the time interval, the time stamp of the current action and the time stamp of the previous action are read and the difference between the two is calculated. If the current time stamp is 208411 and the previous time stamp is 208399, the interval between the two is 12. In this way, the entire action sequence is traversed, and the adjacent time interval values are obtained in turn. For the number of actions, the total number of write instructions appearing under the same page block number is counted. For example, if the page block number is 25 and the corresponding action sequence contains 9 instructions, the number of actions is 9. After completing the above steps, each write sequence needs to be extracted according to the operation set area. That is, taking the page block number as the index, the continuous action fragments with a time interval below a certain threshold in the area are extracted to form an aggregation area. The setting of this threshold needs to be evaluated in combination with the actual cache response frequency and write rate. Generally, it is set to 8 or less. When the time difference between two adjacent actions is ≤8, it is considered as continuous operation. For example, if the adjacent time intervals in the action sequence of a certain page block are 5, 3, 9, 2, and 7, only the action with a time difference of 9 and the previous and next actions are considered as discontinuous, and the rest can be grouped into an aggregation fragment. If the aggregation fragment contains multiple action sequences with a time interval below the threshold, the connected action interval is extracted in its entirety. The start of each extraction is the first action that meets the time threshold, and the end is the previous action that appears above the threshold. In this way, several write fragments with time continuity can be extracted from a page block. For the determination of action-connected page blocks, the continuity of the page block number in each fragment needs to be judged. That is, whether the numbers between the fragments are in an increasing, decreasing, or same state is compared. If the previous fragment number is 26 and the next fragment number is 27, it is determined that the two are connected page blocks, and a continuous operation fragment list is obtained.

[0026] S103: Based on the correspondence between each action fragment in the continuous operation fragment list and the page block number, the action fragments are grouped under the corresponding number, and the action content under each number is output separately to obtain a set of page block write operation streams; First, each segment of the continuous operation segment list needs to be extracted, the page block number attached to the start and end actions in the segment is read, and the number is taken as the attribution label of the current action segment. If a segment of operation contains a start number of 18 and an end number of 18, it means that the action segment belongs to page block number 18. The next step is to traverse all the action segments and complete the number classification operation. The same number matching process is performed for each action segment. The belonging number position is confirmed by comparing the segment label with the items in the preset number set one by one. If the matching is successful, the segment is added to the action sequence group corresponding to the number. The matching method uses character value consistent judgment. If it is completely consistent with the existing number 18, it is considered to be matched successfully. If a number has not appeared in the set during the process, a corresponding number item is newly created and the action segment content is written for the first time. During the execution process, multiple action segments may appear under each number, and the content needs to be sequentially appended. The appending order is sorted according to the appearance order of the segments in the continuous operation segment list. The sorting judgment is based on the ascending comparison of the start time stamp in the action segment. For example, there are three segments under number 21, and the start time stamps are 104012, 104025 and 104018. The sorting needs to be 104012, 104018 and 104025. The corresponding content is also sequentially appended to number 21 in this order. After completing the classification of all action segments, the action segments aggregated under each number are extracted, the operation action instructions in each segment are read, and the instructions are written in the original order in the segment. The time stamp, write value and target address field are preserved during the writing process. Finally, these action instructions are organized to form a write operation set classified by page block number. For example, the operation set under page block number 30 contains instructions [104013, VAL=23, ADDR=0x1A32], [104015, VAL=25, ADDR=0x1A34] and [104020, VAL=26, ADDR=0x1A36]. The page block write operation stream set is obtained.

[0027] Please refer to Figure 3 The specific steps of S2 are: S201: Based on the page block write operation stream set, extract the continuous operation item corresponding to each page block number, analyze the operation instructions in time order, and screen the number sequence that appears continuously and repeatedly in time order to obtain the page block continuous number set; Firstly, all write instruction sequences under the corresponding numbered items in the set are read, and these instructions are rearranged from small to large according to their timestamp fields. During the rearrangement process, the timestamp values of any two instructions are compared. If the timestamp of the former is greater than that of the latter, the positions are exchanged. For example, the timestamps of the three instructions under the page block numbered 32 are 24018, 24012, and 24015, respectively. After comparison, the order is adjusted to 24012, 24015, and 24018. At this time, the continuous operation sequence numbered 32 is formed. Subsequently, all operation items in the sequence are read, and the numbering fields between adjacent operations are compared pair by pair to determine whether the numbering of adjacent items is consistent. When the same number appears continuously, these repeated numbered continuous intervals are recorded as a repeated sequence. The judgment method is to directly compare whether the integer values of the numbering are equal. If the numbering fields of the three consecutive operations are all 32, it is determined that this section is a repeated numbered interval. When the adjacent numbering is inconsistent, the current interval is ended and a new recording section is started. In the judgment process, a time continuity constraint condition needs to be added, that is, the time difference between adjacent operations should not exceed a set time interval threshold. The threshold is set to 10 or less according to the cache operation frequency. When the time difference of consecutive operations is all ≤10 and the numbering is consistent, it is considered as a repeated continuous section. For example, the operation times are 24012, 24015, 24019, and 24032, respectively. The intervals between the first two groups are 3 and 4, both of which are ≤10 and the numbering is consistent, so a continuous repeated sequence is formed. The fourth time difference is 13>10, so the continuous judgment is terminated at this point, the previous continuous interval is recorded, and a new interval recording is started. All page block numbers need to perform the above judgment steps. In the analysis process, the number of continuous repeated sections corresponding to each page block number is counted and stored. In the screening stage, all continuous section statistical results of the numbering are traversed, and the numbering sequence with an occurrence frequency ≥2 is selected as the continuous repeated numbering. The screening operation is based on the number comparison process. If the continuous section count of the numbering 45 is 3, and the count of the numbering 46 is 1, then the numbering 45 meets the condition and the numbering 46 does not meet the condition. Finally, the page block continuous numbering set is obtained.

[0028] S202: Based on the page block continuous numbering set, compare the time interval and action value change state between adjacent operations, split the numbering paragraphs with reduced interval and dense action distribution, extract the numbering sequence with consistent operation time difference amplitude, and obtain the operation behavior trend group; First, read each number corresponding to the operation record sequence in each section, and extract the timestamp field and write value field of each operation instruction. For any pair of adjacent operation records, first calculate the operation time interval by subtracting the timestamp, for example, the timestamp of the two consecutive operations under number 47 is 53120 and 53128, then the interval is 8, then continue to read the write value field and calculate the change amplitude between adjacent values, if the write value changes from 54 to 59, the change amplitude is 5, after completing the extraction of the time difference and action value difference of all adjacent operations in the number, each pair of data is sequentially stored in the time interval sequence and the action value difference sequence corresponding to the number, then the interval change trend identification operation is performed, that is, the difference value is calculated by subtracting the time interval of the two pairs of time interval, to judge the change direction and amplitude, if the previous interval is 10 and the next interval is 7, the difference value is -3, indicating that there is a trend of interval reduction, record all the number paragraphs that meet the time interval difference value as negative as trend paragraphs, when judging the action density, the operation frequency in unit time needs to be calculated, that is, the total number of instructions in a certain time interval of a number segment is counted and divided by the time length, if the judgment window is set to 30 time units, there are 5 operations in this segment, then the action density is 0.167, set the density threshold to 0.15, when the density is greater than or equal to 0.15, it is judged as dense distribution, in actual application, the number paragraphs that meet the above two conditions at the same time need to be split, these paragraphs are separated from the original continuous number set to form a new trend candidate group, after splitting, the remaining operations also need to be judged for time difference consistency, that is, whether all adjacent time differences in a number paragraph are within a certain interval, the fluctuation value of this interval is limited within ±2, if the time difference of the number paragraph is 5, 6, 4, 5, 6, the difference between the maximum difference and the minimum difference is 2, which meets the consistency requirement, then this number paragraph is extracted as the time consistency trend sequence, finally, all the number paragraphs that meet the interval reduction, dense distribution and time difference consistency are aggregated to obtain the operation behavior trend group.

[0029] S203: Based on the operation behavior trend group, analyze the coincidence degree of the action time distribution trend and the instruction number appearance frequency, identify the number set with operation continuity in the same page block number, and obtain the target page block operation identifier list; First, the timestamp sequence and the corresponding instruction quantity sequence are extracted, and the two groups of data are compared one by one according to the numbering correspondence. First, the timestamp sequence is sorted to determine the time distribution range of the action, and the minimum timestamp is taken as the start of the interval and the maximum timestamp is taken as the end of the interval. The entire time axis is divided into equal-width interval segments, and the width of each interval segment is set to 5 time units. The width is determined according to the average write response time of the edge node. If the average response time of the node is 4.7, it is rounded to 5, then every 5 time units is a segment. The number of instructions in each interval segment is counted and recorded in the quantity distribution table. For example, in the trend group with number 52, there are 3 instructions in the time period 0-5, 2 instructions in the time period 5-10, and 5 instructions in the time period 10-15. The quantity distribution table [3, 2, 5] is formed. After obtaining the table, the time distribution trend is analyzed. Specifically, the change difference of the number of instructions in adjacent interval segments is calculated. If the change is positive, it is judged as an increasing trend, and if the change is negative, it is judged as a decreasing trend. For example, the change sequence of [3, 2, 5] is (-1, +3), which indicates that the previous interval decreases and the next interval increases. Then the coincidence degree is judged by comparing whether the time trend change direction and the instruction quantity change direction are consistent. If they are consistent, it is recorded as 1, and if they are not consistent, it is recorded as 0. The proportion of 1 is summed up in all interval segments. If the proportion is ≥0.6, it is considered that the coincidence degree is high, 0.3≤proportion<0.6 is medium, and proportion<0.3 is low. The coincidence judgment threshold is set to 0.6 according to the average time trend synchronization proportion in the test sample. For example, in number 52, the coincidence count is 2, the total comparison segment number is 3, and the proportion is 0.67, which is classified as high coincidence. Then, in the same page block number range, the number groups of high coincidence are identified. The continuity is determined by continuous number comparison judgment operation. When the difference between adjacent numbers is ≤1, it is judged as a continuous number group. For example, the number sequence [51, 52, 53, 55], where 51 and 52 differ by 1, 52 and 53 differ by 1, and 53 and 55 differ by 2, [51, 52, 53] is identified as a group of continuous numbers. The determination of continuous numbers also needs to consider the time continuity, that is, whether the difference between the last operation time of adjacent numbers is ≤a set threshold. The threshold is 10. If the last operation timestamp of number 52 is 30045 and the first operation timestamp of number 53 is 30050, the difference is 5≤10, which satisfies the time continuity condition. Finally, the target page block operation identification list is obtained.

[0030] Please refer to Figure 4 The specific steps of S3 are: S301: Based on the page block number in the target page block operation identification list, the corresponding order index is extracted in the write scheduling buffer pool. The current scheduling state table is called to locate the index position of the page block number. The order value of each number in the buffer area is verified to obtain the page block order index sequence. First, the list is traversed, and the numbered items are read one by one. For each number, the write scheduling buffer pool is called to extract the corresponding index position. When reading, the page block number is used as the query key, and the buffer pool index table is used to perform number matching comparison. If the page block number in the current cache record is exactly the same as the target number, the order index value of the record is extracted, and recorded in the temporary index set. For example, the page block number is 72, the corresponding buffer pool record serial number is 14, and the order index value is 14. The value and the number corresponding record are (72, 14). The extraction operation of all numbers is performed in turn until the list traversal is completed. Then, the extracted index value needs to be verified for scheduling state. The current scheduling state table is called, and the page block number is used as the input parameter to query the real-time scheduling order state field corresponding to the number. This field contains the position sequence value and priority value of the current page block in the buffer area. When performing number positioning operation, the number recorded in the scheduling table is compared with the index number returned by the buffer pool. If they are consistent, the positioning is successful. Otherwise, the updated position index of the buffer pool needs to be re-searched. During the positioning process, the sequence value field is used for judgment. When the sequence value is 0, it means that the page block is located at the head of the scheduling queue. When the sequence value is 1 to 10, it means that it is in the middle section. When the sequence value is greater than 10, it is the end section. For example, the number 72 has a sequence value of 5 in the scheduling state table, which means that the page block is located in the middle of the scheduling queue. After completing the positioning, the accuracy of the order value in the buffer area needs to be verified. The extracted order index is compared with the sequence value in the scheduling state table. If the absolute difference between the two is ≤2, it is determined that the number is in the normal order in the buffer area. Otherwise, it is marked as a position offset number. The setting of the offset threshold is based on the reference of the cache scheduling update period. If the maximum order difference caused by the update period is 2, the threshold is set to 2. During the verification process of all numbers, the order comparison operation is performed for each record, and the comparison result is written into the order sequence list. For example, the page block number 72 has an order index of 14, and the scheduling table has a sequence value of 15. The difference between the two is 1≤2, and the record (72, 14, normal) is recorded. If the number 73 has an order index of 8, and the scheduling table has a sequence value of 11, the difference is 3>2, and the record (73, 8, offset) is recorded. The page block order index sequence is obtained.

[0031] S302: Call the order value in the page block order index sequence, extract the mapping section corresponding to the number in the buffer mapping table, refer to the current mapping relationship of the number in the queue, transfer the corresponding position of the number in the original queue into the execution section, and obtain the buffer mapping position set; First read the entire record of the sequence, for each record execution order value parsing operation, extract the page block number and the corresponding order value field, order value as search key input buffer mapping table, through the index column of mapping table line by line comparison and find the row item which is equal to order value, if match success, read the number corresponding to the mapping section number from the row, for example, the order value of page block number 81 is 12, the 12th row in the mapping table corresponds to section number C05, then record the number and section number as (81, C05), and perform the same operation on all page blocks until the mapping is completed. After the extraction is completed, the mapping relationship of the current number in the queue needs to be verified by reference. Specifically, read the arrangement order of each number in the original scheduling queue, and compare the order value of the number with the order value in the mapping table. If the absolute value of the order difference is ≤1, it is considered that the mapping relationship remains stable, otherwise the current queue position needs to be reacquired. If the order value of number 83 is 9 and the position of number 83 in the scheduling queue is 8, the difference is 1, which is stable. If the order value of number 84 is 6 and the position in the queue is 9, the difference is 3, which is marked as a relationship offset item. The offset threshold is 1, which comes from the maximum allowed order offset number in the buffer queue rearrangement period, which is 1 time. All stable numbers need to perform position transfer operation, that is, the original queue position corresponding to the number is migrated to its corresponding execution section position in the mapping table. In the execution process, the target position number is determined by index mapping operation. If number 81 is currently in queue position 12 and the corresponding execution section is C05, the number in position 12 is marked to move into section C05. Arrange multiple numbers in the same section according to the order value in ascending order. If there are numbers 81, 82 and 85 in section C05, the arrangement order is 85, 81 and 82. After the arrangement is completed, a new execution section number sequence is formed. After the processing of all sections is completed, the number distribution in each section is recorded to obtain the buffer mapping position set.

[0032] S303: Based on the number section mapping result in the buffer mapping position set, synchronize and replace the page block number order in the current scheduling queue, organize the new order after number mapping to the end of the adjustable execution queue, and obtain the reconstructed page block buffer order table; First, the entire item recorded in the mapping position set needs to be traversed, the page block number and the corresponding execution section identifier in each item are extracted, then the order sequence of all page block numbers in the current queue is read from the current scheduling queue, and the current position index mapping relationship table is established, so as to calibrate the original position for reference in the subsequent operation process. After the reading is completed, the order update operation needs to be performed for each mapping number, the original index position of the number is recorded after the number is removed from the original queue position, for example, the page block number 95 has the original index of 17 in the scheduling queue, and is mapped to the section C02 in the buffer mapping set. After the removal is completed, the position is marked as "vacant", and the vacant mark is represented by a special symbol, such as NULL mark position, so that the position cannot be incorrectly reused in the subsequent operation. Next, the number is arranged at the end of the adjustable execution queue according to the ascending order arrangement rule of the section number, if multiple page block numbers are mapped to the same section, the ascending order arrangement is further performed according to the order index value of the number in the mapping set, for example, the order index value of the number 95 under the section C02 is 8, and the order index value of the number 96 is 6, then the numbers are arranged in the order of 96, 95. If the order index values of the numbers in different sections conflict, the section number is given priority, and then the order index value is given priority. When the insertion operation is performed, the number records are added from the end of the original scheduling queue to the rear, and the number arrangement order mark in the queue is updated, the new order number is attached to each inserted number, and the new order number is incremented by 1 from the last order number at the tail of the current queue, for example, the number 101 at the tail of the scheduling queue corresponds to the order number 22, then the order number of the newly inserted number 102 is 23, and the order numbers of the subsequent numbers are 24, 25, etc. The order number field needs to be stored in pairs with the number. After the insertion of all numbers is completed, the reordering operation is performed on the entire scheduling queue, the NULL mark position is compressed and removed, all valid numbers are aggregated forward, and the new order position and index position relationship table of each record is updated, and finally the reconstructed page block buffer order table is obtained.

[0033] Please refer to Figure 5 The specific steps of S4 are as follows: S401: Based on the page block numbers in the reconstructed page block buffer order table, the logical block belonging numbers corresponding to each number are retrieved in the solid state storage distribution structure, the page block numbers are extracted item by item according to the pointing relationship of the logical block belonging, the same belonging items are aggregated into the corresponding list of the same logical block, and the page block logical belonging sequence set is obtained; First, read all the numbered records in the table, and take the page block number as the index item in the solid state storage distribution structure to perform logical block attribution retrieval operation, when retrieving, through reading the logical allocation table in the solid state medium, the corresponding relationship between the page block number field and the logical block number field is matched one by one, when the page block number is completely consistent with the record in the table, the logical block attribution number corresponding to the number is extracted, and is recorded in the form of key value pair, for example, page block number 120 corresponds to logical block number L05, then it is recorded as (120, L05), after completing the attribution retrieval of all numbers, the mapping set of page block and logical block number is formed, then the merging extraction operation is performed according to the pointing relationship of logical block attribution, it is judged whether the attribution number fields of adjacent page block numbers are consistent, if the continuous numbers have the same logical block number, they are aggregated into the same logical block list, if they are not consistent, the current aggregation is ended and a new group record is started, for example, page block numbers 120, 121 and 122 correspond to logical block L05, and number 123 corresponds to logical block L06, then 120 to 122 are attributed to L05 group, and number 123 is separately attributed to L06 group, in the aggregation process, the validity of the logical block pointing relationship needs to be detected, the attribution continuity is judged through the pointing field of the logical allocation table, when the pointing field values of adjacent logical blocks are continuously increased or equal, they are considered to belong to the same section, if the field is interrupted, it is divided into an independent section, in order to prevent number mismatch, attribution verification threshold K needs to be set, when the difference value of the logical block pointing field of continuous attribution is ≤K, it is considered to belong to the same group, the value of K is set to 1 according to the cross-zone continuity of the page block mapping of the solid state medium, when the difference value of the logical block pointing field is 0 or 1, it is considered to belong to the continuous relationship, for example, the difference value of L05 and L06 is 1, then they are considered to belong to the continuous section attribution item, and the number set is merged into the logical block L05-L06 attribution linked list, if the difference value is greater than 1, they are separately split to form independent groups, after merging, the page block numbers in each logical block list are sorted in ascending order, so that the number arrangement meets the storage physical order requirement, for example, logical block L05 group contains page block numbers 120, 121, 122 and 118, after sorting, they are 118, 120, 121 and 122, the sorting rule is realized by comparing the integer values of the numbers, and finally the page block logical attribution sequence set is obtained.

[0034] S402: Based on the page block logical attribution sequence set, the page block items with consistent logical block attribution numbers are screened, the order connection between the page block numbers is compared, the page block groups continuously associated on the logical distribution are identified, and the logical block page block association set is obtained; Firstly, all logical block records in the sequence set are read, and item-by-item filtering operation is performed on the page block number set corresponding to each logical block. In the filtering process, the logical block belonging number field of each page block record is read, and it is judged through equal comparison whether the belonging number of the current page block is consistent with the logical block record number. If consistent, it is retained; if inconsistent, it is rejected. For example, there are page block numbers 140, 141, 142 and 145 in the sequence of logical block L07, among which the number 145 is marked as L08 in the belonging field. The comparison result is inconsistent, that is, the number 145 is rejected, and the remaining numbers 140, 141 and 142 are included in the filtering result group of L07. After the belonging consistency filtering of all logical blocks is completed, the page block numbers in each logical block group need to be sorted in ascending order according to the numerical value, and a sorting list is established. After the sorting is completed, the page block order connection comparison stage is entered, and any two adjacent page block numbers in the sorting list are read in turn, and the number difference calculation operation is performed. If the adjacent number difference is 1, it is judged as sequential and continuous; if the difference is greater than 1, it is considered that there is a breakpoint at this place. For example, the sorting result is 140, 141, 142 and 145, the difference values of the continuous segments 140-142 are all 1, and the difference value of 142 and 145 is 3, forming a breakpoint. When identifying the continuous segment, a number difference threshold T needs to be set. If the difference value ≤ T, it is considered as a continuous segment, and if the difference value > T, it is segmented. T is set to 1 according to the physical continuous feature of the logical mapping of the page block. In the traversal comparison, the start and end numbers of all continuous segments are recorded, and their starting and ending position indexes in the logical block are marked. For example, the continuous segment (140-142) in the L07 block forms the index position (1-3), and the broken segment (145) forms the index position (4-4). Next, the associated judgment operation is performed on each logical block continuous segment. It is judged whether there is a continuous or equal pointing value in the logical pointing field between adjacent continuous segments. If the difference value of the logical pointing field of the adjacent segments ≤ 1, it is considered that the two continuous segments can be merged into a logical association group. If the difference value > 1, it remains independent. For example, the pointing field of the continuous segment (140-142) in the L07 block is 15, the pointing field of the broken segment (145) is 16, the difference value is 1, which meets the merging condition, and the associated group (140-145) is formed after merging. For the case of multiple continuous segments, such as (150-152), (153-155) and (156-157), the adjacent difference value relationship of each segment needs to be compared. If the difference value of the three continuous segments is 1, the whole can be merged into an associated group (150-157). If the intermediate difference value > 1, only the first two segments are merged into a local group. After the logical judgment is completed, the logical block page block association set is obtained.

[0035] S403: Based on the logical block page block association set, the page blocks with consistent belonging numbers are merged into the same logical block channel under the write path, the action connection order between the connected page blocks is connected, the sequence of the page block set under each logical block is extracted, and the page block list in the logical block is obtained. Firstly, all logical block number items in the association set are read, and a belonging matching operation is performed on each logical block number, by reading the belonging number field in the page block record, and performing an equal comparison with the logical block number field, when the field values are completely consistent, the page block number is marked as a belonging consistent item and added to the temporary list under the logical block channel, if the fields are inconsistent, the next record is skipped for comparison, for example, the page block number set under logical block L09 is 160, 161, 162, 165, 166, after comparison, it is found that the belonging number fields of page blocks 165 and 166 are both L09, and the remaining fields are different, only page blocks 165 and 166 are retained to enter the write path group of channel L09, after the belonging import operation of all logical blocks is completed, the imported page block set under each channel needs to perform an action connection sequence confirmation operation, specifically, the operation timestamp field and the action identification field of each page block are read, the timestamps are sorted in ascending order to determine the execution order of actions between page blocks, and the continuity of the time interval between adjacent page blocks is judged, when the time difference is ≤ threshold D, it is considered that the actions of two page blocks are continuous, the threshold D depends on the response period of the solid state cache, and is usually set to within 8 time units, for example, the timestamps corresponding to page blocks 165 and 166 are 51024 and 51030 respectively, the time difference 6≤8, and it is judged that the actions are continuous, after confirming the continuous actions, the action connection sequence between the two page blocks needs to be extracted, the operation type field of the action identification is compared to judge the connection relationship, if the action of the previous page block is write (W) and the action of the next page block is update (U), it is recorded as write-more connection, if the action is write, it is recorded as write-write connection, different combinations are recorded in the action sequence table in turn, so as to completely save the connection process, for example, the action of page block 165 in logical block L09 is W, and the action of page block 166 is U, the connection relationship is marked as (W-U), if there are multiple continuous page blocks, the multi-level connection is recorded in sequence, such as page blocks 160 (W), 161 (W), and 162 (U), which are marked as (W-W-U), forming an action connection sequence chain, after the connection judgment is completed, the page block numbers under the logical block are rearranged according to the order in the action sequence to form a page block set sequence, if there are multiple non-continuous action sequences under the logical block, they are independently stored in the corresponding list items, for example, two sequences [160, 161, 162] and [165, 166] are formed under logical block L09, and two records L09→[160, 161, 162] and L09→[165, 166] are generated in the record table, finally, all the page block set sequences under the logical blocks are integrated and summarized to obtain a logical block page block list.

[0036] Please refer to Figure 6 , the specific steps of S5 are: S501: Based on the logical block number corresponding to each group of page blocks in the logical centralized page block list, call the Internet of Things node identity in the node configuration parameter set, locate the index mapping path between the page block number and the node identification, and obtain the page block node mapping sequence by matching the matching items between the number and the identification; First, each logical block item in the list is traversed, the page block number set under it is read piece by piece, and the logical block number is taken as the retrieval key to search the corresponding Internet of Things node identity in the node configuration parameter set. The logical block field and the node identification field of the node configuration parameter table are compared item by item. When the logical block number field value is completely equal to the target logical block number, the node identification value in the record is extracted as the node belonging item of the current page block group. For example, the logical block number L10 matches the node identification N021 in the parameter set, then L10 and its subordinate page block number set are established in a corresponding relationship with N021 to form a preliminary mapping item (L10→N021). After the pairing of the logical block and the node identification is completed, the index positioning operation is performed on each page block number under the logical block. The page block number field is searched in the configuration parameter set to find its index mapping path field. The field records the logical mapping index value between the page block number and the node identification. The index value is determined by the serial number of the node in the data partition table. If the page block number 180 corresponds to the index value 7, its mapping path is represented as N021[7]. If the number 181 corresponds to the index value 9, it is represented as N021[9]. The index path is determined according to the node allocation interval range. Each node in the node configuration has a fixed number interval range, such as the N021 node covering the number range 160-199. When the page block number is within the interval, it belongs to the node. In the positioning process, the interval of the page block number needs to be compared. When the number value is greater than or equal to the interval start point and less than or equal to the interval end point, it is judged to be the node belonging item. If it is not within the range, the next node record is searched. After the belonging positioning of all page block numbers is completed, the matching verification stage between the number and the identification is entered. The page block number and the node identification of each mapping path record are compared for consistency. It is judged whether there are repeated or missing items. When a page block number appears in multiple nodes, the node with a smaller index value is preferentially attributed. The smaller index value is determined by directly comparing the integer size. If the number 180 exists in N021[7] and N022

[12] , the node with the index value 7 is taken as the belonging node. If the page block number does not exist in any node index interval, it is recorded as an isolated number and does not participate in the generation of the mapping set. Finally, all records verified by comparison are arranged in the form of a three-tuple of logical block number, page block number, and node identification, and the page block node mapping sequence is obtained.

[0037] S502: Based on the page block node mapping sequence, extract the physical storage path set corresponding to the page block number, filter the write path content in the mapping path, and perform joint comparison on the path according to the order of the page block number to obtain the page block write path chain group.First, the page block number recorded in the sequence is read item by item, and the corresponding physical storage path set is extracted through the node identification field. The physical storage path is stored in the form of a path field in the node configuration parameter table. Each node identification corresponds to multiple physical path items, such as the main path, write path, and read path. The path type field needs to be compared during the reading process. Only when the type field is "write" will the path record be extracted into the path set. If the page block number 200 corresponds to node N025, and there is an item " / mnt / n025 / write01" in the path set of the node N025, the item is recorded as the write path of the page block 200. If the path type is read or backup, it is skipped. After completing the path extraction of all page block numbers, a path index table needs to be constructed. The table takes the page block number as the primary key and the write path as the corresponding value, recording the mapping relationship between all page blocks and paths. For example, page block 200 -> / mnt / n025 / write01, page block 201 -> / mnt / n025 / write02, and page block 202 -> / mnt / n025 / write03. After completing the index construction, the path connection comparison needs to be performed according to the order of the page block numbers. The order value field of each page block in the mapping sequence is read, and the order values of adjacent page block numbers are compared. If the order value of the next number is greater than the order value of the previous number and the difference between them is 1, it is judged as a direct continuous path. If the difference is greater than 1, it is judged as a disconnected paragraph. In the continuous path determination, the prefix part of the path field of adjacent page blocks also needs to be compared. If the path prefix is the same, it is considered that the physical location of the path belongs to the same storage channel. If the path prefix is different, it is recorded as a cross-channel path segment. For example, the path prefixes of page blocks 200 and 201 are both " / mnt / n025 / ", so they are continuous. If the path of page block 202 is " / mnt / n026 / ", it is disconnected. When comparing the path connection, a prefix matching threshold L needs to be set to judge whether the prefix similarity of the path string meets the continuity. The value of L ranges from 0 to 1, representing the matching character proportion of the prefix. If the matching proportion is greater than or equal to 0.8, it is judged as a same-channel path. Otherwise, it is considered as disconnected. For example, the path " / mnt / n025 / write01" and " / mnt / n025 / write02" have a common prefix character length of 12, a total length of 15, and a matching proportion of 0.8, which meets the condition and is classified as a same-channel path. If the next path is " / mnt / n026 / write01", the prefix matching proportion is 0.6 discontinuous, record breakpoint position 200-201 as a chain, 202 start a new chain, after traversing all page block numbers and completing path connection, concatenate the paths of each continuous segment to form a path chain structure, record the starting page block number, the ending page block number, and the chain path content, if the chain contains page blocks 200, 201, 202 and all are continuous channel paths, then the chain group is chain (200-202) -> [ / mnt / n025 / write01, / mnt / n025 / write02, / mnt / n025 / write03], finally integrate and output all chain groups to obtain the page block write path chain group.

[0038] S503: Call the node pointing field of the path item in the page block write path chain group, and perform bit-by-bit retrieval with the page block logical block number field to verify the mapping structure connection state between the belonging number and the scheduling order of the page block in the path chain to obtain the data node write organization structure. First, read each path item in the path chain group one by one, and extract its attached node identifier and path pointing field, wherein the node identifier field represents the corresponding Internet of Things node number of the path, and the path pointing field records its terminal target address in the physical writing architecture. All path items in the path chain are aggregated in order of page block number to maintain sequence consistency for subsequent mapping comparison. Next, read the page block logical block number field. Compare whether the page block number and the logical block number corresponding to the path item match one by one. If the page block number 180 mapping path is / mnt / n025 / write01, the node identifier is N025, and the corresponding logical block number is L10, retrieve the home logical block number of page block 180 in the logical mapping table from the scheduling queue. If the retrieval result is L10 and the order position in the path chain is consistent with the logical block order, record this mapping as a consistent item. Otherwise, mark it as a sequence conflict item. In the comparison process, the number consistency and order structure consistency need to be judged respectively. The number consistency is accurately matched by the page block number field and the page block record in the logical block table. The order consistency is judged by the scheduling order field and the path chain order field. For example, the order of page block 180 in the scheduling order table is 12, and the corresponding order of 180 in the path chain is 12, which is consistent. If the difference between the two is greater than the set offset threshold T, it is determined to be inconsistent. T is set to 1 according to the synchronization accuracy of the writing queue, that is, if the page block order value offset is within ±1, it is normal, otherwise it is abnormal. After completing the path item and logical block number alignment retrieval, the structure consistency of all page block path chains under each node identifier is verified. Compare whether the home logical block numbers of all page block numbers under the same node are continuous. If page blocks 180, 181 and 182 under a certain node belong to L10, L10 and L11 respectively, they are considered to belong continuously. If page blocks 180, 181 and 183 belong to L10, L10 and L12 respectively, it is identified that the logical block number of number 183 has a non-continuous jump with the previous segment, and it is marked that there is a logical mapping fault in the node path. Record the start page block number and logical block number of the fault. Finally, record the path structure under each node as a writing organization path structure item, which includes node identifier, path chain start number, path chain order list, corresponding logical block number list and order comparison state label. All node path organization items are output to obtain the data node writing organization structure.

[0039] Please refer to Figure 7 A computer data storage system based on Internet of Things, comprising: The request extraction module obtains the write request sequence in the Internet of Things edge node data cache, filters the same instructions according to the page block number, extracts the operation content according to the time sequence, and obtains the page block write operation stream set; The behavior recognition module is based on the page block write operation flow set, repeatedly appears the page block number, compares the operation number and the adjacent time interval change order, concentrates and repeatedly numbers the action and writes into the output sequence, obtains the target page block operation identification list; The sequential regulation module is based on the target page block operation identification list, extracts the corresponding order index in the write scheduling buffer pool, changes the number position using the buffer mapping table, regulates the page block into the current write execution area specified paragraph, and obtains the reconstructed page block buffer order table; The logic attribution module is based on the reconstructed page block buffer order table, extracts the logic block number corresponding to each number, compares the attribution number, filters the page block number consistent data path, and continuously writes the flow, and obtains the logic concentrated page block list; The structure mapping module is based on the logic concentrated page block list, extracts the corresponding relationship between the page block number and the Internet of Things node identification, extracts the write path and the order information, associates the logic block attribution range and the actual scheduling behavior, and obtains the data node write organization structure.

[0040] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A computer data storage method based on the Internet of Things, characterized by, Comprise the following steps: S1: Obtain the write request sequence in the Internet of Things edge node data cache, filter the same instructions according to the page block number, extract the operation content according to the time sequence, and obtain the page block write operation flow set by the page block write process; S2: Based on the page block write operation flow set, extract the repeatedly appearing page block number, compare the operation number and the adjacent time interval change order, concentrate and output the sequence of the repeated number of actions, and obtain the target page block operation identification list; S3: Based on the target page block operation identification list, extract the corresponding order index in the write scheduling buffer pool, change the number position using the buffer mapping table, call the page block into the current write execution area specified paragraph, and obtain the reconstructed page block buffer order table; S4: Based on the reconstructed page block buffer order table, extract the number corresponding logical block attribution, compare the attribution number to filter the page block number consistent data path, and continuously write the flow, and obtain the logically concentrated page block list; S5: Based on the logically concentrated page block list, extract the corresponding relationship between the page block number and the Internet of Things node identification, extract the write path and sequence information, associate the logical block attribution range with the actual scheduling behavior, and obtain the data node write organization structure; The specific steps of S5 are: S501: Based on the logical block number corresponding to each group of page blocks in the logically concentrated page block list, call the Internet of Things node identity in the node configuration parameter set, locate the index mapping path between the page block number and the node identification, and obtain the page block node mapping sequence by comparing the matching items between the number and the identification; S502: Based on the page block node mapping sequence, extract the physical storage path set corresponding to the page block number, filter the write path content in the mapping path, and connect the paths according to the order sequence of the page block number, and obtain the page block write path chain group; S503: Call the node pointing field of the path item in the page block write path chain group, and perform bit retrieval with the page block logical block number field, verify the mapping structure connection state between the attribution number and the scheduling order of the page block appearing in the path chain, and obtain the data node write organization structure.

2. The computer data storage method based on Internet of Things according to claim 1, characterized in that, The page block write operation flow set includes page block number sequence, time interval sequence, and write number change characteristics, the target page block operation identification list includes repeated page block number, time interval comparison value, and behavior state type, the page block buffer order table includes page block scheduling index, buffer mapping relationship, and execution queue position, the logically concentrated page block list includes page block attribution logical block number, page block grouping structure, and concentrated write sequence, and the data node write organization structure includes node identity mapping relationship, actual write path, and logical block association state.

3. The computer data storage method based on Internet of Things according to claim 1, characterized in that, The write output sequence refers to the page block write instruction sequence formed by concentrating and outputting the page block operation with repeated number in time sequence after identifying the page block operation behavior; The scheduling buffer pool refers to a buffer structure for temporarily storing and sorting page block write tasks, and the scheduling order of the page block in the execution queue is controlled through index and mapping relationship.

4. The computer data storage method based on Internet of Things according to claim 1, characterized in that, The logical block attribution range refers to a range of logical block numbers corresponding to each page block in the solid-state storage structure, and the consistency of the page block number and the logical block number is used to determine the attribution of the logical storage area of the page block; The actual scheduling behavior refers to the execution order of the page block in the write path and the mapping state of the logical block attribution.

5. The computer data storage method based on Internet of Things according to claim 1, characterized in that, The specific steps of S1 are: S101: Obtain the write request sequence in the data cache of the Internet of Things edge node, extract the page block number and timestamp corresponding to each request, extract the write instructions with the same page block number, sequentially arrange the corresponding timestamps in the order of time into the number corresponding structure, and obtain the page block time order action set; S102: Based on the action sequence under the number in the page block time order action set, extract the time value and action number between the continuous actions, extract each write sequence in the operation set area, and obtain the continuous operation fragment list by attributing the write fragments of the actions connected page block; S103: Based on the corresponding relationship between each action fragment in the continuous operation fragment list and the page block number, attribute the action fragment to the corresponding number, separately output the action content under each number, and obtain the page block write operation flow set.

6. The computer data storage method based on Internet of Things according to claim 1, wherein, The specific steps of S2 are: S201: Based on the page block write operation flow set, extract the continuous operation item corresponding to each page block number, analyze according to the time order of the operation instruction, filter the number sequence that continuously appears repeatedly in the time order, and obtain the page block continuous number set; S202: Based on the page block continuous number set, compare the time interval and action number value change state between adjacent operations, split the number paragraph with reduced interval and dense action distribution, extract the number sequence with the same operation time difference amplitude, and obtain the operation behavior trend group; S203: Based on the operation behavior trend group, analyze the coincidence degree of the action time distribution trend and the instruction number appearance frequency, identify the number set that appears operation continuity in the same page block number, and obtain the target page block operation identification list.

7. The computer data storage method based on Internet of Things according to claim 1, characterized in that, The specific steps of S3 are: S301: Based on the page block number in the target page block operation identification list, extract the corresponding order index in the write scheduling buffer pool, call the current scheduling state table to locate the index position of the page block number, verify the order value of each number in the buffer area, and obtain the page block order index sequence; S302: Call the order value in the page block order index sequence, extract the mapping segment corresponding to the number in the buffer mapping table, refer to the current mapping relationship of the number in the queue, transfer the corresponding position of the number in the original queue to the execution segment, and obtain the buffer mapping position set; S303: Based on the number segment mapping result in the buffer mapping position set, synchronize the order of the page block number in the current scheduling queue, organize the new order after the number mapping to the end of the adjustable execution queue, and obtain the reconstructed page block buffer order table.

8. The computer data storage method based on Internet of Things according to claim 1, characterized in that, The specific steps of S4 are: S401: Based on the page block number in the reconstructed page block buffer order table, retrieve the logical block belonging number corresponding to each number in the solid state storage distribution structure, extract the page block number item by item according to the pointing relationship of logical block belonging, merge the same belonging item into the corresponding list of the same logical block, and obtain the page block logical belonging sequence set; S402: Based on the page block logical belonging sequence set, filter the page block items with consistent logical block belonging numbers, compare the order connection between page block numbers, identify the page block group continuously associated in logical distribution, and obtain the logical block page block association set; S403: Based on the logical block page block association set, merge the page blocks with consistent belonging numbers into the same logical block channel under the write path, connect the action connection order between page blocks, extract the sequence in the page block set under each logical block, and obtain the logical centralized page block list.

9. An Internet of Things based computer data storage system, characterized by, The system is used to realize the computer data storage method based on Internet of Things according to any one of claims 1-8, and the system comprises: The request extraction module obtains the write request sequence in the Internet of Things edge node data cache, filters the same instructions according to the page block number, extracts the operation content according to the time sequence, and obtains the page block write operation flow set; The behavior recognition module is based on the page block write operation flow set, repeatedly appears the page block number, compares the operation number and adjacent time interval change order, merges and repeatedly writes the output sequence, obtains the target page block operation identification list, and The order control module is based on the target page block operation identification list, extracts the corresponding order index in the write scheduling buffer pool, changes the number position using the buffer mapping table, adjusts the page block into the current write execution area specified paragraph, and obtains the reconstructed page block buffer order table; The logical belonging module is based on the reconstructed page block buffer order table, extracts the logical block number corresponding to each number, compares the belonging number to filter the page block number consistent data path, and continuously writes the flow, and obtains the logical centralized page block list; The structure mapping module is based on the logical centralized page block list, extracts the corresponding relationship between the page block number and the Internet of Things node identifier, extracts the write path and order information, associates the logical block belonging range and the actual scheduling behavior, and obtains the data node write organization structure.

Images