Edge computing cache updating method integrated with solid state disk

Abstract

The application discloses an edge computing cache updating method integrated with a solid state disk, and relates to the technical field of computers. The application realizes real-time monitoring of the access behavior of a solid state disk physical block, including the access frequency and the last access time interval, through an edge computing node, and constructs a high-heat feature evaluation system of the physical block based on the monitoring data. Then, the application selects the high-heat physical block with frequent access according to the high-heat feature, and performs real-time cache updating on the high-heat physical block, so as to realize the target of optimizing the cache performance. The technical core lies in dynamically monitoring the access behavior of the physical block of the solid state disk, analyzing the data features in combination with the edge computing node, effectively identifying and processing the high-heat physical block, and thus improving the cache efficiency and the data access speed of the solid state disk.

Description

Technical Field

[0001] This invention belongs to the field of computer technology, specifically relating to an edge computing cache update method integrated into a solid-state drive. Background Technology

[0002] With rapid economic development and the deepening of digital transformation, the demand for data generation, storage and processing is growing exponentially; solid-state drives (SSDs), as the core of storage devices, bear the task of efficient management of massive amounts of data and have become an important part of modern information technology.

[0003] In existing technologies, SSD caching management typically relies on static or simple dynamic strategies, such as Least Recently Used (LRU) or First-In-First-Out (FIFO) algorithms. These methods often lack deep integration with edge computing nodes, resulting in inefficient cache updates. Specifically, existing technologies typically cannot monitor changes in access patterns of physical blocks within the SSD in real time. Especially under high concurrency or dynamic workloads, caching strategies struggle to adapt to fluctuations in access frequency and last-access intervals, leading to decreased cache hit rates and increased response latency. Furthermore, existing methods often ignore differences in access patterns at different times of the day. For example, during peak business hours, the cache may not prioritize frequently accessed data, instead retaining infrequently used blocks. This not only wastes valuable cache space but also increases read / write wear on the SSD, shortening its lifespan. Additionally, existing technologies cannot dynamically adjust based on actual access, resulting in outdated or redundant cached data, further reducing overall system performance. Moreover, the potential of edge computing nodes has not been fully explored in existing technologies, lacking real-time interaction mechanisms, making cache updates dependent on centralized processing, introducing unnecessary network latency and computational overhead.

[0004] To address the aforementioned issues, this invention proposes an edge computing cache update method integrated into solid-state drives. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an edge computing cache update method integrated into solid-state drives, which solves the problems of rigid caching strategies, slow response, and uneven resource allocation in existing technologies.

[0006] The objective of this invention can be achieved through the following technical solutions: A method for updating edge computing cache integrated into a solid-state drive, the method comprising: Step 1: Use edge computing nodes that interact with the target SSD in real time to monitor the target SSD in real time, and determine the monitoring data group associated with each physical block in the target SSD within the preset update cycle within the preset monitoring cycle. Step 2: Based on the numerical performance of the monitoring data in the monitoring data group of each physical block within the corresponding monitoring period, evaluate the high-heat physical blocks associated with the corresponding monitoring period, obtain the high-heat data block sequence number, and update the high-heat physical block sequence number originally stored in the edge computing node corresponding to this monitoring period. Step 3: Based on the actual time of the target SSD, determine the monitoring period corresponding to this actual time, extract the corresponding high-heat physical block number from the edge computing node, and extract the data from the corresponding high-heat physical block from the target SSD and cache it in the cache area of ​​the target SSD.

[0007] As a further aspect of the present invention, in step one, the monitoring data in the monitoring data group includes the access frequency and the last access time interval associated with each physical block in the target solid-state drive. The access frequency includes read access frequency and write access frequency; The last access time interval is the time difference between the current time and the timestamp of the last time the physical block was accessed.

[0008] As a further aspect of the present invention, in step one, the duration of the update cycle and the monitoring cycle are both preset values ​​by the operator, denoted as T_jc and T_gx respectively; The update cycle T_gx spans from 0:00 to 24:00 of a day; The monitoring period T_jc spans one hour. Any update cycle T_gx contains m monitoring cycles T_jc, where m=24.

[0009] As a further aspect of the present invention, the specific method for determining the monitoring data group associated with each physical block in the target solid-state drive within a preset monitoring period in step one is as follows: Retrieve each physical block within the target solid-state drive, and assign a sequence number to each physical block based on the retrieval order. Then sort the physical blocks and denote them as the physical block sequence Q1, Q2, ..., Qj. j represents the total number of physical blocks, Qi represents the i-th physical block acquired, and i is the counting index, 1≤i≤j; Take 0:00 on any day as the start time of the update period T_gx and the first monitoring period T_jc within the update period T_gx; Within the first monitoring period T_jc, the access frequency and last access time interval associated with physical block Qi at different times are monitored and denoted as A1_Qi and B1_Qi, respectively. The access frequency A1_Qi and the last access time interval B1_Qi are both data streams, and the length of the data stream is the total number of times within the monitoring period T_jc. The combined access frequency A1_Qi and the last access time interval B1_Qi are used as the monitoring data group FW1_Qi for physical block Qi; Similarly, the monitoring data groups of each physical block in the physical block sequence within the first monitoring period T_jc are determined and sorted according to the order of the physical block sequence to obtain the first monitoring data group sequence FW1={FW1_Q1,FW1_Q2,...,FW1_Qj}; Repeat the above steps to determine the sequence of monitoring data groups in all monitoring periods T_jc within the update period T_gx, totaling m groups, which are represented in chronological order as: FW1, FW2, ..., FWm.

[0010] As a further aspect of the present invention, the specific method for evaluating the high-temperature physical blocks associated with the corresponding monitoring period in step two is as follows: Based on the first monitoring data group sequence, obtain the monitoring data group FW1_Qi of physical block Qi within the first monitoring period T_jc; Decompose the monitoring data set FW1_Qi to obtain the access frequency A1_Qi and the last access time interval B1_Qi; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the monitoring data as the vertical axis. Plot the access frequency A1_Qi and the last access time interval B1_Qi in chronological order on the constructed two-dimensional coordinate system to generate the access frequency change curve SA1_Qi associated with the access frequency A1_Qi and the last access time interval change curve SB1_Qi associated with the last access time interval B1_Qi. The access frequency change curve SA1_Qi does not intersect with the last access time change curve SB1_Qi, and the access frequency change curve SA1_Qi is located above the last access time change curve SB1_Qi. Based on the start and end times of the first monitoring period T_jc, construct straight lines L1 and L2 that are perpendicular to the horizontal axis and parallel to the vertical axis. Calculate the area of ​​the closed region formed by the straight line L1, the straight line L2, the access frequency change curve SA1_Qi, and the last access time change curve SB1_Qi. This area is used as the high-heat characteristic of physical block Qi in the first monitoring period T_jc and is denoted as Z1_Qi. Similarly, determine the high-temperature characteristics of each physical block in the physical block sequence Q1, Q2, ..., Qj within the first monitoring period T_jc, and sort them according to the order of the physical block sequence, denoted as the first high-temperature characteristic sequence Z1={Z1_Q1,Z1_Q2,...,Z1_Qj}; The values ​​of high-temperature features within the first high-temperature feature sequence Z1 are sorted from largest to smallest. The data blocks in the top α% are used as the high-temperature physical blocks within the first monitoring period T_jc, where α% is a percentage preset by the operator.

[0011] As a further aspect of the present invention, in step two, based on the method of determining the first high-temperature feature sequence Z1 within the first monitoring period T_jc, the high-temperature feature sequences of each monitoring period T_jc within the update period T_gx are determined, totaling m, and are represented in chronological order as: Z1, Z2, ..., Zm.

[0012] As a further aspect of the present invention, in step two, based on the method of determining the high-temperature physical block within the first monitoring period T_jc, the high-temperature physical block within each monitoring period T_gx is determined.

[0013] As a further aspect of the present invention, in step two, when any monitoring period T_jc ends, the high-temperature physical block sequence number associated with this monitoring period T_jc is obtained, and the high-temperature physical block sequence number originally stored in the edge computing node corresponding to this monitoring period is deleted and updated.

[0014] As a further aspect of the present invention, in step three, the specific method for extracting the data from the corresponding high-heat physical block of the target solid-state drive and caching it in the cache area of ​​the target solid-state drive is as follows: Obtain the actual time of the target solid-state drive, accurate to the second; The corresponding monitoring cycle identifier is determined based on the actual time, where the monitoring cycle identifier corresponds to the monitoring cycle T_jc; The monitoring cycle identifier takes the values ​​K1, K2, ..., K24, where 0:00 to 1:00 in a day corresponds to the monitoring cycle identifier K1, the monitoring cycle identifier K1 corresponds to the first monitoring cycle T_jc, and so on. Obtain the monitoring cycle corresponding to the actual time and its associated high-thermal physical block; Based on the sequence number of the high-temperature physical block, the data in the corresponding high-temperature physical block is extracted from the target solid-state drive, and the data in the cache area of ​​the target solid-state drive is deleted and updated.

[0015] The beneficial effects of this invention are: This invention achieves dynamic monitoring and evaluation of physical block access patterns through real-time interaction between edge computing nodes and solid-state drives (SSDs), thereby accurately identifying high-hot physical blocks and updating the cache index within a preset period. The advantage of this method is that it can adaptively capture changes in data access hotspots, ensuring that cached content always matches the current workload, significantly improving cache hit rate and data access efficiency. It effectively reduces read / write latency of SSDs, improves overall system response speed, and simultaneously reduces the burden on the main system through edge computing, achieving more efficient resource utilization and adaptive cache management. This invention achieves refined and continuous data collection of the access frequency and last access time interval of each physical block of a solid-state drive by setting a daily update cycle and an hourly monitoring cycle, thereby providing a comprehensive and timely view of the usage status. Its advantages are that it can capture short-term access fluctuations and long-term usage trends of the hard drive in real time, reduce manual intervention, reduce operating costs, and ultimately extend the lifespan of the solid-state drive and ensure the stability of data storage. This invention achieves accurate quantitative evaluation of high-heat physical blocks by dynamically monitoring and analyzing the access frequency and last access time interval of physical blocks, constructing a change curve and calculating the area of ​​the closed region as a high-heat characteristic. Its advantages are that it can automatically and continuously identify and update frequently accessed data blocks, optimize data access efficiency, reduce system latency, and ensure that resource allocation matches the current access mode through real-time updates of edge computing nodes, making it suitable for efficient data management in high-load environments. This invention achieves time-based data caching management by accurately acquiring the actual time of the target solid-state drive (SSD) and associating it with a monitoring cycle identifier. Its core advantage lies in its ability to dynamically identify and extract data from high-heat physical blocks and automatically update the cache, thereby significantly improving data access speed and system responsiveness. This time-driven mechanism ensures optimized resource utilization, reducing access latency by preloading frequently used data and avoiding unnecessary full-disk scans, thus reducing SSD wear and extending its lifespan. Furthermore, automated cycle management reduces manual intervention, improving system reliability and maintainability. Ultimately, while enhancing overall performance, it also achieves energy efficiency optimization and hardware protection, providing users with a smoother and more reliable data storage experience. Attached Figure Description

[0016] The invention will now be further described with reference to the accompanying drawings.

[0017] Figure 1 This is a flowchart illustrating the method described in this invention; Figure 2 This is a flowchart illustrating the method described in Embodiment 2 of the present invention; Figure 3 This is a flowchart illustrating the method described in Embodiment 3 of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1 The edge computing cache update method integrated into solid-state drives, such as Figure 1 As shown, this method includes the following: This method: The edge computing cache update method integrated with solid-state drives is mainly implemented based on solid-state drives and edge computing nodes. It should be noted that the edge computing nodes integrated with solid-state drives and interacting in real time have data analysis and data storage capabilities.

[0020] First, the target solid-state drive (i.e., the solid-state drive to be processed) is obtained. Then, the edge computing node that interacts with the target solid-state drive in real time is obtained. The edge computing node monitors the target solid-state drive and each physical block in the target solid-state drive (the target solid-state drive contains several physical blocks used for storing data) in real time. The edge computing node monitors and collects monitoring data sets of each physical block in each monitoring cycle preset by the operator within the update cycle preset by the operator. The duration of the update cycle and the duration of the monitoring cycle are determined by the operators based on the actual situation, generally based on one day (24 hours) and one hour respectively.

[0021] The monitoring data set is a key indicator for evaluating physical block access behavior, specifically including access frequency and the time interval between the last access. The access frequency includes read access frequency and write access frequency, which respectively record the number of times the physical block is read and written during the monitoring period. It should be noted that the access frequency is stored in the form of a data stream, representing the instantaneous access frequency value at different times, rather than the total number of accesses during the entire monitoring period. The last access time interval is defined as the time difference (in seconds) between the current time and the timestamp of the last access to the physical block. It is also recorded in the form of a data stream to reflect the access timeliness of the physical block. For example, if a physical block was last accessed 30 seconds ago at the current time, then the last access time interval of the physical block at this time is 30 seconds. If the physical block has not been accessed at the next time, then the last access time interval of the physical block at the next time is 31 seconds. If the physical block is accessed at the next time, then the last access time interval of the physical block at the next time is 0 seconds.

[0022] Once the monitoring data groups of all physical blocks within the target SSD in the corresponding monitoring period are determined, the monitoring data of this physical block are evaluated based on their numerical performance. The high-heat physical blocks associated with the corresponding monitoring period are identified, and the serial numbers of the high-heat data blocks are obtained. The serial numbers of the high-heat data blocks are stored in the edge computing node, and the serial numbers of the high-heat physical blocks associated with this monitoring period that were originally stored are deleted and updated.

[0023] By repeating this process, the high-heat physical blocks associated with each monitoring cycle within an update cycle can be identified. After the update cycle ends, the actual operation phase begins, where the target SSD and the current real time are monitored in real time. Based on the determined real time, the monitoring cycle corresponding to this real time is determined. Then, the corresponding high-heat physical block number is extracted from the edge computing node. Once the high-heat physical block number is determined, the data in the corresponding high-heat physical block can be extracted from the target SSD in real time and cached in the target SSD's cache area (such as DRAM cache). This ensures that the cache area always stores data that is currently or will be frequently accessed, thereby reducing the main storage access latency of the SSD and improving I / O performance. This step, through a time-driven caching strategy, achieves the matching of data prefetching and real-time access modes, optimizing the response speed of the storage system.

[0024] Example 2 This embodiment, based on embodiment 1, further discloses a method for determining the monitoring data sets associated with each physical block in different monitoring periods, such as... Figure 2 As shown, it specifically includes the following: Based on the content of Example 1, the update cycle and monitoring cycle preset by the operator are obtained, and the duration preset by the operator for the update cycle and monitoring cycle is determined, which are denoted as T_jc and T_gx respectively. As described in Example 1, the time span of an update cycle T_gx is from 0:00 to 24:00 of a day, which is 24 hours. A monitoring period T_jc spans one hour. It should be noted that one hour is from 0:00 to 1:00, or from 1:00 to 2:00, or similar hours.

[0025] Thus, it can be seen that any update cycle T_gx contains m monitoring cycles T_jc, and m is always equal to 24 (in this scheme, m is always equal to 24), corresponding to the time span of an update cycle T_gx: 24 hours.

[0026] Then, based on the constructed edge computing nodes, all physical blocks in the target solid-state drive are obtained, and each physical block is marked with a sequence number according to the order of acquisition time and sorted (in subsequent processes, the sequence number of the physical block represents the physical block itself). The sorted result is recorded as the physical block sequence and represented as: Q1, Q2, ..., Qj, where j is the total number of physical blocks, Qi represents the i-th acquired physical block, and i is the counting index, with a value range from 1 to j.

[0027] Then, obtain the monitoring period T_jc and update period T_gx preset by the operator, and take 0:00 on any day as the start time of an update period T_gx, which is also the start time of the first monitoring period T_jc within this update period T_gx.

[0028] Based on the above, any one physical block Qi from the physical block sequence Q1, Q2, ..., Qj is processed as an example, and the remaining physical blocks are processed in the same way as physical block Qi.

[0029] Within the first determined monitoring period T_jc, the access frequency and last access time interval associated with physical block Qi at different times are monitored and obtained in real time, and are denoted as A1_Qi and B1_Qi respectively. As described in Example 1, both the access frequency A1_Qi and the last access time interval B1_Qi are in the form of data streams (and the data in the access frequency A1_Qi and the last access time interval B1_Qi are arranged in chronological order), and the length of the data stream is the total number of moments within the monitoring period T_jc.

[0030] The access frequency A1_Qi and the last access time interval B1_Qi of physical block Qi in the first monitoring period T_jc are then combined. The combined result is recorded as the monitoring data group associated with physical block Qi, and it is denoted as FW1_Qi. In FW1, FW represents the monitoring data group, and 1 represents the first monitoring period. Therefore, FW1 represents the monitoring data group of the first monitoring period, and FW1_Qi represents the monitoring data group of physical block Qi in the first monitoring period.

[0031] By repeating the above steps, we can determine the monitoring data groups associated with each physical block in the physical block sequence Q1, Q2, ..., Qj within the first monitoring period T_jc, for a total of j monitoring data groups. Then, we sort the j monitoring data groups according to the order of the physical block sequence Q1, Q2, ..., Qj. The sorted result is denoted as the first monitoring data group sequence (first indicates the first monitoring period), and is represented as: FW1={FW1_Q1,FW1_Q2,...,FW1_Qj}.

[0032] By repeating the above steps, we can determine the monitoring data group sequence associated with each monitoring period T_jc within the update period T_gx, ultimately obtaining m monitoring data group sequences. Arranging the m monitoring data group sequences in chronological order, we get the following: FW1, FW2, ..., FWm, where FW1 is the first monitoring data group sequence, FW2 is the second monitoring data group sequence, FW3 is the third monitoring data group sequence, and so on.

[0033] Example 3 This embodiment discloses a method for evaluating high-thermal-physical blocks based on Embodiment 2, such as... Figure 3 As shown, it specifically includes the following: As described in Example 2, the monitoring data set sequences associated with each monitoring period T_jc within an update period T_gx can be obtained. It should be noted that the monitoring data set sequences are not obtained synchronously, but sequentially over time. Furthermore, after the end of any monitoring period T_jc, the monitoring data set sequences within that monitoring period T_jc are further processed to ensure the real-time performance of the data. The specific processing method is as follows: The first monitoring data group sequence associated with the first monitoring period T_jc within the update period T_gx is obtained for example processing. The monitoring data group sequences associated with other monitoring periods T_jc are processed in the same way. Extract any physical block Qi from the first monitoring data group sequence FW1={FW1_Q1,FW1_Q2,...,FW1_Qj} and its associated monitoring data group FW1_Qi (that is, within the first monitoring period T_jc).

[0034] Next, the monitoring data group FW1_Qi is decomposed to obtain the access frequency A1_Qi and the last access time interval B1_Qi associated with physical block Qi in the first monitoring period; A two-dimensional coordinate system is constructed with the timeline as the horizontal axis (covering the monitoring period T_jc) and the monitoring data as the vertical axis (including access frequency and last access time interval). Although the access frequency and last access time interval are displayed on the same vertical axis, they do not affect each other. In other words, the vertical axis can display two scales and two parameters at the same time, and the specific values ​​of the scales are determined by the operator based on the actual situation.

[0035] Next, the access frequency A1_Qi is obtained, and the data of each moment in the access frequency A1_Qi is obtained and plotted in the constructed two-dimensional coordinate system in the form of data points. Then, the curve is fitted to all the plotted data points in chronological order to obtain a curve, and this curve is recorded as the access frequency change curve SA1_Qi associated with the access frequency A1_Qi. Repeat the above steps to process the last access time interval B1_Qi in the same way, and you can obtain the last access time interval change curve SB1_Qi associated with the last access time interval B1_Qi.

[0036] It should be noted that, due to the operator's preset vertical axis scale, the access frequency change curve SA1_Qi and the last access time change curve SB1_Qi do not intersect in the same two-dimensional coordinate system, and the access frequency change curve SA1_Qi is always above the last access time change curve SB1_Qi.

[0037] Next, determine the start time and end time of the first monitoring period T_jc on the horizontal axis of the two-dimensional coordinate system, and construct a straight line perpendicular to the horizontal axis and parallel to the vertical axis through the start time and end time respectively, and denot it as L1 and L2 respectively, to obtain straight line L1 and straight line L2. Both straight line L1 and straight line L2 pass through the access frequency change curve SA1_Qi and the last access time change curve SB1_Qi.

[0038] At this point, the straight line L1, the straight line L2, the access frequency change curve SA1_Qi, and the last access time change curve SB1_Qi will form a closed region. The area value of this closed region is calculated, and the calculated area value is used as the high heat characteristic of physical block Qi in the first monitoring period T_jc, denoted as Z1_Qi.

[0039] It should be explained that the larger the value of the high-heat feature Z1_Qi, the larger the average distance between the access frequency change curve SA1_Qi and the last access time change curve SB1_Qi. Therefore, it can be determined that the data block Qi has a high access frequency and a short last access time, so the high-heat feature Z1_Qi will be larger, which means that the data block Qi is hotter.

[0040] Repeat the above steps to process all physical blocks in the physical block sequence Q1, Q2, ..., Qj in the same way. Determine the high-temperature features associated with each physical block in the first monitoring period T_jc, and sort the high-temperature features of each physical block according to the sorting order of the physical block sequence Q1, Q2, ..., Qj to obtain the first high-temperature feature sequence, represented as: Z1={Z1_Q1,Z1_Q2,...,Z1_Qj}.

[0041] Then, select the data blocks associated with the first α% of the high-temperature feature values ​​when sorting from largest to smallest from the first high-temperature feature sequence Z1, as the high-temperature physical blocks in the first monitoring period T_jc, where α% is the percentage preset by the operator; It is important to note that not all high-heat physical blocks will be stored in the target SSD's cache. If the data size of the first α% of data blocks exceeds the target SSD's cache size, the number of high-heat physical blocks will be reduced, starting with the physical block with the smallest high-heat characteristic value, until all high-heat physical blocks are large enough to be cached in the target SSD's cache (the cache needs to have sufficient reserve space, and the size of the reserve space is not less than the buffer storage amount preset by the operator). If the size of the first α% of data blocks does not exceed the cache size of the target SSD, and the free storage space is greater than the buffer storage amount preset by the operator, then continue to acquire physical blocks with larger values ​​of high-heat characteristics as high-heat data blocks.

[0042] Based on the above, the detailed steps for determining the first high-temperature feature sequence Z1 within the first monitoring period T_jc are as follows: Continue processing and updating the monitoring data group sequences associated with all monitoring periods T_jc within the update period T_gx, determine the high-temperature feature sequences associated with all monitoring periods T_jc, and finally obtain m high-temperature feature sequences associated with m monitoring periods T_jc. The m high-temperature feature sequences are represented in chronological order as: Z1, Z2, ..., Zm.

[0043] Next, based on the detailed steps for determining the high-temperature physical block within the first monitoring period T_jc as described above, determine the high-temperature physical blocks associated with all monitoring periods T_jc within the update period T_gx, that is, the high-temperature physical blocks associated with each high-temperature feature sequence in Z1, Z2, ..., Zm.

[0044] When any monitoring period T_jc within the update period T_gx ends, immediately obtain the high-temperature physical blocks associated with this monitoring period T_jc, determine the sequence number of all high-temperature physical blocks (Q1, Q2, etc.), and delete and update the high-temperature physical block sequence number originally stored in the edge computing node corresponding to this monitoring period (it is necessary to delete the high-temperature physical block sequence number originally stored in the edge computing node corresponding to this monitoring period).

[0045] This embodiment mainly focuses on real-time monitoring and processing of physical blocks stored in edge computing nodes. It uses high-heat characteristics to filter and update high-heat physical blocks in the cache, thereby optimizing the utilization efficiency of the storage cache. Specifically, by monitoring the access frequency and last access time interval of physical blocks, a two-dimensional coordinate system is constructed, a change curve is plotted, and the area of ​​the closed region enclosed by the curve is calculated as a high-heat characteristic. The high heat characteristic reflects the access activity of the physical block; the larger the characteristic value, the higher its heat. Then, through sorting and filtering mechanisms, physical blocks with larger high-temperature characteristic values ​​are selected as high-temperature physical blocks and stored in the cache. The storage content is adjusted according to the capacity of the cache and the parameters preset by the operator to ensure that the cache not only stores high-temperature data but also retains a certain amount of buffer space, thereby improving the performance and efficiency of the storage system.

[0046] Example 4 This embodiment, based on embodiment 3, further discloses a method for updating data in the cache area of ​​a target solid-state drive, specifically including the following: Once the high-temperature data blocks associated with all monitoring cycles T_jc within an update cycle T_gx are identified, the actual operation phase begins. First, obtain the actual time of the target solid-state drive. It is important to note that the actual time must be accurate to the second.

[0047] Then, the corresponding monitoring cycle identifier is determined based on the actual time (that is, the actual time is mapped to the corresponding hour of the day). The monitoring cycle identifier corresponds to the order of the monitoring cycle T_jc, thereby distinguishing different time periods. The value of the monitoring cycle identifier is K1, K2, ..., K24. The period from 0:00 to 1:00 in a day corresponds to the monitoring cycle identifier K1, the monitoring cycle identifier K1 corresponds to the first monitoring cycle T_jc, and so on.

[0048] In this way, there is no need to deal with complex continuous time calculations; a simple table lookup operation is all that is needed to determine the corresponding monitoring cycle identifier and the corresponding monitoring cycle.

[0049] Next, based on the determined monitoring cycle and in conjunction with the content described in Example 3, the high-temperature physical block associated with this monitoring cycle and the serial number of the high-temperature physical block are determined. Next, the data in the corresponding high-heat physical block can be extracted from the target SSD, and the data in the cache of the target SSD can be updated. In this way, through this mechanism, the target solid-state drive is transformed from a passive data storage device into an intelligent system that can actively optimize its own performance, making it suitable for application scenarios with regular and periodic workloads, such as enterprise OA systems, e-commerce websites, and video-on-demand platforms.

[0050] It should also be noted that the update frequency of high-heat data blocks is determined by the operator.

[0051] All data in the formulas described above are numerical calculations performed with dimensions removed. Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0052] The above description is merely an example and illustration of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0053] It should be stated that all user data collected in this application was collected with the user's consent and authorization. Furthermore, the uses of user data are legal and compliant, and the use and processing of user data comply with the relevant laws, regulations, and standards of the relevant regions.

Claims

1. A method for updating edge computing cache integrated in a solid-state drive, characterized in that, The method includes: Step 1: Utilize edge computing nodes that interact with the target SSD in real time to monitor the target SSD. Within a preset update period T_gx, determine the monitoring data groups associated with each physical block within the target SSD within a preset monitoring period T_jc. Specifically: Retrieve each physical block within the target solid-state drive, and assign a sequence number to each physical block based on the retrieval order. Then sort the physical blocks and denote them as the physical block sequence Q1, Q2, ..., Qj. j represents the total number of physical blocks, Qi represents the i-th physical block acquired, and i is the counting index, 1≤i≤j; Take 0:00 on any day as the start time of the update period T_gx and the first monitoring period T_jc within the update period T_gx; Within the first monitoring period T_jc, the access frequency and last access time interval associated with physical block Qi at different times are monitored and denoted as A1_Qi and B1_Qi, respectively. The access frequency A1_Qi and the last access time interval B1_Qi are both data streams, and the length of the data stream is the total number of times within the monitoring period T_jc. The combined access frequency A1_Qi and the last access time interval B1_Qi are used as the monitoring data group FW1_Qi for physical block Qi; Similarly, the monitoring data groups of each physical block in the physical block sequence within the first monitoring period T_jc are determined and sorted according to the order of the physical block sequence to obtain the first monitoring data group sequence FW1={FW1_Q1,FW1_Q2,...,FW1_Qj}; Repeat the above steps to determine the sequence of monitoring data groups in all monitoring periods T_jc within the update period T_gx, totaling m groups, which are represented in chronological order as: FW1, FW2, ..., FWm; Step two, based on the numerical performance of the monitoring data within the monitoring data group of each physical block in the corresponding monitoring period, evaluate the high-temperature physical blocks associated with the corresponding monitoring period, specifically as follows: Based on the first monitoring data group sequence, obtain the monitoring data group FW1_Qi of physical block Qi within the first monitoring period T_jc; Decompose the monitoring data set FW1_Qi to obtain the access frequency A1_Qi and the last access time interval B1_Qi; Construct a two-dimensional coordinate system with the timeline as the horizontal axis and the monitoring data as the vertical axis. Plot the access frequency A1_Qi and the last access time interval B1_Qi in chronological order on the constructed two-dimensional coordinate system to generate the access frequency change curve SA1_Qi associated with the access frequency A1_Qi and the last access time interval change curve SB1_Qi associated with the last access time interval B1_Qi. The access frequency change curve SA1_Qi does not intersect with the last access time change curve SB1_Qi, and the access frequency change curve SA1_Qi is located above the last access time change curve SB1_Qi. Based on the start and end times of the first monitoring period T_jc, construct straight lines L1 and L2 that are perpendicular to the horizontal axis and parallel to the vertical axis. Calculate the area of ​​the closed region formed by the straight line L1, the straight line L2, the access frequency change curve SA1_Qi, and the last access time change curve SB1_Qi. This area is used as the high-heat characteristic of physical block Qi in the first monitoring period T_jc and is denoted as Z1_Qi. Similarly, determine the high-temperature characteristics of each physical block in the physical block sequence Q1, Q2, ..., Qj within the first monitoring period T_jc, and sort them according to the order of the physical block sequence, denoted as the first high-temperature characteristic sequence Z1={Z1_Q1,Z1_Q2,...,Z1_Qj}; The values ​​of high-temperature features within the first high-temperature feature sequence Z1 are sorted from largest to smallest. The data blocks in the top α% are used as the high-temperature physical blocks within the first monitoring period T_jc, where α% is a percentage preset by the operator. Obtain the high-temperature data block sequence number and update the high-temperature physical block sequence number originally stored in the edge computing node for this monitoring period; Step 3: Based on the actual time of the target SSD, determine the monitoring period corresponding to this actual time, extract the corresponding high-heat physical block number from the edge computing node, and extract the data from the corresponding high-heat physical block from the target SSD and cache it in the cache area of ​​the target SSD.

2. The method according to claim 1, characterized in that, In step one, the monitoring data in the monitoring data group includes the access frequency and last access time interval associated with each physical block in the target solid-state drive. The access frequency includes read access frequency and write access frequency; The last access time interval is the time difference between the current time and the timestamp of the last time the physical block was accessed.

3. The method according to claim 2, characterized in that, In step one, the duration of the update cycle and the monitoring cycle are preset values ​​by the operator, denoted as T_gx and T_jc respectively; The update cycle T_gx spans from 0:00 to 24:00 of a day; The monitoring period T_jc spans one hour. Any update cycle T_gx contains m monitoring cycles T_jc, where m=24.

4. The method according to claim 3, characterized in that, In step two, based on the method of determining the first high-temperature feature sequence Z1 within the first monitoring period T_jc, the high-temperature feature sequences of each monitoring period T_jc within the update period T_gx are determined, totaling m, and are represented in chronological order as: Z1, Z2, ..., Zm.

5. The method according to claim 4, characterized in that, In step two, based on the method of determining the high-temperature physical blocks in the first monitoring period T_jc, the high-temperature physical blocks in each monitoring period T_jc in the update period T_gx are determined.

6. The method according to claim 5, characterized in that, In step two, when any monitoring period T_jc ends, the high-temperature physical block sequence number associated with this monitoring period T_jc is obtained, and the high-temperature physical block sequence number originally stored in the edge computing node corresponding to this monitoring period is deleted and updated.

7. The method according to claim 6, characterized in that, In step three, the specific method for extracting the data from the corresponding high-heat physical block of the target solid-state drive and caching it in the cache area of ​​the target solid-state drive is as follows: Obtain the actual time of the target solid-state drive, accurate to the second; The corresponding monitoring cycle identifier is determined based on the actual time, where the monitoring cycle identifier corresponds to the monitoring cycle T_jc; The monitoring cycle identifier takes the values ​​K1, K2, ..., K24, where 0:00 to 1:00 in a day corresponds to the monitoring cycle identifier K1, the monitoring cycle identifier K1 corresponds to the first monitoring cycle T_jc, and so on. Obtain the monitoring cycle corresponding to the actual time and its associated high-thermal physical block; Based on the sequence number of the high-temperature physical block, the data in the corresponding high-temperature physical block is extracted from the target solid-state drive, and the data in the cache area of ​​the target solid-state drive is deleted and updated.

Images