Real-time big data intelligent tuning database system based on deep learning

Abstract

The application discloses a real-time big data intelligent optimization database system based on deep learning, and relates to the technical field of databases.The application solves the problem that load changes cannot be dynamically adapted according to real-time monitoring data, and the problem that resource allocation is prone to excess or deficiency.The closed-loop mechanism of'monitoring-verification-evaluation-optimization-secondary verification' can quickly respond to real-time load changes, accurately allocate redundant IO resources to interfaces in short supply, and can also prioritize high-priority interfaces when resources are tight, effectively improving the overall throughput of the database, reducing query delay, and ensuring the stable operation of core business data interfaces, and is especially suitable for dynamic load regulation in real-time big data scenarios.

Description

Technical Field

[0001] This invention relates to the field of database technology, specifically to a real-time big data intelligent optimization database system based on deep learning. Background Technology

[0002] With the rapid development of big data technology and the popularization of real-time business scenarios (such as e-commerce real-time transactions, IoT data monitoring, financial risk control, etc.), the performance of databases, as the core carrier of data storage and processing, directly determines the response efficiency and stability of business systems. In scenarios with high concurrency and massive data throughput, the load faced by databases exhibits dynamic and complex characteristics. For example, a sudden increase in data volume during peak business periods, frequent switching of query types (from simple point queries to complex multi-table join queries), and resource competition between different data interfaces (such as IO, CPU, and memory) can all lead to fluctuations in database performance and even cause problems such as soaring query latency and decreased throughput.

[0003] Traditional database tuning methods largely rely on human experience, with DBAs manually adjusting parameters (such as cache size and connection count), optimizing indexes, or SQL statements to improve performance. However, this approach has significant limitations: on the one hand, manual tuning is slow to respond and struggles to handle real-time load changes; on the other hand, the tuning process is subjective and dependent on personal experience, making it difficult to consider multiple metrics (such as simultaneously balancing response time, resource utilization, and interface priority). Although some automated tuning tools have emerged, they often only optimize a single performance metric (such as IO throughput) or a fixed scenario, lacking a comprehensive understanding of the overall database operating status.

[0004] For example, some tools adjust resource allocation by monitoring CPU usage but ignore the differences in business priorities of different data interfaces, resulting in performance degradation of core interfaces due to resource contention. In addition, the resource scheduling strategies of existing tools are mostly based on preset rules and cannot dynamically adapt to load changes based on real-time monitoring data. This can easily lead to over- or under-allocation of resources, making it difficult to meet the fine-tuning needs of complex business scenarios.

[0005] Therefore, how to achieve real-time and comprehensive awareness of the database's operating status, intelligently determine the resource supply and demand relationship by combining multi-dimensional indicators, and dynamically allocate resources according to business priorities has become a key issue in improving the performance and stability of databases in real-time big data scenarios. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a real-time big data intelligent optimization database system based on deep learning, which solves the problem of being unable to dynamically adapt to load changes based on real-time monitoring data, and the tendency for resource allocation to be excessive or insufficient.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a real-time big data intelligent optimization database system based on deep learning, comprising: The data monitoring terminal monitors the parameters of multiple data interfaces associated with the database. The monitoring feature verification end compares and verifies the parameters monitored by the data interface with preset standard parameters, and confirms the verification ratio associated with the corresponding data interface based on the comparison and verification process. The specific method is as follows: The different parameters monitored by different data interfaces are denoted as CS. i-k Where i represents different data interfaces, k represents different parameter items, and the preset standard parameters are denoted as BZ. i-k ; Based on the total number of standard parameters, a corresponding number of polygons are generated. Starting from the center point of each polygon, the lines connecting the center point to the corner points of the polygon are identified. Different identified lines serve as measurement standards for different standard parameters, and the parameter values ​​recorded for line segments of different unit lengths within different lines are different. The generated polygons and the different parameters (CS) monitored by the corresponding data interface are then used as the basis for the calculation. i-k Generate the parameter polygon associated with the corresponding data interface, and denot the area of ​​the standard polygon as MB. i The area of ​​the parametric polygon is denoted as Mz. i Using: Mz i ÷MB i =ZB i Confirm the checksum ZB belonging to the current data interface. i ; If the corresponding data interface verification ratio ZB i If the value is less than 1, then the current data interface is recorded as the data interface to be adjusted. If the checksum of the corresponding data interface is ZB i If the value is greater than 1, then the current data interface is recorded as the data interface to be supplemented. If the corresponding data interface's verification ratio ZB i =1, no calibration is performed; The comprehensive evaluation end identifies the calibration status of different data interfaces, confirms the IO resources that can be scheduled or need to be scheduled, and evaluates whether the database can allocate IO resources on its own to complete the optimization process. If it can, the optimization center is executed; if not, the secondary verification end is executed. The specific method is as follows: Confirm the original IO resources ZY of the corresponding data interface i : If the corresponding data interface is the data interface to be adjusted, then use: ZY i ×(1-ZB i =DY i Confirm that the corresponding data interface can schedule the single resource DY. iAnd for different data interfaces to be adjusted, single resource DY i Confirm in sequence, then transfer the confirmed groups of single resources DY i Perform summation to confirm the total IO resources Dz that can be scheduled for the corresponding database; If the corresponding data interface is a data interface to be supplemented, then use: ZY i ×(ZB i -1) = DQ i Confirm that the corresponding data interface can be scheduled for a single resource DQ. i And for different data interfaces to be adjusted, single resource DQ i Confirm in sequence, and then process the confirmed groups of single-resource DQ. i Perform summation to confirm the total IO resources Db associated with the corresponding database that need to be scheduled; If Dz≥Db, it means that the database can allocate IO resources on its own to complete the tuning process, and the tuning center can be directly executed. The optimization center determines the total IO resources Dz that can be scheduled based on the data interface to be adjusted, and the single IO resources DQ that need to be scheduled based on the data interface to be supplemented. i The corresponding IO resources are allocated sequentially from Dz to the corresponding data interface to be supplemented, thus completing the database's self-tuning process and enabling the corresponding database to complete the tuning process on its own.

[0008] If Dz < Db, it means that the database can allocate IO resources on its own to complete the tuning process, and a secondary verification process needs to be performed. The secondary verification end performs feature verification on the response time associated with the confirmed data interfaces to be supplemented in the historical process, confirms the feature latency associated with different data interfaces to be supplemented, and locks down the resource ratio columns associated with several data interfaces to be supplemented based on the resource ratio columns to be supplemented, and confirms the IO resources to be supplemented for each data interface to be supplemented based on the resource ratio columns to be supplemented. The specific method is as follows: Using the current time as the base time, a set of tracing cycles is defined, and the tracing cycle is a preset cycle. The response time associated with different data interfaces to be supplemented within the tracing cycle is confirmed, and the time change curve of the response time of the data interface to be supplemented is generated according to the chronological relationship of time. Identify the maximum and minimum response times from the time variation curves associated with different data interfaces to be supplemented. Then, make the maximum and minimum response times change in opposite directions to identify several time ranges. Denote the curve segment included in each time range as the built-in range segment, and denot the length of the corresponding time range as F. q Let C be the length of the built-in range segment. q Where q represents different time ranges, using: Cq ÷F q =M q Confirm process characteristic M q And from the different process characteristics M associated with different time ranges q In the middle, select the maximum value, and record the time range associated with the maximum value as the standard range. Then, average the response times of several groups associated with the standard range to confirm that they belong to the characteristic delay associated with the corresponding data interface to be supplemented. Then, the feature delays associated with different data interfaces to be supplemented are processed into ratios to generate feature delay ratio columns belonging to several data interfaces to be supplemented. Then extract the checksum ZB associated with different interfaces for supplementing data. i , using: (ZB i -1) =DM i Confirm the undetermined ratio DM associated with the interface for supplementing data. i Confirm the correlation ratio of the corresponding data interface to be supplemented in the feature delay ratio column, and compare the correlation ratio with the undetermined ratio DM. i Perform product processing to confirm the comprehensive ratio, then sort the comprehensive ratios associated with several data interfaces to be supplemented to generate a column of resource ratios to be supplemented; Based on the total IO resources Dz confirmed by several data interfaces to be adjusted and the confirmed ratio of resources to be supplemented, the IO resources to be supplemented associated with each data interface to be supplemented are confirmed.

[0009] Preferably, the optimization center performs IO resource adjustment processing on each data interface to be supplemented based on the IO resources associated with each data interface to be supplemented.

[0010] Preferably, the parameters monitored in real time by the data interface include request volume, response time, and CPU usage.

[0011] This invention provides a real-time big data intelligent optimization database system based on deep learning. Compared with existing technologies, it has the following advantages: Firstly, it enables dynamic and precise scheduling of database IO resources. By monitoring parameters such as QPS, response time, and CPU usage of each data interface in real time through the data monitoring terminal, and combining the verification ratio calculated by the polygon area ratio using the monitoring feature verification terminal, it can comprehensively and objectively identify data interfaces to be adjusted (redundant IO resources) and data interfaces to be supplemented (insufficient IO resources), providing a precise basis for resource allocation and avoiding the blindness of resource allocation in traditional optimization.

[0012] Secondly, it improves the intelligence and automation of database optimization; the comprehensive evaluation end automatically determines whether the database can complete optimization on its own by calculating the matching relationship between the total schedulable resources and the total resources to be scheduled, reducing manual intervention; for scenarios that require secondary verification, the secondary verification end generates a column of resource ratios to be supplemented by combining historical response time characteristics and verification ratios, realizing differentiated resource allocation based on the priority of actual interface needs, making optimization more in line with the actual business operation status.

[0013] Third, it ensures the efficient and stable operation of the database under complex loads. Through a closed-loop mechanism of "monitoring-verification-evaluation-adjustment-secondary verification", the system can quickly respond to real-time load changes and accurately allocate redundant IO resources to scarce interfaces. It can also prioritize high-priority interfaces (such as interfaces with high latency and urgent needs) when resources are scarce. This effectively improves the overall throughput of the database, reduces query latency, and ensures the stable operation of core business data interfaces. It is especially suitable for dynamic load control in real-time big data scenarios. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the principle framework of the present invention. Detailed Implementation

[0015] 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.

[0016] First Embodiment Please see Figure 1 This application provides a real-time big data intelligent optimization database system based on deep learning, including a data monitoring end, a monitoring feature verification end, a comprehensive evaluation end, an optimization center, and a secondary verification end. The data monitoring end, the monitoring feature verification end, the comprehensive evaluation end, and the optimization center are electrically connected from the output node to the input node in sequence. The data monitoring end and the comprehensive evaluation end are both electrically connected to the input node of the secondary verification end, and the secondary verification end is electrically connected to the input node of the optimization center. The data monitoring end monitors the parameters of multiple data interfaces associated with the database, confirming the real-time request volume (QPS), response time (RT), and CPU usage of each data interface. The confirmed parameter data is then transmitted to the monitoring feature verification end. Specifically, different data interfaces monitor different parameters, and each set of data interfaces is associated with three sets of parameters. In the actual monitoring and processing process, each different parameter has an associated data standard. Based on the corresponding data standard, the supply and demand status of the corresponding data interface at the corresponding time can be confirmed. If the standard is exceeded, supply and demand adjustment is required, and excess IO resources are allocated. If the standard is not exceeded, excess IO resources can be extracted and allocated to the data interface that needs supply and demand adjustment. The monitoring feature verification end compares and verifies the parameters monitored by the data interface with preset standard parameters. Based on the comparison and verification process, it confirms the verification ratio associated with the corresponding data interface. The verification ratio is the overall feature ratio between the corresponding parameter and the standard parameter. This overall feature ratio effectively displays the specific numerical characteristics of the corresponding data interface, achieving better numerical representation. The specific method for confirming the verification ratio of the data interface is as follows: The different parameters monitored by different data interfaces are denoted as CS. i-k Where i represents different data interfaces, k represents different parameter items (corresponding to request volume QPS, response time RT, and CPU usage respectively), and the preset standard parameters are denoted as BZ. i-k ; Based on the total number of standard parameters, a corresponding number of polygons are generated. Starting from the center point of each polygon, the lines connecting the center point to the corner points of the polygon are identified. Different identified lines serve as measurement standards for different standard parameters, and the parameter values ​​recorded for line segments of different unit lengths within different lines are different. The generated polygons and the different parameters (CS) monitored by the corresponding data interface are then used as the basis for the calculation. i-k Generate the parameter polygon associated with the corresponding data interface, and denot the area of ​​the standard polygon as MB. i The area of ​​the parametric polygon is denoted as Mz. i Using: Mz i ÷MB i =ZB i Confirm the checksum ZB belonging to the current data interface. i ; If the corresponding data interface verification ratio ZB i If the value is less than 1, then the current data interface is recorded as the data interface to be adjusted. If the checksum of the corresponding data interface is ZB i If the value is greater than 1, then the current data interface is recorded as the data interface to be supplemented. If the corresponding data interface's verification ratio ZB i =1, no calibration is performed; Specifically, different data interfaces have their own verification standards. In order to effectively identify the actual operating status of the corresponding data interface, it is necessary to confirm the operating status of the corresponding data interface to assess whether there is a load during the data throughput process of such data interface, and take timely countermeasures so that the database can perform a good adjustment process in the future.

[0017] Among them, the comprehensive evaluation end identifies the calibration status of different data interfaces, confirms the IO resources that can be scheduled or the IO resources that need to be scheduled (here, the letters I and O), and evaluates whether the database can allocate IO resources on its own to complete the optimization process. If it can, the optimization center is executed; if it cannot, the secondary verification end is executed. The specific methods for conducting the evaluation are as follows: Confirm the existing IO resources of the corresponding data interface (which can be understood as the set disk utilization percentage, etc.) ZY i : If the corresponding data interface is the data interface to be adjusted, then use: ZY i ×(1-ZB i =DY i Confirm that the corresponding data interface can schedule the single resource DY. i And for different data interfaces to be adjusted, single resource DY i Confirm in sequence, then transfer the confirmed groups of single resources DY i Perform summation to confirm the total IO resources Dz that can be scheduled for the corresponding database; If the corresponding data interface is a data interface to be supplemented, then use: ZY i ×(ZB i -1) = DQ i Confirm that the corresponding data interface can be scheduled for a single resource DQ. i And for different data interfaces to be adjusted, single resource DQ i Confirm in sequence, and then process the confirmed groups of single-resource DQ. i Perform summation to confirm the total IO resources Db associated with the corresponding database that need to be scheduled; If Dz≥Db, it means that the database can allocate IO resources on its own to complete the tuning process, and the tuning center can be executed directly. If Dz<Db, it means that the database can allocate IO resources on its own to complete the tuning process, and the secondary verification terminal needs to be executed for secondary verification processing.

[0018] The optimization center determines the total IO resources (Dz) that can be scheduled based on the data interface to be adjusted, and the single IO resources (DQ) that need to be scheduled based on the data interface to be supplemented. iThe corresponding IO resources are allocated sequentially from Dz to the corresponding data interface to be supplemented, thus completing the database's self-tuning process and enabling the corresponding database to complete the tuning process on its own.

[0019] Second Embodiment In the specific implementation process, compared with the above embodiments, this embodiment mainly focuses on the secondary verification process of the corresponding database. It allocates computing power according to the priority of different data interfaces in the corresponding database to achieve the best comprehensive optimization effect. The secondary verification end performs feature verification on the response time associated with the confirmed data interfaces to be supplemented in the historical process, confirms the feature latency associated with different data interfaces to be supplemented, and locks down the resource ratio columns associated with several data interfaces to be supplemented based on the feature latency and the verification ratio. Based on the resource ratio columns, it confirms the IO resources to be supplemented for each data interface to be supplemented. The specific method for detailed confirmation is as follows: Using the current time as the base time, a set of traceability periods is defined. The traceability period is a preset period, and its specific value is determined by the operator based on experience. The response time associated with different data interfaces to be supplemented within the traceability period is confirmed, and the time change curve of the data interface to be supplemented with respect to the response time is generated according to the chronological relationship of time. Identify the maximum and minimum response times from the time variation curves associated with different data interfaces to be supplemented. Then, cause the maximum and minimum response times to change in opposite directions, identifying several time ranges (each time range is different in each variation process). Denote the curve segment included in each time range as the built-in range segment, and denote the length of the corresponding time range as F. q Let C be the length of the built-in range segment. q Where q represents different time ranges, using: C q ÷F q =M q Confirm process characteristic M q And from the different process characteristics M associated with different time ranges q In the middle, select the maximum value, and record the time range associated with the maximum value as the standard range. Then, average the response times of several groups associated with the standard range to confirm that they belong to the characteristic delay associated with the corresponding data interface to be supplemented. Then, the feature delays associated with different data interfaces to be supplemented are processed into ratios to generate feature delay ratio columns belonging to several data interfaces to be supplemented. Then extract the checksum ZB associated with different interfaces for supplementing data. i , using: (ZB i -1) =DM i Confirm the undetermined ratio DM associated with the interface for supplementing data.i Confirm the correlation ratio of the corresponding data interface to be supplemented in the feature delay ratio column, and compare the correlation ratio with the undetermined ratio DM. i Perform product processing to confirm the comprehensive ratio, then sort the comprehensive ratios associated with several data interfaces to be supplemented to generate a column of resource ratios to be supplemented; Based on the total IO resources Dz confirmed by several data interfaces to be adjusted and the confirmed ratio of resources to be supplemented, the IO resources to be supplemented associated with each data interface to be supplemented are confirmed. Specifically, the confirmed ratio of resources to be supplemented is proposed to be 1:2:3:4:5, where the associated DZ value is 300. Then, by dividing the total resource DZ equally from front to back, the specific IO resources to be supplemented associated with each ratio can be confirmed as 20, 40, 60, 80, and 100, respectively. Then, resource allocation is performed according to the IO resources to be supplemented associated with the corresponding data interface to be adjusted.

[0020] The optimization center adjusts the IO resources associated with each data interface to ensure that the database is in optimal operating condition and effectively guarantees the corresponding data throughput process.

[0021] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0022] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A real-time big data intelligent optimization database system based on deep learning, characterized in that, include: The data monitoring terminal monitors the parameters of multiple data interfaces associated with the database. The monitoring feature verification end compares and verifies the parameters monitored by the data interface with the preset standard parameters, and confirms the verification ratio associated with the corresponding data interface based on the comparison and verification process. The comprehensive evaluation end identifies the calibration status of different data interfaces, confirms the IO resources that can be scheduled or need to be scheduled, and evaluates whether the database can allocate IO resources on its own to complete the optimization process. If it can, the optimization center is executed; if it cannot, the secondary verification end is executed. The secondary verification end performs feature verification on the response time associated with the confirmed data interfaces to be supplemented in the historical process, confirms the feature latency associated with different data interfaces to be supplemented, and locks down the resource ratio columns associated with several data interfaces to be supplemented based on the resource ratio columns to be supplemented, and confirms the IO resources to be supplemented for each data interface to be supplemented based on the resource ratio columns to be supplemented. The specific method is as follows: Using the current time as the base time, a set of tracing cycles is defined, and the tracing cycle is a preset cycle. The response time associated with different data interfaces to be supplemented within the tracing cycle is confirmed, and the time change curve of the response time of the data interface to be supplemented is generated according to the chronological relationship of time. Identify the maximum and minimum response times from the time variation curves associated with different data interfaces to be supplemented. Then, make the maximum and minimum response times change in opposite directions to identify several time ranges. Denote the curve segment included in each time range as the built-in range segment, and denot the length of the corresponding time range as F. q Let C be the length of the built-in range segment. q Where q represents different time ranges, using: C q ÷F q =M q Confirm process characteristic M q And from the different process characteristics M associated with different time ranges q In the middle, select the maximum value, and record the time range associated with the maximum value as the standard range. Then, average the response times of several groups associated with the standard range to confirm that they belong to the characteristic delay associated with the corresponding data interface to be supplemented. Then, the feature delays associated with different data interfaces to be supplemented are processed into ratios to generate feature delay ratio columns belonging to several data interfaces to be supplemented. Then extract the checksum ZB associated with different interfaces for supplementing data. i , using: (ZB i -1) =DM i Confirm the undetermined ratio DM associated with the interface for supplementing data. i Confirm the correlation ratio of the corresponding data interface to be supplemented in the feature delay ratio column, and compare the correlation ratio with the undetermined ratio DM. i Perform product processing to confirm the comprehensive ratio, then sort the comprehensive ratios associated with several data interfaces to be supplemented to generate a column of resource ratios to be supplemented; Based on the total IO resources Dz confirmed by several data interfaces to be adjusted and the confirmed ratio of resources to be supplemented, the IO resources to be supplemented associated with each data interface to be supplemented are confirmed.

2. The real-time big data intelligent optimization database system based on deep learning according to claim 1, characterized in that, The parameters monitored in real time by the data interface include request volume, response time, and CPU usage.

3. The real-time big data intelligent optimization database system based on deep learning according to claim 1, characterized in that, The specific method by which the monitoring feature verification terminal confirms the verification ratio corresponding to the data interface is as follows: The different parameters monitored by different data interfaces are denoted as CS. i-k Where i represents different data interfaces, k represents different parameter items, and the preset standard parameters are denoted as BZ. i-k ; Based on the total number of standard parameters, a corresponding number of polygons are generated. Starting from the center point of each polygon, the lines connecting the center point to the corner points of the polygon are identified. Different identified lines serve as measurement standards for different standard parameters, and the parameter values ​​recorded for line segments of different unit lengths within different lines are different. The generated polygons and the different parameters (CS) monitored by the corresponding data interface are then used as the basis for the calculation. i-k Generate the parameter polygon associated with the corresponding data interface, and denot the area of ​​the standard polygon as MB. i The area of ​​the parametric polygon is denoted as Mz. i Using: Mz i ÷MB i =ZB i Confirm the checksum ZB belonging to the current data interface. i ; If the corresponding data interface verification ratio ZB i If the value is less than 1, then the current data interface is recorded as the data interface to be adjusted. If the checksum of the corresponding data interface is ZB i If the value is greater than 1, then the current data interface is recorded as the data interface to be supplemented. If the corresponding data interface's verification ratio ZB i =1, no calibration is performed.

4. The real-time big data intelligent optimization database system based on deep learning according to claim 1, characterized in that, The comprehensive evaluation system assesses whether the database can automatically allocate IO resources to complete the optimization process in the following way: Confirm the original IO resources ZY of the corresponding data interface i : If the corresponding data interface is the data interface to be adjusted, then use: ZY i ×(1-ZB i =DY i Confirm that the corresponding data interface can schedule the single resource DY. i And for different data interfaces to be adjusted, single resource DY i Confirm in sequence, then transfer the confirmed groups of single resources DY i Perform summation to confirm the total IO resources Dz that can be scheduled for the corresponding database; If the corresponding data interface is a data interface to be supplemented, then use: ZY i ×(ZB i -1) = DQ i Confirm that the corresponding data interface can be scheduled for a single resource DQ. i And for different data interfaces to be adjusted, single resource DQ i Confirm in sequence, and then process the confirmed groups of single-resource DQ. i Perform summation to confirm the total IO resources Db associated with the corresponding database that need to be scheduled; If Dz≥Db, it means that the database can allocate IO resources on its own to complete the tuning process, and the tuning center can be directly executed.

5. The real-time big data intelligent optimization database system based on deep learning according to claim 4, characterized in that, The optimization center determines the total IO resources Dz that can be scheduled based on the data interface to be adjusted, and the single IO resources DQ that need to be scheduled based on the data interface to be supplemented. i The corresponding IO resources are allocated sequentially from Dz to the corresponding data interface to be supplemented, thus completing the database's self-tuning process and enabling the corresponding database to complete the tuning process on its own.

6. The real-time big data intelligent optimization database system based on deep learning according to claim 4, characterized in that, If Dz < Db, it means that the database cannot allocate IO resources on its own to complete the tuning process, and a secondary verification process needs to be performed.

7. The real-time big data intelligent optimization database system based on deep learning according to claim 1, characterized in that, The optimization center performs IO resource adjustment processing on each data interface to be supplemented based on the IO resources associated with each data interface to be supplemented.

Images