CN114138469B - Thread self-adaptive distribution method and device for solid state disk and solid state disk - Google Patents

Abstract

The present application relates to a thread adaptive allocation method, device, computer equipment and solid state drive of a solid state drive. The method includes: obtaining input and output characteristic data within a preset time interval; determining the input and output characteristic vector range in the thread allocation parameter mapping table according to the input and output characteristic data, and the thread allocation parameter mapping table maps each input and output. The feature vector range corresponds to the thread allocation parameter with the best performance; according to the input and output feature vector range, the first thread allocation parameter is obtained from the thread allocation parameter mapping table; compare whether the current thread allocation parameter is consistent with the first thread allocation parameter; when they are inconsistent , adjust the current thread allocation parameters to the first thread allocation parameters, so that the solid-state drive can adaptively modify the thread allocation parameters according to the current application scenario, so that the thread allocation ratio of the solid-state drive in each application scenario is optimal and the performance is optimal. Improved the overall performance of solid-state drives.

Description

Thread adaptive allocation method, device and solid state drive of solid state drive

Technical field

The present application relates to the field of electronic technology, and in particular to a thread adaptive allocation method, device, computer equipment and solid state drive of a solid state drive.

Background technique

With the development of electronic technology, solid-state drives have emerged. Solid-state drives (Solid State Disk or SolidState Drive, referred to as SSD), also known as solid-state drives, are hard drives made of solid-state electronic storage chip arrays. In terms of interface specifications and definitions, The functions and usage methods are exactly the same as ordinary hard drives. The product appearance and size are basically the same as ordinary hard drives. Solid-state drives have the characteristics of fast reading and writing, light weight, low energy consumption and small size that traditional mechanical hard drives do not have. Therefore, it is widely used in military, vehicle, industrial control, video surveillance, network monitoring, network terminals, electric power, medical, aviation, navigation equipment and many other fields.

Users have increasingly higher performance requirements for solid-state drives, and there are many factors that affect SSD performance. In particular, different business scenarios have a great impact on performance. At present, the main focus is on optimizing the firmware of solid-state drives for specific business scenarios, so that the solid-state drives can achieve optimal performance in this business scenario.

However, when SSDs are applied to other business scenarios, they often fail to meet the performance requirements of other business scenarios, resulting in low overall performance of SSDs.

Contents of the invention

Based on this, it is necessary to address the above technical problems and provide a thread adaptive allocation method, device, computer equipment and solid state drive for a solid state drive that can improve the overall performance of the solid state drive.

On the one hand, a thread adaptive allocation method for solid state drives is provided, and the method includes:

Obtain input and output feature data within a preset time interval;

According to the input and output feature data, determine the input and output feature vector ranges belonging to the thread allocation parameter mapping table, and the thread allocation parameter mapping table maps the thread allocation parameters with the best performance corresponding to each input and output feature vector range;

Obtain the first thread allocation parameter from the thread allocation parameter mapping table according to the input and output feature vector range;

Compare whether the current thread allocation parameter is consistent with the first thread allocation parameter;

When inconsistent, the current thread allocation parameters are adjusted to the first thread allocation parameters.

In one embodiment, the step of determining the input-output feature vector range belonging to the thread allocation parameter mapping table based on the input-output feature data includes:

Perform denoising integration on each input and output feature vector in the input and output feature data to obtain the average value of the input and output feature vectors;

According to the average value of the input and output feature vectors, the corresponding input and output feature vector ranges are matched in the thread allocation parameter mapping table.

In one embodiment, the method further includes: when receiving an input/output request, recording the input/output feature vector of the input/output request.

In one embodiment, the method for determining the mapping relationship of the thread allocation parameter mapping table includes:

From the pre-selected thread allocation parameter set, select the pre-selected thread allocation parameters as the target thread allocation parameters;

Adjust the thread allocation parameters in the first solid-state drive according to the target thread allocation parameters;

Simulate different application scenarios to perform performance testing on the input and output requests issued by the first solid-state drive, and obtain the performance results of the first solid-state drive in each application scenario;

Collect input and output characteristic parameters corresponding to the first solid-state drive in each application scenario;

Return to the step of sequentially selecting target thread allocation parameters from the pre-selected thread allocation parameter set until all pre-selected thread allocation parameters in the pre-selected thread allocation parameter set are selected, and obtain performance result data, where the performance result data includes each pre-selected thread allocation parameter When assigning parameters as a target thread, the performance results of the first solid-state drive in various application scenarios;

Analyze the performance result data to determine the optimal thread allocation parameters corresponding to each application scenario;

According to the input and output feature parameters collected in each application scenario, determine the corresponding input and output feature vector range in each application scenario;

According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

In one embodiment, the step of determining the mapping relationship of the thread allocation parameter mapping table based on the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario includes:

According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, a preliminary mapping table is constructed for performance testing and the testing results are obtained;

When the test results meet the performance requirements, the primary thread allocation parameters with the best performance corresponding to each application scenario are determined as the thread allocation parameters with the best performance corresponding to each application scenario;

The mapping relationship of the thread allocation parameter mapping table is determined based on the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario.

In one embodiment, the steps of allocating parameters and input and output feature vector ranges to the primary threads with the best performance corresponding to each application scenario, constructing a preliminary mapping table for verification, and obtaining verification results include:

Select application scenarios from each application scenario in sequence as the target application scenario;

Perform a performance test on the input and output requests of the target application scenario issued by the second solid-state drive in which the preliminary mapping table is written, and obtain the performance results of the second solid-state drive in the target application scenario. The preliminary mapping table It is established based on the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario;

Return to the step of selecting application scenarios as target application scenarios in sequence from each application scenario until all application scenarios are used as target application scenarios, and obtain the performance results of the second solid-state drive in each application scenario;

When the performance results under each application scenario reach the preset threshold, it is determined that the inspection results meet the performance requirements;

When the performance results under each application scenario do not reach the preset threshold, it is determined that the inspection results do not meet the performance requirements.

In one embodiment, the thread allocation parameters include: the number of threads for processing input and output requests issued by the GC side and the number of threads for processing the input and output requests issued by the host side.

On the other hand, a thread adaptive allocation device for a solid state drive is provided, and the device includes:

The data acquisition module is used to acquire input and output characteristic data within a preset time interval;

A range determination module, configured to determine the input and output feature vector ranges belonging to the thread allocation parameter mapping table according to the input and output feature data. The thread allocation parameter mapping table maps the input and output feature vector ranges corresponding to the optimal performance Thread allocation parameters;

An allocation parameter acquisition module, configured to acquire the first thread allocation parameter from the thread allocation parameter mapping table according to the input and output feature vector range;

A comparison module, used to compare whether the current thread allocation parameter is consistent with the first thread allocation parameter;

A parameter adjustment module, configured to adjust the current thread allocation parameters to the first thread allocation parameters when they are inconsistent.

On the other hand, a computer device is provided, including a memory, a processor, and a solid-state drive. When the firmware of the solid-state drive is running, the following steps are implemented:

The data acquisition module is used to acquire input and output characteristic data within a preset time interval;

A range determination module, configured to determine the input and output feature vector ranges belonging to the thread allocation parameter mapping table according to the input and output feature data. The thread allocation parameter mapping table maps the input and output feature vector ranges corresponding to the optimal performance Thread allocation parameters;

An allocation parameter acquisition module, configured to acquire the first thread allocation parameter from the thread allocation parameter mapping table according to the input and output feature vector range;

A comparison module, used to compare whether the current thread allocation parameter is consistent with the first thread allocation parameter;

A parameter adjustment module, configured to adjust the current thread allocation parameters to the first thread allocation parameters when they are inconsistent.

On the other hand, a solid state drive is provided, a computer program is stored thereon, and when the computer program is executed by a processor, the following steps are implemented:

The data acquisition module is used to acquire input and output characteristic data within a preset time interval;

A range determination module, configured to determine the input and output feature vector ranges belonging to the thread allocation parameter mapping table according to the input and output feature data. The thread allocation parameter mapping table maps the input and output feature vector ranges corresponding to the optimal performance Thread allocation parameters;

An allocation parameter acquisition module, configured to acquire the first thread allocation parameter from the thread allocation parameter mapping table according to the input and output feature vector range;

A comparison module, used to compare whether the current thread allocation parameter is consistent with the first thread allocation parameter;

A parameter adjustment module, configured to adjust the current thread allocation parameters to the first thread allocation parameters when they are inconsistent.

The above-mentioned solid-state drive method, device, computer equipment and solid-state drive can reflect the scenarios in which the solid-state drive is mainly used by obtaining input and output characteristic data within a preset time interval; based on the input and output characteristic data, determine the thread allocation parameter mapping The input and output feature vector ranges in the table, the thread allocation parameter mapping table maps the optimal thread allocation parameters corresponding to each input and output feature vector range; according to the input and output feature vector range, the first thread is obtained from the thread allocation parameter mapping table Allocation parameters; compare whether the current thread allocation parameters are consistent with the first thread allocation parameters; when they are inconsistent, adjust the current thread allocation parameters to the first thread allocation parameters, so that the solid state drive can adaptively modify the thread allocation parameters according to the current application scenario. This optimizes the thread allocation ratio of the solid-state drive in various application scenarios and achieves optimal performance, improving the overall performance of the solid-state drive.

Description of the drawings

Figure 1 is a schematic flowchart of a thread adaptive allocation method for a solid state drive in an embodiment;

Figure 2 is a structural block diagram of a thread adaptive allocation device for a solid state drive in one embodiment;

Figure 3 is an internal structure diagram of a computer device in one embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application and are not used to limit the present application.

The thread adaptive allocation method of solid-state drives provided by this application can be applied to solid-state drives. Among them, the solid-state drive can be connected to the terminal for use. During use, the firmware in the solid-state drive obtains the input and output characteristic data within the preset time interval during operation; according to the input and output characteristic data, the input belonging to the thread allocation parameter mapping table is determined. Output feature vector range, thread allocation parameter mapping table maps the optimal thread allocation parameters corresponding to each input and output feature vector range; according to the input and output feature vector range, obtain the first thread allocation parameter from the thread allocation parameter mapping table; compare the current Whether the thread allocation parameters are consistent with the first thread allocation parameters; if they are inconsistent, adjust the current thread allocation parameters to the first thread allocation parameters. The terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, portable wearable devices, etc.

In one embodiment, as shown in Figure 1, a method for thread adaptive allocation of solid-state drives is provided. The application of this method to solid-state drives is used as an example to illustrate, including the following steps:

Step 202: Obtain input and output characteristic data within a preset time interval.

Among them, the preset time interval can be determined according to the user's frequency of use of the solid state drive, such as one hour, two hours, or one day, etc. Input and output feature data are all input and output feature vectors collected within a preset time interval. The input and output feature vectors include the randomness of input and output requests, block size, queue depth, etc. Input and output are also called I/O, the full name It is Input/Output (i.e. input/output).

Step 204: Determine the input and output feature vector ranges in the thread allocation parameter mapping table based on the input and output feature data. The thread allocation parameter mapping table maps the thread allocation parameters with the best performance corresponding to each input and output feature vector range.

Among them, each input and output feature vector range represents an application scenario. Application scenarios can include: database service, mailbox service, web service, streaming media service, etc. The thread allocation parameter mapping table is stored in the solid state drive. Thread allocation parameters include the number of threads for processing input and output requests issued by the GC and the number of threads for processing input and output requests issued by the host, etc. GC (Garbage Collection) is a garbage collection mechanism. The host is the main control chip in the solid state drive. The input and output feature vectors with the highest frequency of occurrence can be analyzed from the input and output feature data as the representative value, and the range of input and output feature vectors in the thread allocation parameter mapping table can be analyzed, or the input and output feature vector can be analyzed from the input and output feature data. The average value of the input and output feature vectors is used as the representative value. The range of the input and output feature vectors in the thread allocation parameter mapping table can also be denoised in other ways to produce the input and output feature vectors that can represent the input and output feature data as the basis for judgment. The mapping relationship of the thread allocation parameter mapping table can be automatically obtained by executing the test program, or can be obtained by manual testing.

Step 206: Obtain the first thread allocation parameter from the thread allocation parameter mapping table according to the input and output feature vector range.

The first thread allocation parameter is a thread allocation parameter that can optimize the performance of the solid state drive when the solid state drive is in an application scenario corresponding to the input and output feature vector range. For example: if the application scenario is a streaming media service, which requires high bandwidth and relatively little garbage data, reduce the number of input and output request threads issued by the GC to reduce the resource occupation of the GC; if the application scenario is a database service, the latency and IOPS The requirements are high and there are a lot of random writes, so there will be a lot of garbage data, and the pressure on the GC will be great, so a relatively large number of threads need to be allocated to the GC. IOPS (Input/Output Operations PerSecond) refers to the number of read and write operations per second.

Step 208: Compare whether the current thread allocation parameter is consistent with the first thread allocation parameter.

Among them, the current thread allocation parameters in the firmware of the solid state disk are obtained, compared with the first thread allocation parameters, and it is determined whether the current thread allocation parameters are consistent with the first thread allocation parameters. For example: the current thread allocation parameters are 4 threads for processing input and output requests issued by the GC and 8 threads for processing input and output requests issued by the host. The first thread allocation parameter is also 8 threads for processing input and output requests issued by the GC. 4 and the number of threads processing input and output requests issued by the host is 8, then the current thread allocation parameters are consistent with the first thread allocation parameters; the current thread allocation parameters are the number of threads processing input and output requests issued by the GC is 4 and the number of threads processing the host The number of input and output request threads issued is 8. The first thread allocation parameter is to process the input and output request issued by the GC. The number of threads is 8 and the number of input and output request threads issued by the host is 4. Then the current thread allocation parameter is the same as the first thread allocation parameter. Thread allocation parameters are inconsistent.

Step 300: When inconsistent, adjust the current thread allocation parameters to the first thread allocation parameters.

In the thread adaptive allocation method of the solid-state drive mentioned above, by obtaining the input and output characteristic data within the preset time interval, it can be reflected in what scenarios the solid-state drive is mainly used; according to the input and output characteristic data, it is determined in the thread allocation parameter mapping table According to the input and output feature vector range, the thread allocation parameter mapping table maps the optimal thread allocation parameter corresponding to each input and output feature vector range; according to the input and output feature vector range, the first thread allocation parameter is obtained from the thread allocation parameter mapping table ; Compare whether the current thread allocation parameters are consistent with the first thread allocation parameters; when they are inconsistent, adjust the current thread allocation parameters to the first thread allocation parameters, so that the solid state drive can adaptively modify the thread allocation parameters according to the current application scenario, so that each In application scenarios, the thread allocation ratio of the solid-state drive is optimal and the performance is optimal, improving the overall performance of the solid-state drive.

In one embodiment, the step of determining the range of input and output feature vectors in the thread allocation parameter mapping table according to the input and output feature data includes: denoising and integrating each input and output feature vector in the input and output feature data to obtain The average value of the input and output feature vectors; according to the average value of the input and output feature vectors, match the corresponding input and output feature vector range in the thread allocation parameter mapping table.

Among them, the average value of the input and output feature vectors is used as the representative value, which can more accurately reflect the characteristics of the application scenario used by the solid-state drive, and further more accurately corresponds to the acquisition of the first thread allocation parameters, so that the performance of the solid-state drive can be maximized. excellent.

In one embodiment, the step of denoising and integrating each input and output feature vector in the input and output feature data to obtain the average value of the input and output feature vectors includes: removing extreme input and output feature vectors from the input and output feature data, Obtain the eliminated input and output feature data; perform average analysis based on the eliminated input and output feature data to obtain the average value of the input and output feature vectors.

Among them, eliminating the extreme input and output feature vectors can avoid the low-probability input and output feature vectors from affecting the average analysis, so that the average value of the input and output feature vectors can more accurately reflect the characteristics of the application scenarios used by the solid-state drive, making it more accurate Corresponding to obtain the first thread allocation parameters, so that the performance of the solid state drive can be optimized.

In one embodiment, the thread adaptive allocation method of the solid state drive further includes: when an input/output request is received, recording the input/output feature vector of the input/output request.

Among them, the input and output requests can be input and output requests issued by the GC or input and output requests issued by the host.

In one embodiment, taking the execution of a test program to obtain the mapping relationship of the thread allocation parameter mapping table as an example, the method for determining the mapping relationship of the thread allocation parameter mapping table includes the following steps:

From the pre-selected thread allocation parameter set, select the pre-selected thread allocation parameters as the target thread allocation parameters; adjust the thread allocation parameters in the first solid-state drive according to the target thread allocation parameters; simulate various application scenarios to deliver input and output to the first solid-state drive Request a performance test to obtain the performance results of the first solid-state drive in each application scenario; collect the corresponding input and output characteristic parameters of the first solid-state drive in each application scenario; return the target thread allocation parameters selected from the pre-selected thread allocation parameter set. Steps until all pre-selected thread allocation parameters in the pre-selected thread allocation parameter set are selected to obtain performance result data. The performance result data includes the performance of the first solid-state drive in each application scenario when each pre-selected thread allocation parameter is used as the target thread allocation parameter. Results: Analyze the performance result data to determine the optimal thread allocation parameters corresponding to each application scenario; determine the input and output feature vector range corresponding to each application scenario based on the input and output feature parameters collected in each application scenario; According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

Among them, simulating different application scenarios can build real scenarios, and also use third-party simulation tools, such as fio (supports multiple input and output engine workload generators), vdbench (input and output workload generator), etc. Performance results include bandwidth, iops (ie, input and output per second), latency and other data. Each input and output feature vector range corresponds to an application scenario. The determined primary thread allocation parameters corresponding to the optimal performance for each application scenario can be directly used as the thread allocation parameters corresponding to the optimal performance for each input and output feature vector range, and the mapping relationship of the determined thread allocation parameter mapping table. It is also possible to verify the primary thread allocation parameters with the best performance corresponding to each application scenario, and then determine the mapping relationship of the thread allocation parameter mapping table after passing the verification.

In one embodiment, the step of determining the mapping relationship of the thread allocation parameter mapping table according to the primary thread allocation parameters and the input and output feature vector ranges corresponding to the optimal performance of each application scenario includes: according to the optimal performance corresponding to each application scenario Based on the primary thread allocation parameters and input and output feature vector ranges, a preliminary mapping table is constructed for performance testing, and the inspection results are obtained; when the inspection results meet the performance requirements, the primary thread allocation parameters with the best performance corresponding to each application scenario are determined as each The application scenario corresponds to the thread allocation parameter with the best performance; according to the thread allocation parameter with the best performance and the input and output feature vector range corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

Among them, after verifying the primary thread allocation parameters with the best performance for each application scenario, and then determining the mapping relationship of the thread allocation parameter mapping table after passing the verification, it can improve the more accurate correspondence to adjust the thread allocation parameters, so that the solid-state drive performance reaches its optimum.

In one embodiment, the step of constructing a preliminary mapping table for verification based on the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, and obtaining the verification results includes: selecting applications from each application scenario in turn The scenario is used as the target application scenario; perform a performance test on the input and output requests of the target application scenario issued by the second solid-state drive written in the preliminary mapping table, and obtain the performance results of the second solid-state drive in the target application scenario. The preliminary mapping table is based on Each application scenario corresponds to the optimal primary thread allocation parameters and input and output feature vector ranges; return to the steps of selecting application scenarios as target application scenarios from each application scenario until all application scenarios are used as target application scenarios. Obtain the performance results of the second solid-state drive in each application scenario; when the performance results in each application scenario reach the preset threshold, determine that the inspection results meet the performance requirements; when the performance results in each application scenario do not reach the preset threshold, It is determined that the inspection results do not meet the performance requirements.

Among them, after verifying the primary thread allocation parameters with the best performance for each application scenario, and then determining the mapping relationship of the thread allocation parameter mapping table after passing the verification, it can improve the more accurate correspondence to adjust the thread allocation parameters, so that the solid-state drive performance reaches its optimum.

It should be understood that although various steps in the flowchart of FIG. 1 are shown in sequence as indicated by arrows, these steps are not necessarily executed in the order indicated by arrows. Unless explicitly stated in this article, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in Figure 1 may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but may be executed at different times. The execution of these sub-steps or stages The sequence is not necessarily sequential, but may be performed in turn or alternately with other steps or sub-steps of other steps or at least part of the stages.

In one embodiment, as shown in Figure 2, a thread adaptive allocation device for a solid state drive is provided, including: a data acquisition module, a range determination module, an allocation parameter acquisition module, a comparison module and a parameter adjustment module, wherein:

The data acquisition module is used to obtain the input and output feature data within the preset time interval; the range determination module is used to determine the input and output feature vector range belonging to the thread allocation parameter mapping table based on the input and output feature data, and the thread allocation parameter mapping The table maps the optimal thread allocation parameters corresponding to each input and output feature vector range; the allocation parameter acquisition module is used to obtain the first thread allocation parameter from the thread allocation parameter mapping table according to the input and output feature vector range; the comparison module is used To compare whether the current thread allocation parameters are consistent with the first thread allocation parameters; the parameter adjustment module is used to adjust the current thread allocation parameters to the first thread allocation parameters when they are inconsistent.

Among them, the thread allocation parameters include: the number of threads for processing input and output requests issued by the GC side and the number of threads for processing input and output requests issued by the host side.

In one embodiment, the range determination module is also used to: denoise and integrate each input and output feature vector in the input and output feature data to obtain the average value of the input and output feature vectors; based on the average value of the input and output feature vectors, in the thread Assign the range of input and output feature vectors to which matches in the parameter mapping table belong.

In one embodiment, the thread adaptive allocation device of the solid state drive further includes a recording module configured to record the input and output feature vectors of the input and output requests when an input and output request is received.

In one embodiment, the method for determining the mapping relationship of the thread allocation parameter mapping table includes:

From the pre-selected thread allocation parameter set, select the pre-selected thread allocation parameters as the target thread allocation parameters; adjust the thread allocation parameters in the first solid-state drive according to the target thread allocation parameters; simulate various application scenarios to deliver input and output to the first solid-state drive Request a performance test to obtain the performance results of the first solid-state drive in each application scenario; collect the corresponding input and output characteristic parameters of the first solid-state drive in each application scenario; return the target thread allocation parameters selected from the pre-selected thread allocation parameter set. Steps until all pre-selected thread allocation parameters in the pre-selected thread allocation parameter set are selected to obtain performance result data. The performance result data includes the performance of the first solid-state drive in each application scenario when each pre-selected thread allocation parameter is used as the target thread allocation parameter. Results: Analyze the performance result data to determine the optimal thread allocation parameters corresponding to each application scenario; determine the input and output feature vector range corresponding to each application scenario based on the input and output feature parameters collected in each application scenario; According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

In one embodiment, the step of determining the mapping relationship of the thread allocation parameter mapping table according to the primary thread allocation parameters and the input and output feature vector ranges corresponding to the optimal performance of each application scenario includes: according to the optimal performance corresponding to each application scenario Based on the primary thread allocation parameters and input and output feature vector ranges, a preliminary mapping table is constructed for performance testing, and the inspection results are obtained; when the inspection results meet the performance requirements, the primary thread allocation parameters with the best performance corresponding to each application scenario are determined as each The application scenario corresponds to the thread allocation parameter with the best performance; according to the thread allocation parameter with the best performance and the input and output feature vector range corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

In one embodiment, the step of constructing a preliminary mapping table for verification based on the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, and obtaining the verification results includes: selecting applications from each application scenario in turn The scenario is used as the target application scenario; perform a performance test on the input and output requests of the target application scenario issued by the second solid-state drive written in the preliminary mapping table, and obtain the performance results of the second solid-state drive in the target application scenario. The preliminary mapping table is based on Each application scenario corresponds to the optimal primary thread allocation parameters and input and output feature vector ranges; return to the steps of selecting application scenarios as target application scenarios from each application scenario until all application scenarios are used as target application scenarios. Obtain the performance results of the second solid-state drive in each application scenario; when the performance results in each application scenario reach the preset threshold, determine that the inspection results meet the performance requirements; when the performance results in each application scenario do not reach the preset threshold, It is determined that the inspection results do not meet the performance requirements.

For specific limitations on the thread adaptive allocation device of the solid state drive, please refer to the limitations on the thread adaptive allocation method of the solid state drive mentioned above, which will not be described again here. Each module in the above-mentioned thread adaptive allocation device of the solid state drive can be implemented in whole or in part by software, hardware, and combinations thereof. Each of the above modules may be embedded in or independent of the processor of the computer device in the form of hardware, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. 3 . The computer device includes a processor, memory, a network interface, a display screen, a solid state drive, and an input device connected via a system bus. Wherein, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile memory and internal memory. This non-volatile memory stores the operating system and computer programs. This internal memory provides an environment for the execution of operating systems and computer programs in non-volatile memory. The network interface of the computer device is used to communicate with external terminals through a network connection. The firmware of the solid state drive implements a thread adaptive allocation method of the solid state drive when running. The display screen of the computer device may be a liquid crystal display or an electronic ink display. The input device of the computer device may be a touch layer covered on the display screen, or may be a button, trackball or touch pad provided on the computer device shell. , it can also be an external keyboard, trackpad or mouse, etc.

Those skilled in the art can understand that the structure shown in Figure 3 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. The specific computer equipment can May include more or fewer parts than shown, or combine certain parts, or have a different arrangement of parts.

In one embodiment, a solid state drive is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented:

Obtain the input and output feature data within the preset time interval; according to the input and output feature data, determine the input and output feature vector ranges in the thread allocation parameter mapping table. The thread allocation parameter mapping table maps the corresponding performance of each input and output feature vector range. Optimal thread allocation parameters; according to the input and output feature vector range, obtain the first thread allocation parameters from the thread allocation parameter mapping table; compare whether the current thread allocation parameters are consistent with the first thread allocation parameters; when they are inconsistent, change the current thread allocation parameters Adjust the allocation parameters for the first thread.

In one embodiment, the computer program, when executed by the processor, also implements the following steps:

Denoise and integrate each input and output feature vector in the input and output feature data to obtain the average value of the input and output feature vectors; match the corresponding input and output feature vector range in the thread allocation parameter mapping table based on the average value of the input and output feature vectors .

In one embodiment, when executed by the processor, the computer program further implements the following steps: when an input-output request is received, recording the input-output feature vector of the input-output request.

Among them, the mapping relationship determination method of the thread allocation parameter mapping table includes:

From the pre-selected thread allocation parameter set, select the pre-selected thread allocation parameters as the target thread allocation parameters; adjust the thread allocation parameters in the first solid-state drive according to the target thread allocation parameters; simulate various application scenarios to deliver input and output to the first solid-state drive Request a performance test to obtain the performance results of the first solid-state drive in each application scenario; collect the corresponding input and output characteristic parameters of the first solid-state drive in each application scenario; return the target thread allocation parameters selected from the pre-selected thread allocation parameter set. Steps until all pre-selected thread allocation parameters in the pre-selected thread allocation parameter set are selected to obtain performance result data. The performance result data includes the performance of the first solid-state drive in each application scenario when each pre-selected thread allocation parameter is used as the target thread allocation parameter. Results: Analyze the performance result data to determine the optimal thread allocation parameters corresponding to each application scenario; determine the input and output feature vector range corresponding to each application scenario based on the input and output feature parameters collected in each application scenario; According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

The step of determining the mapping relationship of the thread allocation parameter mapping table according to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario includes: according to the optimal thread allocation parameters corresponding to each application scenario. and the range of input and output feature vectors, construct a preliminary mapping table for performance testing, and obtain the testing results; when the testing results meet the performance requirements, the primary thread allocation parameters corresponding to the optimal performance of each application scenario are determined as the optimal thread allocation parameters corresponding to each application scenario. Thread allocation parameters; according to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, a preliminary mapping table is constructed for verification, and the steps for obtaining the verification results include: selecting application scenarios from each application scenario in turn as the target application scenario; Perform a performance test on the input and output requests of the target application scenario delivered by the second solid-state drive written in the preliminary mapping table to obtain the performance results of the second solid-state drive in the target application scenario. The preliminary mapping table is based on the best performance corresponding to each application scenario. The optimal primary thread allocation parameters and input and output feature vector ranges are established; return to the steps of selecting application scenarios as target application scenarios from each application scenario in turn, until all application scenarios are used as target application scenarios, and obtain the second solid state drive in Performance results under each application scenario; when the performance results under each application scenario reach the preset threshold, it is determined that the inspection results meet the performance requirements; when the performance results under each application scenario do not reach the preset threshold, it is determined that the inspection results do not meet the performance requirements Require.

Thread allocation parameters include: the number of threads for processing input and output requests issued by the GC side and the number of threads for processing input and output requests issued by the host side.

In one embodiment, a computer device is provided, including a memory, a processor, and a solid state drive. When the firmware of the solid state drive is running, the following steps are implemented:

Obtain the input and output feature data within the preset time interval; according to the input and output feature data, determine the input and output feature vector ranges in the thread allocation parameter mapping table. The thread allocation parameter mapping table maps the corresponding performance of each input and output feature vector range. Optimal thread allocation parameters; according to the input and output feature vector range, obtain the first thread allocation parameters from the thread allocation parameter mapping table; compare whether the current thread allocation parameters are consistent with the first thread allocation parameters; when they are inconsistent, change the current thread allocation parameters Adjust the allocation parameters for the first thread.

In one embodiment, the processor also implements the following steps when executing the computer program:

Denoise and integrate each input and output feature vector in the input and output feature data to obtain the average value of the input and output feature vectors; match the corresponding input and output feature vector range in the thread allocation parameter mapping table based on the average value of the input and output feature vectors .

In one embodiment, when the processor executes the computer program, the processor also implements the following steps: when receiving the input/output request, recording the input/output feature vector of the input/output request.

Among them, the mapping relationship determination method of the thread allocation parameter mapping table includes:

From the pre-selected thread allocation parameter set, select the pre-selected thread allocation parameters as the target thread allocation parameters; adjust the thread allocation parameters in the first solid-state drive according to the target thread allocation parameters; simulate various application scenarios to deliver input and output to the first solid-state drive Request a performance test to obtain the performance results of the first solid-state drive in each application scenario; collect the corresponding input and output characteristic parameters of the first solid-state drive in each application scenario; return the target thread allocation parameters selected from the pre-selected thread allocation parameter set. Steps until all pre-selected thread allocation parameters in the pre-selected thread allocation parameter set are selected to obtain performance result data. The performance result data includes the performance of the first solid-state drive in each application scenario when each pre-selected thread allocation parameter is used as the target thread allocation parameter. Results: Analyze the performance result data to determine the optimal thread allocation parameters corresponding to each application scenario; determine the input and output feature vector range corresponding to each application scenario based on the input and output feature parameters collected in each application scenario; According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

The step of determining the mapping relationship of the thread allocation parameter mapping table according to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario includes: according to the optimal thread allocation parameters corresponding to each application scenario. and the range of input and output feature vectors, construct a preliminary mapping table for performance testing, and obtain the testing results; when the testing results meet the performance requirements, the primary thread allocation parameters corresponding to the optimal performance of each application scenario are determined as the optimal thread allocation parameters corresponding to each application scenario. Thread allocation parameters; according to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, the mapping relationship of the thread allocation parameter mapping table is determined.

According to the optimal thread allocation parameters and input and output feature vector ranges corresponding to each application scenario, a preliminary mapping table is constructed for verification, and the steps for obtaining the verification results include: selecting application scenarios from each application scenario in turn as the target application scenario; Perform a performance test on the input and output requests of the target application scenario delivered by the second solid-state drive written in the preliminary mapping table to obtain the performance results of the second solid-state drive in the target application scenario. The preliminary mapping table is based on the best performance corresponding to each application scenario. The optimal primary thread allocation parameters and input and output feature vector ranges are established; return to the steps of selecting application scenarios as target application scenarios from each application scenario in turn, until all application scenarios are used as target application scenarios, and obtain the second solid state drive in Performance results under each application scenario; when the performance results under each application scenario reach the preset threshold, it is determined that the inspection results meet the performance requirements; when the performance results under each application scenario do not reach the preset threshold, it is determined that the inspection results do not meet the performance requirements Require.

Thread allocation parameters include: the number of threads for processing input and output requests issued by the GC side and the number of threads for processing input and output requests issued by the host side.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage. In the media, when executed, the computer program may include the processes of the above method embodiments. Any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and/or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous chain Synchlink DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The technical features of the above embodiments can be combined in any way. To simplify the description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, all possible combinations should be used. It is considered to be within the scope of this manual.

The above-described embodiments only express several implementation modes of the present application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all fall within the protection scope of the present application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims (9)

1. A thread self-adaptive allocation method of a solid state disk comprises the following steps:

acquiring input and output characteristic data in a preset time interval;

determining the range of the input and output feature vectors which belong to a thread allocation parameter mapping table according to the input and output feature data, wherein the thread allocation parameter mapping table maps the thread allocation parameters with optimal performance corresponding to the range of the input and output feature vectors;

Acquiring a first thread allocation parameter from the thread allocation parameter mapping table according to the input/output characteristic vector range;

comparing whether the current thread allocation parameters are consistent with the first thread allocation parameters;

when the current thread allocation parameters are inconsistent, the current thread allocation parameters are adjusted to the first thread allocation parameters;

the method for determining the mapping relation of the thread allocation parameter mapping table comprises the following steps:

sequentially selecting preselected thread allocation parameters from a preselected thread allocation parameter set as target thread allocation parameters;

adjusting thread allocation parameters in the first solid state disk according to the target thread allocation parameters;

simulating different application scenes to perform performance test on the input/output requests issued by the first solid state disk, and obtaining performance results of the first solid state disk in the application scenes;

collecting corresponding input and output characteristic parameters of the first solid state disk under each application scene;

returning to the step of sequentially selecting target thread allocation parameters from the preselected thread allocation parameter set until all preselected thread allocation parameters in the preselected thread allocation parameter set are selected, and obtaining performance result data, wherein the performance result data comprises performance results of the first solid state disk under each application scene when each preselected thread allocation parameter is used as the target thread allocation parameter;

Analyzing according to the performance result data, and determining initial thread allocation parameters with optimal performance corresponding to each application scene;

according to the input and output characteristic parameters collected under each application scene, determining the corresponding input and output characteristic vector range under each application scene;

and determining the mapping relation of the thread allocation parameter mapping table according to the initial thread allocation parameters with optimal performance and the input and output characteristic vector range corresponding to each application scene.

2. The method of claim 1, wherein the step of determining the range of the input-output feature vector to which the thread allocation parameter map belongs according to the input-output feature data comprises:

denoising and integrating each input and output characteristic vector in the input and output characteristic data to obtain an average value of the input and output characteristic vectors;

and matching the range of the input and output characteristic vectors in a thread allocation parameter mapping table according to the average value of the input and output characteristic vectors.

3. The method according to claim 1, wherein the method further comprises: when an input-output request is received, recording an input-output feature vector of the input-output request.

4. The method according to claim 1, wherein the step of determining the mapping relation of the thread allocation parameter mapping table according to the initial thread allocation parameter and the input/output feature vector range with optimal performance corresponding to each application scenario comprises:

according to the initial thread allocation parameters and the input/output characteristic vector range with optimal performance corresponding to each application scene, constructing a preliminary mapping table to perform performance test, and obtaining a test result;

when the test result meets the performance requirement, determining the initially selected thread allocation parameters with the optimal performance corresponding to each application scene as the thread allocation parameters with the optimal performance corresponding to each application scene;

and determining the mapping relation of the thread allocation parameter mapping table according to the thread allocation parameter with the optimal corresponding performance of each application scene and the input/output characteristic vector range.

5. The method of claim 4, wherein the step of constructing a preliminary mapping table for verification according to the initial thread allocation parameter and the input/output feature vector range with optimal performance corresponding to each application scenario, and obtaining a verification result comprises the following steps:

sequentially selecting an application scene from all application scenes as a target application scene;

Performing performance test on an input/output request of a second solid state disk written with a preliminary mapping table for issuing the target application scene, and obtaining a performance result of the second solid state disk in the target application scene, wherein the preliminary mapping table is established according to a preliminary thread allocation parameter and an input/output feature vector range with optimal corresponding performance of each application scene;

returning to the step of sequentially selecting application scenes from all application scenes as target application scenes until all application scenes are used as target application scenes, and obtaining a performance result of the second solid state disk under all application scenes;

when the performance result in each application scene reaches a preset threshold value, determining that the test result meets the performance requirement;

and when the performance result in each application scene does not reach the preset threshold value, determining that the test result does not meet the performance requirement.

6. The method of claim 1, wherein the thread allocation parameters comprise: the method comprises the steps of processing the number of the input/output request threads issued by a GC end and the number of the input/output request threads issued by a host end.

7. A thread adaptive allocation device for a solid state disk, the device comprising:

The data acquisition module is used for acquiring input and output characteristic data in a preset time interval;

the range determining module is used for determining the range of the input and output feature vectors which belong to a thread allocation parameter mapping table according to the input and output feature data, and the thread allocation parameter mapping table maps the thread allocation parameters with optimal performance corresponding to the range of the input and output feature vectors;

the distribution parameter acquisition module is used for acquiring a first thread distribution parameter from the thread distribution parameter mapping table according to the input/output characteristic vector range;

the comparison module is used for comparing whether the current thread allocation parameters are consistent with the first thread allocation parameters or not;

and the parameter adjustment module is used for adjusting the current thread allocation parameters to the first thread allocation parameters when the current thread allocation parameters are inconsistent, wherein the mapping relation determination mode of the thread allocation parameter mapping table comprises the steps of sequentially selecting the preselected thread allocation parameters from a preselected thread allocation parameter set as target thread allocation parameters, adjusting the thread allocation parameters in the first solid state disk according to the target thread allocation parameters, simulating each different application scene to perform performance test on the input and output requests issued by the first solid state disk, acquiring performance results of the first solid state disk under each application scene, collecting corresponding input and output characteristic parameters of the first solid state disk under each application scene, returning to the step of sequentially selecting the target thread allocation parameters from the preselected thread allocation parameter set until all the preselected thread allocation parameters in the preselected thread allocation parameter set are selected, obtaining performance result data, wherein the performance result data comprises the performance results of each preselected thread allocation parameter under each application scene when the preselected thread allocation parameters are used as the target thread allocation parameters, analyzing according to the performance result data, determining initial preferred performance parameters of each application scene, determining the initial preferred performance of each application scene according to the initial performance parameters, collecting the corresponding to the input and output characteristic vectors of each application scene, and determining the optimal performance vectors according to the application scene, and the input characteristic vectors of each application scene.

8. A solid state disk having a computer program stored thereon, which when executed by a processor performs the steps of the method of any of claims 1 to 7.

9. A computer device comprising a memory, a processor and a solid state disk, the firmware of the solid state disk implementing the steps of the method of any of claims 1 to 7 when run.