Parameter recommendation method and device, storage medium and program product

Abstract

The embodiment of the invention provides a parameter recommendation method and device, a storage medium and a program product. The method comprises the steps that a resource recovery command is sent to a target solid state disk at the ith moment; obtaining a difference value between the ith performance recovery time of the target solid state disk and the ith write-in time of executing write-in operation on the target solid state disk to obtain the ith performance recovery duration; recording an association relationship between the ith performance recovery duration and an ith interval in a first test result, wherein the ith interval is an interval between the ith write-in moment and the ith moment; and generating a recommendation interval parameter based on the first test result, and sending the recommendation interval parameter to the reference account, so that the reference account adjusts the reference solid state disk based on the recommendation interval parameter, the recommended interval parameter is used for representing a time interval between sending a resource recovery command to the reference solid state disk and executing a write-in operation on the reference solid state disk; the problem that the optimization efficiency of the SSD is low is solved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computers, and more specifically, to a parameter recommendation method and device, a storage medium, and a program product. Background Art

[0002] Solid State Drive (SSD) has become a key technology in the field of high-performance computing and big data storage due to its excellent read and write performance and high reliability. However, SSD performance optimization is a complex process, especially in the performance recovery phase after the TRIM operation. How to select the configuration parameters for adjusting the SSD has become a difficult problem for R&D personnel. Improper parameter selection will extend the performance recovery time and affect the stability and user experience of the SSD. Among them, the TRIM operation is a data management technology for SSDs. Its main purpose is to optimize the performance and life of solid-state drives. The TRIM operation allows the operating system to inform the SSD which data blocks are no longer needed, so that the SSD can clean up these data blocks in the background, thereby improving write performance and extending the life of the hard drive.

[0003] Currently, there is a lack of a method to guide R&D personnel in configuring the configuration parameters of the SSD, especially in optimizing the performance after the SSD reaches a steady state, which leads to a technical problem of low optimization efficiency of the SSD.

[0004] Therefore, there is a technical problem in the prior art that the optimization efficiency of SSD is low. Summary of the invention

[0005] The embodiments of the present application provide a parameter recommendation method and device, a storage medium, and a program product to at least solve the technical problem of low optimization efficiency of SSD.

[0006] According to an embodiment of the present application, a parameter recommendation method is provided, comprising: sending a resource recovery command to a target solid-state drive at the i-th moment, wherein the resource recovery command is used to instruct to perform a recovery operation on a data block to be recovered included in the target solid-state drive, and i is a positive integer and i is less than N; obtaining the difference between the i-th performance recovery moment of the target solid-state drive and the i-th write moment of performing a write operation on the target solid-state drive, and obtaining the i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state drive at the i-th performance recovery moment is a target threshold; recording the association between the i-th performance recovery duration and the i-th interval in a first test result, wherein the i-th interval is the interval between the i-th write moment and the i-th moment; generating a recommended interval parameter based on the first test result, and sending the recommended interval parameter to a reference account, so that the reference account adjusts the reference solid-state drive based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state drive and performing a write operation on the reference solid-state drive.

[0007] According to another embodiment of the present application, a parameter recommendation device is provided, including: a first sending unit, used to send a resource recovery command to a target solid-state hard disk at the i-th moment, wherein the resource recovery command is used to instruct to perform a recovery operation on a data block to be recovered included in the target solid-state hard disk, i is a positive integer and i is less than N; an acquisition unit, used to obtain the difference between the i-th performance recovery moment of the target solid-state hard disk and the i-th write moment of performing a write operation on the target solid-state hard disk, and obtain the i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state hard disk at the i-th performance recovery moment is the target threshold; a recording unit, used to record the association between the i-th performance recovery duration and the i-th interval in a first test result, wherein the i-th interval is the interval between the i-th write moment and the i-th moment; a second sending unit, used to generate a recommended interval parameter based on the first test result, and send the recommended interval parameter to a reference account, so that the reference account adjusts the reference solid-state hard disk based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state hard disk and performing a write operation on the reference solid-state hard disk.

[0008] According to another embodiment of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0009] According to another embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0010] According to another embodiment of the present application, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps of the method described in each embodiment of the present application are implemented.

[0011] Through the embodiments provided by the present application, a resource recovery command is sent to a target solid-state drive at the i-th moment, wherein the resource recovery command is used to instruct to perform a recovery operation on a data block to be recovered included in the target solid-state drive, and i is a positive integer and i is less than N; the difference between the i-th performance recovery moment of the target solid-state drive and the i-th write moment of performing a write operation on the target solid-state drive is obtained to obtain the i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state drive at the i-th performance recovery moment is the target threshold; the association between the i-th performance recovery duration and the i-th interval is recorded in a first test result, wherein the i-th interval is the interval between the i-th write moment and the i-th moment; a recommended interval parameter is generated based on the first test result, and the recommended interval parameter is sent to a reference account, so that the reference account adjusts the reference solid-state drive based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state drive and performing a write operation on the reference solid-state drive. In other words, using the embodiments of the present application, traditional parameter adjustment usually requires a lot of trial and error process, while this method can quickly identify the key time parameters that affect performance recovery by presetting the test cycle until the time interval (N) is reached, thereby greatly shortening the exploration time of R&D personnel on parameter configuration and improving the efficiency of SSD performance optimization. By systematically testing the impact of different post-TRIM write time intervals (P) on the performance recovery time of the SSD, it is possible to generate recommended interval parameters to guide R&D personnel to reasonably arrange the time interval between sending resource recovery commands and subsequent write operations in actual applications to reduce performance recovery time, thereby further improving the optimization efficiency of the SSD. In summary, the use of the embodiments of the present application solves the technical problem of low optimization efficiency of the SSD and achieves the technical effect of improving the optimization efficiency of the SSD. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a flow chart of a parameter recommendation method according to an embodiment of the present application;

[0013] Figure 2 is a schematic diagram of a parameter recommendation method according to an embodiment of the present application;

[0014] Figure 3 is a schematic diagram of another parameter recommendation method according to an embodiment of the present application;

[0015] Figure 4It is a structural block diagram of a parameter recommendation device according to an embodiment of the present application. DETAILED DESCRIPTION

[0016] The embodiments of the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0017] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0018] As an optional solution, the recommended method for the above parameters is as follows: Figure 1 Shown include:

[0019] S102, sending a resource recovery command to the target solid state drive at the i-th time, wherein the resource recovery command is used to instruct to perform a recovery operation on the data blocks to be recovered included in the target solid state drive, and i is a positive integer and i is less than N.

[0020] It should be noted that the above parameter recommendation method can be applied to at least the following scenarios but is not limited to:

[0021] 1) Data center and enterprise storage system optimization scenarios: In data centers or enterprise storage systems, SSDs are widely used to improve data processing speed and storage efficiency. Through the above method, system administrators can test and determine the time interval between the execution of the data block release notification (TRIM) operation and the write operation required for the shortest performance recovery time, thereby optimizing the scheduling strategy of the solid-state drive, reducing business interruption time, and improving the overall performance of the storage system. Among them, TRIM is a key function for optimizing performance and durability on solid-state drives (SSDs). At a more detailed technical level, TRIM is a storage device command that allows the operating system to notify the SSD which data blocks can be considered free so that they can be reused in future write operations. When a file is deleted or overwritten, the operating system usually marks the storage area of ​​the file as logically free, but the physical storage blocks on the SSD are not immediately erased or released. Due to the erase and write characteristics of the SSD, that is, the storage blocks must be erased before writing new data, so if there is no TRIM command, the SSD may undergo additional erase steps when writing new data, which will reduce the write speed and increase wear. The role of TRIM is that when the operating system recognizes that there are data blocks that are no longer in use, it will send a TRIM command to the SSD, instructing the SSD controller that these data blocks can be erased for reuse. After receiving the TRIM command, the SSD can perform garbage collection in the background to erase the data blocks that are no longer needed, and update its internal mapping table to reflect which storage blocks are free. Since the TRIM command enables the SSD to know in advance which areas can be reused, when actually writing new data, the SSD can directly write the new data to these erased "free" blocks without the need for an additional erase step, which significantly improves the write performance of the SSD and helps extend the service life of the SSD.

[0022] Optionally, the basic process of TRIM operation is as follows: Operating system side: When a file is deleted, the operating system updates its internal file system metadata and marks the file's data blocks as deleted. However, the SSD controller does not receive this information immediately. Send TRIM command: The operating system tells the SSD controller which data blocks are no longer needed by sending TRIM command. This process is usually triggered automatically after the file is deleted, the system is idle or under certain conditions. SSD controller response: After receiving the TRIM command, the SSD controller updates its internal mapping table and marks those data blocks that the operating system considers to be free. Subsequently, the SSD controller can give priority to erasing these marked blocks in its background garbage collection process to facilitate future data writing. Garbage collection: When the SSD is idle, the controller will perform garbage collection operations to erase all data blocks marked as invalid and reorganize valid data blocks to optimize the use of storage space. The information from the TRIM operation helps the SSD controller to perform garbage collection more efficiently because it already knows which blocks can be safely erased.

[0023] 2) Server virtualization environment: In a server virtualization environment, multiple virtual machines (VMs) may share the same physical SSD. The above method can help identify the optimal TRIM operation interval to ensure that the VM can quickly recover to normal performance when performing write operations, avoiding performance bottlenecks caused by TRIM operations that affect the VM's response speed and user experience.

[0024] Optionally, in some embodiments, multiple target solid-state drives of the same model may be selected to perform multiple tests, thereby ensuring the accuracy of the test results.

[0025] Optionally, the resource recovery command here can be but not limited to being used to indicate a global garbage collection notification command (global TRIM), and the global TRIM command can be but not limited to being issued through an I / O test tool (Flexible I / O Tester, referred to as FIO) tool, wherein the parameters are set as follows: size = 100%, TRIM full disk. When the target solid-state drive receives the above-mentioned global TRIM command, it will record all data blocks as data blocks to be recovered and recycle them. Here, FIO is a tool for testing and measuring the read and write speed and I / O performance of storage devices (such as hard disks, SSDs, etc.), which provides a flexible and configurable way to simulate different types of I / O loads to evaluate the performance of storage systems. FIO supports multi-threaded and multi-process I / O testing, and can set different test modes, such as random read and write, sequential read and write, etc., as well as various parameters.

[0026] S104, obtaining the difference between the i-th performance recovery moment of the target solid-state hard disk and the i-th write moment of executing the write operation on the target solid-state hard disk, and obtaining the i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state hard disk at the i-th performance recovery moment is the target threshold.

[0027] It should be noted that the above-mentioned target threshold can be, but is not limited to, used to indicate the maximum performance level that can be achieved when the hard disk performs random write operations again after a certain recovery time after the SSD performs the TRIM operation. Specifically, when the SSD recovers from the low performance state after the TRIM operation to its optimal performance state, the highest OPS value recorded is the maximum IOPS. This value is crucial for evaluating the performance recovery speed and performance stability of the SSD after the TRIM operation, because it directly reflects the upper limit of the read and write performance of the SSD when it reaches the optimal working state. For example, when the 1.92TP series SSD performs random write operations immediately after TRIM, the maximum input / output operations per second (IOPS) can reach 686.907k, which means that under optimal conditions, this SSD can process up to 686,907 read and write operations per second. When testing at different waiting times (such as 10 minutes, 30 minutes, and 60 minutes), the maximum IOPS value and the time required to reach this performance level will be different. This data helps analyze the impact of TRIM operations on SSD performance and the performance recovery characteristics of the hard drive at different waiting times.

[0028] It should be noted that IOPS is an important indicator to measure the performance of storage devices (such as hard disk drives, solid-state drives (SSDs) or storage systems). It indicates the number of read or write operations that the device can complete in one second. The higher the IOPS value, the faster the storage device reads and writes, and the better the performance.

[0029] It should be noted that the above-mentioned i-th performance recovery duration is obtained by calculating the time difference between the performance recovery moment and the moment when the write operation is performed on the solid-state hard drive. This reflects the time required for the hard drive to reach stable performance after the TRIM operation (that is, all data blocks will be recorded as data blocks to be recovered and recycled).

[0030] S106, recording the correlation between the i-th performance recovery duration and the i-th interval in the first test result, wherein the i-th interval is the interval between the i-th writing moment and the i-th moment.

[0031] It should be noted that the association between the above-mentioned i-th performance recovery time and the i-th interval can be recorded in the above-mentioned first reference test result in the form of a table, but is not limited to it, and can also be recorded in other forms, and there is no limitation on this in the present embodiment. Furthermore, the above-mentioned i-th interval can be flexibly set according to actual needs. For example, the i-th interval can be set to 0, 10min, 30min, 60min, etc., respectively, and there is no limitation on this in the present embodiment. For example, assuming that the above-mentioned first association relationship is recorded in the form of a table, as shown in Table 1, a first association relationship is recorded in each row of the table:

[0032] Table 1

[0033] Time interval (minutes) Performance recovery time (seconds) 0 4099 10 3579 30 2418 60 211

[0034] S108, generating a recommended interval parameter based on the first test result, and sending the recommended interval parameter to the reference account, so that the reference account adjusts the reference solid-state hard disk based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state hard disk and performing a write operation on the reference solid-state hard disk.

[0035] Optionally, the above-mentioned generation of recommended interval parameters based on the first test result may include but is not limited to: S1, determining based on the first test result that the performance recovery time of the solid-state drive is inversely proportional to the time interval for performing a write operation after sending a resource recovery command to the solid-state drive; S2, obtaining first adjustment requirement information of the reference solid-state drive and the load level of the reference solid-state drive, wherein the load level is used to characterize the load condition of the reference solid-state drive; S3, determining the recommended interval parameters based on the load level of the reference solid-state drive and the first test result when the first adjustment requirement information is used to indicate that the performance of the reference solid-state drive is to be improved.

[0036] It should be noted that the above-mentioned reference account can be but is not limited to a social account used to indicate professional technicians used to develop or adjust the reference solid-state drive or a system account of a specific optimization system, etc., and this is not limited to this in the present embodiment.

[0037] Optionally, the reference account may adjust the reference solid state drive based on the recommended interval parameters, including but not limited to: adjusting the reference solid state drive's post-TRIM write time interval (i.e., the time interval between write operations after executing the TRIM command) to a time interval that matches the recommended time interval.

[0038] For example, the above steps can be explained based on, but not limited to, the following examples:

[0039] Assume that a solid-state drive (SSD) with a capacity of 1.92TB is being tested. First, the SSD is formatted using a low-level formatting tool with a block size of 512k. Then, the SSD is written to a steady state by sequentially writing 128k data blocks twice and randomly writing 4k data blocks for 2 hours.

[0040] Then, in the steady state, a global TRIM command is sent to the SSD. At four different time points P, 0 minutes, 10 minutes, 30 minutes, and 60 minutes after the TRIM command is sent, random write operations are performed on the SSD and the IOPS parameters are recorded. The time point when the IOPS parameter reaches the target threshold is the performance recovery moment.

[0041] For example, it is recorded that when P is 0 minutes, the performance recovery time is 4099 seconds; when P is 10 minutes, the performance recovery time is 5310 seconds. This shows that there is an inverse relationship between the performance recovery time and the interval time P of the write operation after TRIM.

[0042] By changing P and repeating the above test steps until P reaches the set N value (for example, N = 60 minutes), a series of performance recovery time data are obtained, which constitute the first test result. Based on these data, the first test result is generated, revealing the inverse relationship between P and performance recovery time.

[0043] Finally, based on the first test result, the system generates the first recommended parameter, for example, recommending that the time interval between TRIM and write operation is 30 minutes to achieve the optimal performance recovery speed. This recommended parameter will be sent to the reference account via email or platform notification, so that the user or system administrator can adjust the performance strategy of the SSD to optimize its performance after the TRIM operation.

[0044] It should be noted that the above example is an optional example provided for the convenience of explaining the above example, and there is no limitation on the recommended method for the above parameters.

[0045] Through the embodiments provided by the present application, a resource recovery command is sent to a target solid-state drive at the i-th moment, wherein the resource recovery command is used to instruct to perform a recovery operation on a data block to be recovered included in the target solid-state drive, and i is a positive integer and i is less than N; the difference between the i-th performance recovery moment of the target solid-state drive and the i-th write moment of performing a write operation on the target solid-state drive is obtained to obtain the i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state drive at the i-th performance recovery moment is the target threshold; the association between the i-th performance recovery duration and the i-th interval is recorded in a first test result, wherein the i-th interval is the interval between the i-th write moment and the i-th moment; a recommended interval parameter is generated based on the first test result, and the recommended interval parameter is sent to a reference account, so that the reference account adjusts the reference solid-state drive based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state drive and performing a write operation on the reference solid-state drive. In other words, using the embodiments of the present application, traditional parameter adjustment usually requires a lot of trial and error process, while this method can quickly identify the key time parameters that affect performance recovery by presetting the test cycle until the time interval (N) is reached, thereby greatly shortening the exploration time of R&D personnel on parameter configuration and improving the efficiency of SSD performance optimization. By systematically testing the impact of different post-TRIM write time intervals (P) on the performance recovery time of the SSD, it is possible to generate recommended interval parameters to guide R&D personnel to reasonably arrange the time interval between sending resource recovery commands and subsequent write operations in actual applications to reduce performance recovery time, thereby further improving the optimization efficiency of the SSD. In summary, the use of the embodiments of the present application solves the technical problem of low optimization efficiency of the SSD and achieves the technical effect of improving the optimization efficiency of the SSD.

[0046] As an optional solution, generating the recommended time interval parameter based on the first test result includes:

[0047] S1, based on the first test result, determining that the performance recovery time of the solid state drive is inversely proportional to the time interval between executing a write operation after sending a resource recovery command to the solid state drive.

[0048] S2, obtaining first adjustment requirement information of a reference solid state drive and a load level of the reference solid state drive, wherein the load level is used to characterize a load condition of the reference solid state drive.

[0049] S3, when the first adjustment requirement information is used to indicate that the performance of the reference solid state drive is to be improved, determining a recommended interval parameter based on the load level of the reference solid state drive and the first test result.

[0050] Optionally, the following method may be used, but is not limited to, to determine, based on the first test result, that the performance recovery time of the solid state drive is inversely proportional to the time interval for performing a write operation after sending a resource recovery command to the solid state drive:

[0051] S1, Data Analysis:

[0052] Chart drawing: Use a scatter plot to intuitively represent the relationship between the time interval P value and the performance recovery time. In the scatter plot, the P value is the X-axis, the performance recovery time is the Y-axis, and each test data point is represented as a point on the chart;

[0053] Trend identification: Observe the scatter plot. If the performance recovery time shows a downward trend (i.e., the Y value decreases) as the P value increases, this preliminarily indicates that there may be an inverse relationship between the two.

[0054] S3, Statistical Modeling:

[0055] Regression analysis: Apply linear regression or other appropriate regression models, such as power law regression, to fit the relationship between P value and performance recovery time. If it is an inverse relationship, the regression model may be in the form of Y = a / P + b, where a and b are regression coefficients;

[0056] Model evaluation: By calculating the R 2 The value (coefficient of determination) was used to assess the goodness of fit of the model. 2 Values ​​close to 1 indicate a good model fit, confirming the hypothesis of an inverse relationship;

[0057] Residual analysis: Check the residual plot of the regression model to ensure that the residual distribution is random and there is no obvious pattern or trend, which further verifies the validity of the model;

[0058] S3, Result confirmation and reporting:

[0059] The results confirm that if the coefficients and R 2 If the value supports the hypothesis of an inverse relationship and the residual analysis does not reveal systematic errors in the model, then we can confirm that there is an inverse relationship between performance recovery time and P value.

[0060] Report writing: Organize the analysis results into a data analysis report, which may include but is not limited to: description of the test method and data collection process, scatter plots and regression line charts, detailed parameters of the regression model, R 2 Results of value and residual analysis, clear conclusions and interpretation of the inverse relationship.

[0061] It should be noted that the first adjustment requirement information may be used, but is not limited to, to indicate that a user or system administrator wishes to adjust the performance of the solid-state drive, and the load level is used to characterize the load of the solid-state drive, that is, how busy the SSD is in processing data read and write requests. The load level may be determined based on the usage of the disk (for example, whether it is used for lightweight read and write or high-load read and write intensive applications) and the frequency of read and write requests. It helps the system understand the working pressure of the SSD in actual application scenarios so as to adjust performance parameters more accurately.

[0062] Optionally, based on the first test result and the load level of the reference SSD, the system recommends a time interval between the TRIM operation and the subsequent write operation. The selection of this time interval is intended to minimize the performance recovery time while ensuring that the SSD can maintain stable and efficient performance under different load levels.

[0063] As an optional implementation, the above-mentioned determination of the first recommended time interval based on the load level of the reference solid state drive and the first test result may include but is not limited to:

[0064] When the load information indicates that the load level of the reference solid state drive is less than the first predetermined level, adding the first recommended time interval to the first recommended parameter, wherein the higher the load level, the greater the load of the solid state drive, the first recommended time interval is greater than the first threshold, the first recommended time interval is less than the second threshold, and the second threshold is greater than the first threshold;

[0065] When the load information indicates that the load level of the reference solid state drive is greater than the first predetermined level, a second recommended time interval is added to the first recommended parameter, wherein the second recommended time interval is greater than the third threshold, and the second recommended time interval is less than the fourth threshold, the first threshold is greater than the fourth threshold, and the fourth threshold is greater than the third threshold.

[0066] Optionally, as an optional example, the above steps may be explained by, but not limited to, the following examples:

[0067] Assume the following first test result:

[0068] When a write operation is performed immediately after TRIM (ie, time interval P=0), the performance recovery time T=4000 seconds.

[0069] When the write operation is performed after waiting for 10 minutes (ie, the time interval P=600 seconds) after TRIM, the performance recovery time T=3000 seconds.

[0070] When the write operation is performed after waiting for 30 minutes (ie, time interval P=1800 seconds) after TRIM, the performance recovery time T=2000 seconds.

[0071] When the write operation is performed after waiting for 60 minutes (ie, time interval P=3600 seconds) after TRIM, the performance recovery time T=1500 seconds.

[0072] Determine the first adjustment requirement information: Under high load conditions, the performance recovery time of the SSD is too long, which affects the business response speed. Therefore, the first adjustment requirement information is set, that is, to improve the performance recovery speed of the SSD.

[0073] Obtaining the load level: Using the system monitoring tool, we obtained the load level information of the reference SSD. For example, if the SSD is mainly responsible for high-load read-write intensive tasks, its load level may be marked as "high".

[0074] Analysis of the first test results and load levels: Based on the first test results, we can see that as the waiting time after TRIM increases, the performance recovery time gradually decreases. Considering that administrators want to improve the performance recovery speed under high load conditions, combined with the above test results, it can be inferred that under high load levels, a longer waiting time is more conducive to rapid performance recovery.

[0075] Determine the first recommended time interval: Based on the above analysis, if under high load levels, in order to minimize performance recovery time, the system may recommend a longer TRIM wait time, such as 30 minutes or 60 minutes. This is because a longer wait time allows the SSD to have more time to perform internal optimization and garbage collection after the TRIM operation, thereby recovering to the best performance state faster.

[0076] It should be noted that the above example is an optional example provided for the convenience of explaining the above example, and there is no limitation on the recommended method for the above parameters.

[0077] According to the embodiment of the present application, a flexible time interval recommendation is made based on the "first adjustment requirement information" of a specific solid-state drive. If the performance of the solid-state drive needs to be improved, the system will intelligently calculate an optimal time interval based on its load level and previous test results to ensure that the TRIM operation will not be performed during high-load periods, thereby avoiding performance bottlenecks.

[0078] As an optional solution, before sending the resource recovery command to the target solid state drive at the i-th time, the following is also included:

[0079] S1, formatting the candidate target solid state drive.

[0080] It should be noted that the above formatting process can be, but is not limited to, using low-level formatting technology to clear all data in the disk. Specifically, in this embodiment, the low-level formatting function of the FIO tool can be used to format the solid-state hard disk, but is not limited to it. The block size of the operation can be set to 512 kilobytes or other sizes, which is not limited in this embodiment. The purpose of the formatting process is to restore the solid-state hard disk to its initial state and clear all data.

[0081] S2, performing a write operation on the formatted candidate solid-state hard disk until the formatted candidate solid-state hard disk reaches a steady state, thereby obtaining a target solid-state hard disk.

[0082] It should be noted that steady state means that the hard disk enters a state in which any further write operations will trigger the garbage collection mechanism.

[0083] As an optional solution, a write operation is performed on the formatted candidate solid-state hard disk until the formatted candidate solid-state hard disk reaches a steady state, and the target solid-state hard disk is obtained including:

[0084] Perform j full-disk sequential write operations on the formatted candidate solid-state hard disk to obtain a first solid-state hard disk, where j is a positive integer;

[0085] Perform K hours of random write operations on the first solid state drive to obtain a target solid state drive, where K is a positive integer.

[0086] Optionally, before the formatted candidate SSD reaches a steady state, it is first necessary to perform j full-disk sequential write operations. Here, j is a positive integer representing the number of write cycles that need to be performed. The purpose is to ensure that the SSD has written enough data so that its internal storage units reach a certain level of utilization, creating a benchmark for subsequent performance tests. Sequential writing means that data is written continuously in the order of LBA (logical block address), which is usually used to fill all available space on the SSD. For example, j can be 2, and the size of the above sequential write can be 128K, that is, a full-disk 128k sequential write is performed twice.

[0087] It should be noted that after completing j times of full disk sequential writes, the SSD is then subjected to K hours of random write operations in order to reach a steady state. Here, K is also a positive integer, indicating the duration of the random write. Random write means that data is written to any location of the SSD, rather than to a continuous address. This operation is closer to the data write mode in actual application scenarios. Through K hours of random writes, the storage units inside the SSD will experience a more complex usage pattern, which will eventually trigger the garbage collection mechanism of the SSD, ensuring that the SSD enters a stable working state, namely a "steady-state solid-state drive". For example, K here can be 2, and bs can be 4K.

[0088] By adopting the embodiment of the present application, by performing multiple full-disk sequential write operations (j times) and subsequent random write operations (K hours) on the formatted SSD, it is possible to ensure that the SSD reaches a stable performance state, namely steady state. Steady state is an ideal benchmark for evaluating and comparing the performance of solid-state drives, because at this time, indicators such as the read and write speed and latency of the hard disk will remain in a relatively stable and predictable range, and will not be affected by initial formatting, empty disk effect or write mode. This provides a consistent starting point for subsequent performance testing and analysis. Furthermore, the combination of sequential write operations and random write operations can activate the garbage collection mechanism of the SSD. In steady state, due to the limited available space on the SSD, each write operation will trigger garbage collection, which helps to clean up data blocks that are no longer used, thereby optimizing the read and write performance of the hard disk. By ensuring that the SSD has reached a steady state before testing, the actual effect of performance recovery after the TRIM operation and the impact of the garbage collection mechanism on performance can be more accurately evaluated.

[0089] As an optional solution, after sending the resource recovery command to the target solid state drive at the i-th time, the method further includes:

[0090] At the i-th writing time, a random write operation is performed on the target solid state drive for w hours, and the number of input / output operations per second of the target solid state drive during the random write operation is recorded, where w is a positive integer.

[0091] For example, w here can be 3, that is, random writing is performed for 3 hours, the IOPS value during this period is recorded, and the time of the maximum IOPS is recorded.

[0092] By using the embodiment of the present application, performing random write operations for w hours and recording IOPS, the performance stability of the solid-state drive after TRIM can be evaluated. Long-term random write tests can reveal the performance fluctuations of the solid-state drive under high load, which is crucial for the application of solid-state drives in critical business scenarios because it directly affects the system's response time and data processing efficiency.

[0093] As an optional solution, the recommended method for the above parameters also includes:

[0094] S1, determine each reference solid state drive among H reference solid state drives included in the reference solid state drive set as a current solid state drive and perform the following steps, wherein the hard drive capacities of the H reference solid state drives are increased one by one.

[0095] It should be noted that the above-mentioned reference solid-state drive set can be, but is not limited to, a series of SSDs to form a candidate solid-state drive set, which includes H SSDs, and the capacities of these SSDs increase in sequence. For example, the set may include SSDs of different capacity points such as 0.96T, 1.60T, 3.2T, and 6.4T.

[0096] S2, sending a resource recovery command to the current solid state drive at a first moment, wherein the current solid state drive, upon receiving the resource recovery command, will perform a recovery operation on the data blocks to be recovered included in the current solid state drive.

[0097] Optionally, before the above S2, but not limited to, the following may be included: formatting the current solid state drive, and then performing a write operation on it until it reaches a steady state. For example, performing a low-level format of FIO on the current solid state drive, and then performing a full disk 128k sequential write twice and a 4k random write for 2 hours on the current solid state drive, so that the disk enters a steady state as quickly as possible.

[0098] S3, obtain the difference between the current performance recovery moment of the current solid-state hard disk and the second moment of performing a write operation on the current solid-state hard disk, and obtain the current performance recovery duration, wherein the number of input / output operations per second of the current solid-state hard disk reaches a reference threshold at the current performance recovery moment.

[0099] Optionally, in some embodiments, a fixed time interval between the second moment and the first moment can be selected for testing, or multiple different time intervals can be used for testing, which is not limited in this embodiment. It should be noted that the value of the above reference threshold is related to the current solid-state drive, and is used to indicate the highest performance level that can be achieved when the hard disk performs a random write operation again after a certain recovery time after the current solid-state drive performs a TRIM operation.

[0100] S4, recording the correlation between the current storage capacity of the solid state drive and the current performance recovery time in the second test result.

[0101] It should be noted that the correlation between the storage capacity of the current solid-state hard disk and the current performance recovery time can be recorded in the second test result in any form such as a table or report, but is not limited to it, and can also be recorded in other forms. This is not limited to any limitation in the present embodiment.

[0102] S5, generating a recommended capacity parameter based on the second test result, and sending the recommended capacity parameter to the reference account, so that the reference account adjusts the reference solid state drive based on the recommended capacity parameter.

[0103] As an optional example, the above steps can also be explained by, but not limited to, the following examples: first, the disk is put into steady state, then TRIM is performed on the entire disk, after waiting for a certain period of time, random write operation is performed, and how long it takes to achieve the highest performance of the disk: S1, low-level formatting of the disk, bs=512k, clearing all data in the disk. S2, perform two sequential writes of 128k on the entire disk, and 4k random writes for 2 hours, so that the disk enters steady state as quickly as possible. (This sequential write bs=128k and random write bs=4k is the optimal way for the disk to enter steady state). S3, use the FIO tool to issue a TRIM command, size=100%, TRIM the entire disk; S4, after the above command is completed, wait for 0min / 10min / 30min / 60min respectively. S5, random write for 3 hours, record the IOPS value during this period, and record the time of the maximum IOPS as T. S6, repeat the above steps multiple times according to the different waiting times in step d. S7, draw an IOPS fluctuation chart according to the values ​​in e. The test results are shown in Table 2:

[0104] Table 2

[0105]

[0106] From this we can draw the following conclusions:

[0107] 1) According to the comparison of IOPS recovery time results, it is found that the performance recovery time of random write immediately after TRIM, and after 10 minutes, 30 minutes, and 60 minutes is better than that of X2 products, indicating that the performance of this product is outstanding.

[0108] 2) As the waiting time increases, the performance recovery time gradually decreases. That is, the waiting time T1 is inversely proportional to the performance recovery time T2 (used to represent the first test result above);

[0109] 3) As the capacity increases, the performance recovery time gradually decreases, that is, the disk capacity Ca is inversely proportional to the performance recovery time T2 (used to represent the second test result above).

[0110] Optionally, the above-mentioned generation of recommended capacity parameters based on the second test result may include but is not limited to: determining, based on the second test result, that the performance recovery time is inversely proportional to the capacity of the solid-state drive; obtaining second adjustment requirement information of the reference solid-state drive and the amount of adjustment resources held by the reference account, wherein the adjustment resources are used to adjust the capacity of the reference solid-state drive; and determining the recommended capacity parameters based on the amount of resources and the second test result when the second adjustment requirement information is used to indicate that the performance of the reference solid-state drive is to be improved.

[0111] It should be noted that the reference account may adjust the configuration parameters of the reference solid-state drive based on the recommended capacity parameters, including but not limited to: adjusting the storage capacity of the reference solid-state drive to a capacity that matches the recommended capacity parameters.

[0112] It should be noted that, in some embodiments, the performance of different reference solid-state drives may be compared by comparing the performance recovery times of different reference solid-state drives.

[0113] As an optional example, the above steps may be explained by, but not limited to, the following examples:

[0114] Assume that the test of claim 5 is being performed on a group of SSDs with increasing capacities, and the SSD capacities are 0.96T, 1.60T, 3.2T, 6.4T, and 12.8T, respectively, that is, H=5. Perform a low-level formatting operation (bs=512k) on all SSDs, and then perform a full disk write operation until the SSD reaches a steady state. After reaching the steady state, send a global TRIM command to each SSD, and then immediately start to perform a random write operation, record the IOPS parameters of each SSD, until the IOPS reaches the target threshold, and record this moment. Calculate the time difference from sending the TRIM command to the IOPS reaching the target threshold, that is, the current performance recovery duration. For example, for an SSD with a capacity of 1.60T, the performance recovery duration is 2266 seconds, and record the second association relationship between this result and the capacity of the SSD (1.60T). Repeat this process until all SSDs with capacity points (that is, H candidate SSDs) are tested. Generate a second reference test result, which contains the second association relationship between the performance recovery duration and capacity of all SSDs. The second reference test result is analyzed to generate a second test result, and it is confirmed that there is an inverse relationship between the performance recovery time and the SSD capacity. For example, it is found that as the capacity increases, the performance recovery time gradually decreases. Based on the second test result, a second recommended parameter is generated, and it is recommended to use a larger capacity SSD in a specific business scenario to shorten the performance recovery time after the TRIM operation. The second recommended parameter is sent to the system administrator or SSD user so that they can adjust the capacity and configuration of the SSD according to the recommended parameter to optimize the read and write performance of the system.

[0115] It should be noted that the above example is an optional example provided for the convenience of explaining the above example, and there is no limitation on the recommended method for the above parameters.

[0116] By using the embodiments of the present application, the ability to predict the performance recovery time of hard disks of different capacities is provided. This is particularly important for hard disk management in data centers or high-performance computing environments, and can help plan storage resources in advance to avoid unexpected downtime or service degradation caused by hard disk performance fluctuations. The ability to generate recommended capacity parameters based on test results helps users or system administrators select the most appropriate SSD capacity based on actual needs and performance goals to achieve optimal storage performance and cost-effectiveness. In addition, by generating recommended capacity parameters, performance bottlenecks or excessive costs caused by selecting inappropriate SSD capacity can be avoided, thereby improving customer satisfaction and user experience with SSD products.

[0117] As an optional solution, generating the second recommended parameter based on the second test result includes:

[0118] S1, based on the second test result, determines that the performance recovery time is inversely proportional to the capacity of the solid state drive.

[0119] S2, obtaining second adjustment requirement information of the reference solid state drive and the resource amount of adjustment resources held by the reference account, wherein the adjustment resources are used to adjust the capacity of the reference solid state drive.

[0120] S3, when the second adjustment requirement information is used to indicate that the performance of the reference solid state drive is to be improved, determining a recommended capacity parameter based on the resource amount and the second test result.

[0121] Optionally, the specific implementation of determining that the performance recovery time is inversely proportional to the capacity of the solid-state drive based on the second test result may include but is not limited to referring to the following steps:

[0122] Data integration: Extract the performance recovery time and corresponding storage capacity data of all H reference SSDs from the second test results to ensure data integrity, including the performance recovery time of each capacity point SSD.

[0123] Correlation analysis: Analyze the correlation between performance recovery time and SSD capacity, and use statistical methods to verify whether they are inversely proportional. You can use a scatter plot to visually display data distribution, and use linear or nonlinear regression analysis to fit the data to find a model for the inverse relationship.

[0124] Model construction: If the initial analysis supports the inverse relationship hypothesis, a mathematical model is constructed to describe this relationship. For example, the inverse relationship between performance recovery time T and SSD capacity C can be expressed as T = a / C + b, where a and b are regression coefficients.

[0125] Model validation: Use a set of data that was not used in model building to verify the accuracy of the model. Check the difference between the model's predicted value and the actual performance recovery time to ensure that the model has good generalization ability.

[0126] Conclusion Summary: If the validation results support the inverse relationship, summarize the findings. Clearly indicate that as the SSD capacity increases, the performance recovery time decreases; and when the SSD capacity decreases, the performance recovery time increases.

[0127] Generate second test results: Write a test report that includes the results of data analysis, model building and validation, and the conclusions about the inverse relationship. The report may include a scatter plot, the equation of the regression model, a comparison chart of the validation data, and an explanation and discussion of the inverse relationship.

[0128] It should be noted that the second adjustment requirement information indicates the user or system administrator's requirement for adjusting the performance of the solid-state drive, and the resource amount may be, but is not limited to, used to indicate the resources available to the system administrator or user when optimizing the SSD performance, including but not limited to budget, storage space, hardware upgrade capability, etc. The size of the resource amount will affect the scope of the optimization solutions that can be implemented.

[0129] Optionally, the above step of determining the recommended capacity parameter based on the resource amount and the second test result may be implemented based on but not limited to the following steps:

[0130] S1, formulate strategy: Based on the inverse relationship revealed in the second test result, combined with the user's specific performance improvement requirements and resource volume, formulate a recommended strategy. The strategy should take into account the resource volume constraints and the performance improvement effect brought by the capacity increase.

[0131] S2, generate recommended parameters: extract the performance recovery time data of different capacity points from the second test results, combine with resource quantity evaluation, and use data modeling methods (such as regression analysis) to predict which capacity SSD can best meet user needs.

[0132] S3, check constraints: Ensure that the recommended capacity parameters meet the user's resource constraints, such as budget and storage space availability. If the amount of resources is limited, you may need to balance performance improvement with cost and select the capacity point with the best cost-effectiveness.

[0133] S4, sending recommended parameters: sending the generated recommended capacity parameters to the reference account, i.e., the account of the system administrator or user. The recommended parameters should include specific recommendations for the SSD capacity and the reasons for selecting the capacity, such as the performance improvement percentage, the expected performance recovery time reduction, etc.

[0134] For example, the above steps can be explained by the following examples, but are not limited to:

[0135] Assume that an enterprise-level data center is being managed, which contains SSDs of different capacities for processing a large number of data read and write requests. Through the above method, it is found that large-capacity SSDs can recover performance faster after TRIM operations. At the same time, it is known that the budget and storage space resources of the data center are sufficient to support the upgrade or replacement of SSDs. Therefore, based on the second test result, a second recommended parameter is generated, which recommends upgrading the SSD capacity in the data center to 6.40T to reduce the performance recovery time and improve the overall data processing efficiency. This recommended parameter is then sent to the reference account, that is, the system administrator of the data center, so that they can adjust or upgrade the reference solid-state drive based on the recommended parameter. The administrator received the recommended parameter and decided to adopt the recommendation and upgrade a batch of SSDs to 6.40T capacity, considering that the business requirements of the data center are high-performance and low-latency data processing. After the upgrade, the performance recovery time of the SSDs in the data center after the TRIM operation was significantly shortened, thereby improving the overall business processing speed and user satisfaction.

[0136] By adopting the embodiment of the present application, by obtaining the resource amount of the adjustment resource held by the reference account, the resource constraints that the user may face when actually deploying the SSD are taken into account, such as budget, space, energy consumption, etc. This mechanism allows finding the optimal capacity point under resource constraints, ensuring that the recommended SSD capacity not only meets the performance requirements, but is also feasible in the user's actual environment.

[0137] As an optional solution, after obtaining the difference between the current performance recovery time of the current solid state drive and the second time when the write operation is performed on the current solid state drive to obtain the current performance recovery time, the method further includes:

[0138] S1, based on the set of input / output operations per second corresponding to the current solid-state hard disk when the write operation is performed, determine the current stability level of the current solid-state hard disk, wherein the set of input / output operations per second includes multiple input / output operations per second recorded when the current solid-state hard disk is written, and the current stability level is used to characterize the stability of the current solid-state hard disk, and the higher the current stability level is, the stronger the stability of the current solid-state hard disk is.

[0139] Optionally, in some embodiments, the above-mentioned input / output operations per second set may be used, but not limited to, to indicate that during the process of performing a write operation on the solid-state drive, multiple IOPS values ​​are recorded, and the set of these values ​​reflects the read and write performance of the hard disk at different time points. The input / output operations per second set is specifically composed of multiple IOPS (input and output operations per second) parameters recorded when the hard disk is written, and is used to evaluate the read and write speed of the hard disk.

[0140] Furthermore, the above current stability level can be, but is not limited to, the stability level of the current steady-state solid-state hard disk that the system can determine based on the changes in the input / output operations per second set. The stability level can be determined by analyzing the fluctuations, average values, standard deviations and other statistical information of the input / output operations per second set, which characterizes the performance stability of the solid-state hard disk after the TRIM operation. The higher the stability level, the smaller the fluctuation of the hard disk during the performance recovery process, and the more stable the data processing.

[0141] S2, recording the correlation between the current stability level and the current performance recovery time in the third test result, wherein the third test result is used to characterize that the performance recovery time is proportional to the stability of the solid state drive.

[0142] It should be noted that, in some embodiments, the third test result collects a set of performance recovery time and stability level test results of H reference solid-state drives after TRIM operation, and records the correlation between each performance recovery time and each stability level.

[0143] S3: Send the third test result to the reference account, so that the reference account adjusts the reference solid state drive based on the third test result.

[0144] Optionally, the specific implementation of determining that the performance recovery time is proportional to the stability of the solid state drive may include but is not limited to referring to the following steps:

[0145] Data summary: The data from the third test results are summarized, including the performance recovery time and stability level (determined based on the set of input / output operations per second) collected from all H reference SSDs;

[0146] Correlation analysis: Based on the test results of H reference SSDs, the third correlation between performance recovery time and stability level is analyzed to confirm whether this relationship is positively proportional. That is, when the performance recovery time of the SSD after the TRIM operation increases, its stability level should also increase, indicating that there is a positive correlation between longer recovery time and more stable performance;

[0147] Statistical analysis: Apply statistical methods to quantify the strength of the association, such as calculating a correlation coefficient or using a linear regression model. If the statistical analysis shows a significant positive relationship, then this finding will be included in the third test results;

[0148] Construct the third test results: Write a test report, including data summary, correlation analysis results, and statistical analysis conclusions. The report should clearly point out that there is a positive relationship between performance recovery time and SSD stability, and may include charts, regression model equations, and correlation coefficients to intuitively demonstrate this relationship;

[0149] Result interpretation: Explain in detail in the report what this proportional relationship means, for example, why longer performance recovery time is associated with more stable SSD performance. Possible reasons include that longer post-TRIM recovery time allows the SSD to perform more thorough internal optimization, thereby improving the stability and consistency of data processing;

[0150] Recommendations and guidance: Based on the test results, the report should include recommendations and guidance for adjusting SSD configuration parameters. For example, if stable performance is required, it is recommended to select SSD types or capacity points that have longer recovery time after TRIM. The report may also recommend that system administrators or users adjust the capacity of the SSD based on their business needs and resource constraints to balance performance recovery time and stability.

[0151] Optionally, the reference account may adjust the configuration parameters of the reference solid state drive based on the third test result by, but is not limited to, the following steps:

[0152] Determine the performance recovery time target: Based on business needs, performance requirements, and cost budget, refer to the account to determine a target performance recovery time. For example, if the business requires a quick response, you may need to set a shorter performance recovery time target.

[0153] Identify key capacity points: Based on the third test results, identify which SSD capacity points can meet the performance recovery time target while having a high stability level. This may involve locating data points in the chart that correspond to the target performance recovery time, or finding combinations of performance recovery time and stability level through data analysis that meet business needs.

[0154] Evaluate configuration parameters: Considering the capacity, performance recovery time, and stability level of the SSD, refer to the account to analyze the current SSD configuration parameters, including but not limited to write cache size, garbage collection policy, read and write policy, and TRIM operation waiting time, to determine which parameters may need to be adjusted to optimize performance recovery time and stability.

[0155] Generate a configuration parameter adjustment plan: Develop a configuration parameter adjustment plan. For example, if the third test results show that increasing the write cache size can extend the performance recovery time and thus improve the stability level, then the plan may include increasing the write cache size. The plan should list in detail the parameters that need to be adjusted, the target value of the adjustment, and the expected performance impact.

[0156] Implement parameter adjustments: Implement configuration parameter adjustments in a real environment. This may involve modifying SSD firmware settings, storage management parameters in the operating system, or hardware configuration. Ensure that the adjustment process is safe and preferably performed during off-peak hours to reduce the impact on business.

[0157] Performance monitoring and verification: After adjusting configuration parameters, perform performance monitoring to ensure that changes in performance recovery time and stability level are in line with expectations. Use the same or similar tools and methods as the test to perform performance benchmarking to verify the effect of the adjustment.

[0158] Feedback and optimization: Based on the performance monitoring results, if it is found that the parameter adjustment does not achieve the expected effect or other performance problems occur, the reference account should feedback to the development or SSD vendor for more in-depth analysis and necessary optimization. This is a continuous iterative process, and multiple adjustments may be required to achieve the optimal configuration.

[0159] Optionally, as an optional example, the above steps may be explained based on, but not limited to, the following examples:

[0160] Assume that the third test result is as shown in Table 3:

[0161] Table 3

[0162]

[0163] Assume that the IOPS fluctuation of the above H candidate fixed hard disks is as follows Figure 2 , Figure 3 shown.

[0164] Referring to the third test result and the IO PS fluctuation chart above, we can draw the following conclusions:

[0165] The longer the SSD performance recovery time after TRIM and the greater the SSD performance fluctuation value, the more unstable the SSD performance is, indicating that there is still room for optimization of the reference fixed hard disk at the code level.

[0166] In some cases, such as when a request for performance evaluation of a reference fixed hard disk is received from a reference account, the reference fixed hard disk can be, but is not limited to, used to indicate the candidate fixed hard disk. In this embodiment, the above test process provides data support and actual measured value for development.

[0167] Optionally, as another optional example, the above steps may be explained by, but not limited to, the following example:

[0168] Assume that a series of SSDs are being tested, labeled 1.92T-X2, 3.84T-X2, 7.68T-X2, and 15.36T-X2, as well as P series hard drives of the same capacity point, for a total of H = 8 candidate SSDs.

[0169] Perform a low-level format and write operation to each hard disk to a steady state, then send a global TRIM command.

[0170] At different time points (eg, 0 minutes, 10 minutes, 30 minutes, 60 minutes) after the TRIM command, the IOPS parameter set is recorded.

[0171] Based on the IOPS parameter set, the current stability level of each hard disk is determined. For example, if the fluctuation of the IOPS parameter set is small and the average value is high, then the current stability level of the hard disk may be high.

[0172] The third correlation between the performance recovery time and the current stability level is calculated and recorded to form a third test result for each hard disk.

[0173] Continue testing until all candidate SSDs have been tested.

[0174] Analysis of the third test results revealed a positive correlation between the performance recovery time and the hard drive stability level.

[0175] Specific numerical examples:

[0176] After assuming the test, the following data is obtained:

[0177] For the 1.92T-X2 hard disk, the data is written immediately after waiting for 0 minutes after TRIM. The recorded IOPS parameter set fluctuates greatly, with an average value of 170k and a maximum value of 210k. The stability level is medium; the performance recovery time is 4099 seconds.

[0178] For a 1.92TP hard disk, with the same TRIM and 0-minute wait for writing, the recorded IOPS parameter set fluctuates slightly, with an average value of 230k, a maximum value of 690k, and a high stability level; the performance recovery time is 755 seconds.

[0179] It is assumed that hard disks of other capacity points are also subjected to similar tests, and the corresponding IO PS parameter sets and performance recovery time are recorded.

[0180] Generate the third test result:

[0181] After comparative analysis, it is found that the performance recovery time of hard disks with high stability levels is generally longer, which shows that there is a positive relationship between the performance recovery time and the stability level of the hard disk. For example, the stability level of P series hard disks is generally higher than that of X2 series, and the performance recovery time is also longer.

[0182] Send the third test result:

[0183] The third test results are sent to the reference account, and the reference account can adjust the configuration parameters of the reference solid-state drive based on these results, such as the garbage collection strategy, the wear leveling algorithm, or the read-write cache strategy, to optimize the hard drive performance and stability after the TRIM operation to meet specific application requirements.

[0184] By using an embodiment of the present application, the relationship between the performance recovery time and the stability level of the solid-state drive after the TRIM operation is analyzed to quantify the stability performance of the SSD after the TRIM operation. This helps to more comprehensively understand the performance of the SSD when facing the TRIM operation, rather than just focusing on the length of the recovery time. Based on the proportional relationship between the performance recovery time and the stability level, the optimization adjustment of the SSD configuration parameters can be guided. For example, if the performance recovery time after the TRIM operation is long but the stability level is high, this may mean that the SSD exhibits higher data consistency when processing the TRIM command, which helps to improve the reliability of the SSD during long-term use. By sending this analysis result to a reference account, the system administrator can adjust the configuration parameters of the SSD accordingly, such as the garbage collection strategy, wear leveling algorithm, etc., to improve the stability performance of the SSD.

[0185] In this embodiment, a parameter recommendation device is also provided, which is used to implement the above embodiments and preferred implementations, and will not be repeated here. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0186] Figure 4 is a structural block diagram of a parameter recommendation device according to an embodiment of the present application, such as Figure 4 As shown, the device comprises:

[0187] The first sending unit 402 is used to send a resource recovery command to the target solid state drive at the i-th time, wherein the resource recovery command is used to instruct to perform a recovery operation on the data block to be recovered included in the target solid state drive, and i is a positive integer and i is less than N;

[0188] The acquisition unit 404 is used to acquire the difference between the i-th performance recovery moment of the target solid state drive and the i-th write moment of executing the write operation on the target solid state drive, and obtain the i-th performance recovery duration, wherein the number of input / output operations per second of the target solid state drive at the i-th performance recovery moment is the target threshold;

[0189] A recording unit 406, configured to record the association between the i-th performance recovery duration and the i-th interval in the first test result, wherein the i-th interval is the interval between the i-th writing moment and the i-th moment;

[0190] The second sending unit 408 is used to generate a recommended interval parameter based on the first test result, and send the recommended interval parameter to the reference account, so that the reference account adjusts the reference solid-state hard disk based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state hard disk and performing a write operation on the reference solid-state hard disk.

[0191] As an optional scheme, the above-mentioned second sending unit includes: a first determination module, used to determine, based on the first test result, that the performance recovery time of the solid-state hard disk is inversely proportional to the time interval for performing a write operation after sending a resource recovery command to the solid-state hard disk; a first acquisition module, used to obtain first adjustment requirement information of the reference solid-state hard disk and the load level of the reference solid-state hard disk, wherein the load level is used to characterize the load condition of the reference solid-state hard disk; a second determination module, used to determine the recommended interval parameters based on the load level of the reference solid-state hard disk and the first test result when the first adjustment requirement information is used to indicate that the performance of the reference solid-state hard disk is to be improved.

[0192] As an optional solution, the above-mentioned device also includes: a formatting processing unit, used to format the candidate target solid-state hard disk; a first writing unit, used to perform a write operation on the formatted candidate solid-state hard disk until the formatted candidate solid-state hard disk reaches a steady state to obtain a target solid-state hard disk.

[0193] As an optional scheme, the above-mentioned first writing unit includes: a first writing module, used to perform j full-disk sequential write operations on the formatted candidate solid-state hard disk to obtain a first solid-state hard disk, where j is a positive integer; a second writing module, used to perform K hours of random write operations on the first solid-state hard disk to obtain a target solid-state hard disk, where K is a positive integer.

[0194] As an optional solution, the above-mentioned device also includes: a second writing unit, which is used to perform w hours of random write operations on the target solid-state hard disk at the i-th writing time, and record the number of input / output operations per second of the target solid-state hard disk during the random write operation, where w is a positive integer.

[0195] As an optional solution, the above-mentioned device also includes: a first determination unit, which is used to determine each reference solid-state hard disk among the H reference solid-state hard disks included in the reference solid-state hard disk set as the current solid-state hard disk and perform a test step, wherein the hard disk capacity of the H reference solid-state hard disks increases one by one; a second sending unit, which is used to send a resource recovery command to the current solid-state hard disk at a first moment, wherein the current solid-state hard disk, when receiving the resource recovery command, will perform a recovery operation on the data blocks to be recovered included in the current solid-state hard disk; a first acquisition unit, which is used to obtain the difference between the current performance recovery moment of the current solid-state hard disk and the second moment when the write operation is performed on the current solid-state hard disk, and obtain the current performance recovery duration, wherein at the current performance recovery moment, the number of input / output operations per second of the current solid-state hard disk reaches the reference threshold; a first recording unit, which is used to record the association between the storage capacity of the current solid-state hard disk and the current performance recovery duration in the second test result; a third sending unit, which is used to generate a recommended capacity parameter based on the second test result, and send the recommended capacity parameter to the reference account, so that the reference account adjusts the reference solid-state hard disk based on the recommended capacity parameter.

[0196] As an optional scheme, the above-mentioned third sending unit includes: a third determination module, used to determine, based on the second test result, that the performance recovery time is inversely proportional to the capacity of the solid-state hard drive; a second acquisition module, used to obtain the second adjustment requirement information of the reference solid-state hard drive and the resource quantity of the adjustment resources held by the reference account, wherein the adjustment resources are used to adjust the capacity of the reference solid-state hard drive; a fourth determination module, used to determine the recommended capacity parameters based on the resource quantity and the second test result when the second adjustment requirement information is used to indicate that the performance of the reference solid-state hard drive is to be improved.

[0197] As an optional solution, the above-mentioned device also includes: a second determination unit, used to determine the current stability level of the current solid-state hard disk based on the set of input / output operations per second corresponding to the current solid-state hard disk when a write operation is performed, wherein the set of input / output operations per second includes multiple input / output operations per second recorded when the current solid-state hard disk is written, and the current stability level is used to characterize the stability of the current solid-state hard disk, and the higher the current stability level, the stronger the stability of the current solid-state hard disk; a second recording unit, used to record the correlation between the current stability level and the current performance recovery time in the third test result, wherein the third test result is used to characterize that the performance recovery time is proportional to the stability of the solid-state hard disk; a fourth sending unit, used to send the third test result to the reference account, so that the reference account adjusts the reference solid-state hard disk based on the third test result.

[0198] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0199] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0200] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0201] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0202] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0203] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0204] In an exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0205] An embodiment of the present application further provides a computer program product, comprising a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program product, and when the computer program is executed by a processor, the steps of the method described in each embodiment of the present application are implemented.

[0206] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0207] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0208] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A parameter recommendation method, characterized in that: include: Sending a resource recovery command to the target solid-state drive at the i-th time, wherein the resource recovery command is used to instruct to perform a recovery operation on the data block to be recovered included in the target solid-state drive, and i is a positive integer and i is less than N; Obtaining a difference between an i-th performance recovery moment of the target solid-state hard disk and an i-th write moment of executing a write operation on the target solid-state hard disk, to obtain an i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state hard disk at the i-th performance recovery moment is a target threshold; Recording the correlation between the i-th performance recovery duration and the i-th interval in the first test result, wherein the i-th interval is the interval between the i-th writing moment and the i-th moment; Generate a recommended interval parameter based on the first test result, and send the recommended interval parameter to a reference account, so that the reference account adjusts the reference solid-state hard drive based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state hard drive and performing a write operation on the reference solid-state hard drive.

2. The parameter recommendation method according to claim 1, characterized in that: The generating a recommended time interval parameter based on the first test result comprises: Determining, based on the first test result, that the performance recovery time of the solid-state drive is inversely proportional to the time interval between executing the write operation after sending the resource recovery command to the solid-state drive; Acquire first adjustment requirement information of the reference solid state drive and a load level of the reference solid state drive, wherein the load level is used to characterize a load condition of the reference solid state drive; In a case where the first adjustment requirement information is used to indicate that the performance of the reference solid state drive is to be improved, the recommended interval parameter is determined based on the load level of the reference solid state drive and the first test result.

3. The parameter recommendation method according to claim 1, characterized in that: Before sending the resource recovery command to the target solid state drive at the i-th time, the method further includes: Format the candidate target solid state drive; A write operation is performed on the formatted candidate solid-state hard disk until the formatted candidate solid-state hard disk reaches a steady state, thereby obtaining the target solid-state hard disk.

4. The parameter recommendation method according to claim 3, characterized in that: Performing a write operation on the formatted candidate solid-state hard disk until the formatted candidate solid-state hard disk reaches a steady state, and obtaining the target solid-state hard disk includes: Performing j full-disk sequential write operations on the formatted candidate solid-state hard disk to obtain a first solid-state hard disk, where j is a positive integer; Performing K hours of random write operations on the first solid-state drive to obtain the target solid-state drive, where K is a positive integer.

5. The parameter recommendation method according to any one of claims 1 to 4, characterized in that: After sending the resource recovery command to the target solid state drive at the i-th time, the method further includes: At the i-th writing moment, a random write operation is performed on the target solid state drive for w hours, and the number of input / output operations per second of the target solid state drive during the random write operation is recorded, where w is a positive integer.

6. The parameter recommendation method according to any one of claims 1 to 4, characterized in that: The recommended method of the parameters also includes: Determine each of the H reference solid-state hard disks included in the reference solid-state hard disk set as a current solid-state hard disk and perform the following steps, wherein the hard disk capacities of the H reference solid-state hard disks are increased one by one: Sending the resource recovery command to the current solid state drive at a first moment, wherein the current solid state drive, upon receiving the resource recovery command, will perform a recovery operation on the data blocks to be recovered included in the current solid state drive; Obtaining a difference between a current performance recovery moment of the current solid-state hard disk and a second moment at which a write operation is performed on the current solid-state hard disk, to obtain a current performance recovery duration, wherein at the current performance recovery moment, the number of input / output operations per second of the current solid-state hard disk reaches a reference threshold; Recording the correlation between the current storage capacity of the solid-state hard disk and the current performance recovery time in the second test result; Generate a recommended capacity parameter based on the second test result, and send the recommended capacity parameter to the reference account, so that the reference account adjusts the reference solid-state drive based on the recommended capacity parameter.

7. The parameter recommendation method according to claim 6, characterized in that: Generating the recommended capacity parameter based on the second test result includes: Determining, based on the second test result, that the performance recovery time is inversely proportional to the capacity of the solid-state drive; Acquire second adjustment requirement information of the reference solid-state hard disk and the resource amount of adjustment resources held by the reference account, wherein the adjustment resources are used to adjust the capacity of the reference solid-state hard disk; In a case where the second adjustment requirement information is used to indicate that the performance of the reference solid state drive is to be improved, the recommended capacity parameter is determined based on the resource amount and the second test result.

8. The parameter recommendation method according to claim 6, characterized in that: After obtaining the difference between the current performance recovery time of the current solid state hard disk and the second time of performing a write operation on the current solid state hard disk to obtain the current performance recovery time, the method further includes: Determine a current stability level of the current solid-state hard disk based on a set of input / output operations per second corresponding to the current solid-state hard disk when a write operation is performed, wherein the set of input / output operations per second includes a plurality of input / output operations per second recorded when the current solid-state hard disk is written, and the current stability level is used to characterize the stability of the current solid-state hard disk, and a higher current stability level indicates a stronger stability of the current solid-state hard disk; Recording the correlation between the current stability level and the current performance recovery time in a third test result, wherein the third test result is used to indicate that the performance recovery time is proportional to the stability of the solid state drive; The third test result is sent to the reference account, so that the reference account adjusts the reference solid state drive based on the third test result.

9. A parameter recommendation device, characterized in that: include: A first sending unit is used to send a resource recovery command to a target solid state drive at an i-th time, wherein the resource recovery command is used to instruct to perform a recovery operation on a data block to be recovered included in the target solid state drive, and i is a positive integer and i is less than N; an acquisition unit, configured to acquire a difference between an i-th performance recovery moment of the target solid-state hard disk and an i-th write moment of executing a write operation on the target solid-state hard disk, to obtain an i-th performance recovery duration, wherein the number of input / output operations per second of the target solid-state hard disk at the i-th performance recovery moment is a target threshold; a recording unit, configured to record the association between the i-th performance recovery duration and the i-th interval in a first test result, wherein the i-th interval is an interval between the i-th writing moment and the i-th moment; A second sending unit is used to generate a recommended interval parameter based on the first test result, and send the recommended interval parameter to a reference account, so that the reference account adjusts the reference solid-state hard disk based on the recommended interval parameter, wherein the recommended interval parameter is used to characterize the time interval between sending a resource recovery command to the reference solid-state hard disk and performing a write operation on the reference solid-state hard disk.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program implements the steps of the method according to any one of claims 1 to 8 when executed by a processor.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

12. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images