A hybrid read-write performance diagnosis method, device and medium

Abstract

The application discloses a mixed read-write performance diagnosis method and device and a medium. Through running a test script on a client and analyzing OPS statistics and metadata operation time delay statistics, in the case of detecting OPS fluctuation or metadata operation time delay mutation, it is judged on a server whether corresponding OPS fluctuation or time fluctuation exists. If no corresponding fluctuation is detected on the server, it is confirmed that a performance bottleneck point exists in metadata operation. If corresponding fluctuation exists on the server, it is confirmed that a performance bottleneck point exists in data operation. By adopting the technical scheme, OPS statistics and metadata operation time delay statistics are respectively performed on the client and the server, whether corresponding regular fluctuation exists on the server is detected after fluctuation is detected on the client, and the position of the performance bottleneck point is confirmed. When corresponding fluctuation exists, it is confirmed that a performance bottleneck point exists in metadata operation, and if not, it is confirmed that a performance bottleneck point exists in data operation.

Description

Technical Field

[0001] This application relates to the field of storage testing technology, and in particular to a method, apparatus and medium for diagnosing mixed read and write performance. Background Technology

[0002] With technological innovations in computing, networking, and storage, as well as the diversification of business types, data centers have undergone a long evolution from mainframes to x86 servers, and from virtualization technology to cloud computing, where everything is a service. Storage technology, in order to meet the massive data storage demands of ever-evolving applications, has also undergone a radical transformation, from physical media to technical architecture. Regardless of how technology advances, the goal remains the same: to better provide high-performance, high-capacity, and highly available data services. High performance is a crucial factor in IT decision-making, and a good solution can save both money and time.

[0003] However, in a file business model that is a hybrid read-write model, a large number of files are read and written concurrently, resulting in unsatisfactory business performance. It is necessary to identify the performance bottleneck and perform targeted performance tuning.

[0004] Therefore, it is evident that identifying performance bottlenecks in hybrid read / write business models is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, and medium for diagnosing hybrid read / write performance, used to locate performance bottlenecks in hybrid read / write business models.

[0006] To address the aforementioned technical problems, this application provides a hybrid read / write performance diagnostic method, applied to a client, comprising:

[0007] Run the test script and analyze OPS statistics and metadata operation latency statistics;

[0008] If OPS fluctuations or sudden changes in metadata operation latency are detected, the server determines whether there are corresponding OPS fluctuations or time fluctuations.

[0009] If no corresponding OPS fluctuations or time fluctuations are detected on the server side, it is confirmed that there is a performance bottleneck in the metadata operation.

[0010] If there are corresponding fluctuations in OPS or time on the server side, it confirms that there is a performance bottleneck in data operations.

[0011] Preferably, if a performance bottleneck is confirmed in metadata operations, the following measures are also included:

[0012] The percentage of time spent on client locks is statistically analyzed. If the client lock time reaches a first threshold and the percentage reaches a second threshold, then the performance bottleneck is identified as client lock conflict.

[0013] Preferably, if the performance bottleneck is identified as client lock contention, the number of clients should be increased to reduce client lock contention.

[0014] Preferably, if a performance bottleneck is confirmed in the data operation, the following measures are also included:

[0015] By analyzing the network processing time and the underlying processing time, if the total processing time reaches the third threshold, it is confirmed that there is a performance bottleneck in the underlying disk.

[0016] Preferably, before the step of running the test script and analyzing OPS statistics and metadata operation latency statistics, the method further includes:

[0017] Based on performance statistics analysis of the latency of each operation word, the operation words include metadata operation words and data operation words.

[0018] To address the aforementioned technical issues, this application also provides another hybrid read / write performance diagnostic method, applied to the server side, including:

[0019] Run test scripts on the client and analyze OPS statistics and metadata operation latency statistics. If OPS fluctuations or metadata operation latency changes are detected, determine whether there are corresponding OPS fluctuations or time fluctuations.

[0020] If no corresponding OPS fluctuations or time fluctuations are detected, it is confirmed that there is a performance bottleneck in the metadata operation.

[0021] If there are corresponding OPS fluctuations or time fluctuations, it confirms that there is a performance bottleneck in data operations.

[0022] To address the aforementioned technical problems, this application also provides a hybrid read / write performance diagnostic device, applied to a client, comprising:

[0023] The analysis module is used to run test scripts and analyze OPS statistics and metadata operation latency statistics;

[0024] The detection module is used to determine on the server side whether there is a corresponding OPS fluctuation or time fluctuation if an OPS fluctuation or metadata operation latency change is detected.

[0025] The processing module is used to confirm that there is a performance bottleneck in metadata operations if no corresponding OPS fluctuation or time fluctuation is detected on the server side; and to confirm that there is a performance bottleneck in data operations if there is corresponding OPS fluctuation or time fluctuation on the server side.

[0026] Preferably, it also includes: a first statistics module, used to count the proportion of client lock time consumption; if the client lock time consumption reaches a first threshold and the proportion reaches a second threshold, then the performance bottleneck is confirmed as client lock conflict.

[0027] Preferably, it also includes: a second statistics module, used to count network time and underlying processing time; if the total time reaches a third threshold, it is confirmed that the underlying disk has a performance bottleneck.

[0028] Preferably, it further includes: a statistics module, used to perform performance statistical analysis on the latency of each operation word, wherein the operation words include metadata operation words and data operation words.

[0029] To address the aforementioned technical problems, this application also provides another hybrid read / write performance diagnostic device, applied to the server side, comprising:

[0030] The judgment module is used to run test scripts on the client and analyze OPS statistics and metadata operation latency statistics. When OPS fluctuations or metadata operation latency changes are detected, it determines whether there are corresponding OPS fluctuations or time fluctuations.

[0031] The confirmation module is used to confirm that there is a performance bottleneck in metadata operations if no corresponding OPS fluctuation or time fluctuation is detected; and to confirm that there is a performance bottleneck in data operations if corresponding OPS fluctuation or time fluctuation is detected.

[0032] To address the aforementioned technical problems, this application also provides another hybrid read / write performance diagnostic device, including a memory for storing computer programs;

[0033] A processor, used to implement the steps of the hybrid read / write performance diagnostic method described above when executing the computer program.

[0034] To address the aforementioned technical problems, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the hybrid read / write performance diagnostic method described above.

[0035] The hybrid read / write performance diagnostic method provided in this application is applied to the client side. By running test scripts and analyzing OPS statistics and metadata operation latency statistics, when OPS fluctuations or sudden changes in metadata operation latency are detected, the server side determines whether corresponding OPS fluctuations or time fluctuations exist. If no corresponding OPS fluctuations or time fluctuations are detected on the server side, a performance bottleneck in metadata operations is confirmed; if corresponding OPS fluctuations or time fluctuations exist on the server side, a performance bottleneck in data operations is confirmed. This technical solution performs OPS statistics and metadata operation latency statistics on both the client and server sides. When fluctuations are detected on the client side, the server side checks for corresponding regular fluctuations to identify the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0036] Furthermore, the hybrid read / write performance diagnostic device and medium provided in this application correspond to the hybrid read / write performance diagnostic method described above, and have the same effect. Attached Figure Description

[0037] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 A flowchart illustrating a hybrid read / write performance diagnostic method provided in this application embodiment;

[0039] Figure 2 A flowchart of another hybrid read / write performance diagnosis method provided in this application embodiment:

[0040] Figure 3 A structural diagram of a hybrid read / write performance diagnostic device provided in an embodiment of this application;

[0041] Figure 4 A structural diagram of another hybrid read / write performance diagnostic device provided in an embodiment of this application;

[0042] Figure 5 This is a structural diagram of another hybrid read / write performance diagnostic device provided in an embodiment of this application. Detailed Implementation

[0043] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0044] With technological innovations in computing, networking, and storage, as well as the diversification of business types, data centers have undergone a long evolution from mainframes to x86 servers, and from virtualization technology to cloud computing, where everything is a service. Storage technology, in order to meet the massive data storage demands of ever-evolving applications, has also undergone a radical transformation, from physical media to technical architecture. Regardless of how technology advances, the goal remains the same: to better provide high-performance, high-capacity, and highly available data services. High performance is a crucial factor in IT decision-making, and a good solution can save both money and time.

[0045] Current technologies typically use vdbench for I / O performance testing. vdbench is an I / O workload generator, commonly used to verify data integrity and measure the performance of directly attached (or network-connected) storage. It can run on Windows and Linux environments and can be used to test file system or block device benchmark performance. Different business models have a significant impact on storage system performance and are therefore crucial performance considerations.

[0046] However, in a file business model that is a hybrid read-write model, a large number of files are read and written concurrently, resulting in unsatisfactory business performance. It is necessary to identify the performance bottleneck and perform targeted performance tuning.

[0047] Therefore, it is evident that identifying performance bottlenecks in hybrid read / write business models is a problem that urgently needs to be solved by those skilled in the art.

[0048] The core of this application is to provide a hybrid read / write performance diagnosis method, device, and medium for locating performance bottlenecks in hybrid read / write business models.

[0049] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] Figure 1 A flowchart illustrating a hybrid read / write performance diagnostic method provided in this application embodiment, applied to a client, such as... Figure 1 As shown, the method includes:

[0051] S10: Run the test script and analyze OPS statistics and metadata operation latency statistics;

[0052] S11: If an OPS fluctuation or a sudden change in metadata operation latency is detected, determine on the server whether there is a corresponding OPS fluctuation or time fluctuation.

[0053] S12: If no corresponding OPS fluctuation or time fluctuation is detected on the server side, then it is confirmed that there is a performance bottleneck in the metadata operation.

[0054] S13: If there are corresponding OPS fluctuations or time fluctuations on the server side, it is confirmed that there is a performance bottleneck in data operations.

[0055] Optical Performance Per Second (OPS) is a typical performance metric in the storage world, used to indicate how many requests can be processed per unit of time. OPS is usually considered in conjunction with latency: for the same latency, a higher OPS indicates a stronger storage system processing capability. However, simply achieving a high OPS but also having high latency does not necessarily indicate better storage system performance.

[0056] Latency refers to the time it takes from initiating an I / O request to receiving a response, including service time and waiting time. In most cases, latency is the single most critical metric reflecting the performance of a storage system.

[0057] In practical implementation, before running the test script and analyzing OPS statistics and metadata operation latency statistics, the following may also be included: analyzing the latency of each operation word based on performance statistics. Operation words include metadata operation words and data operation words. Specific metadata operation words include: readdir, lookup, setattr, getattr, open, release, etc.; specific data operation words include read, write, sync, etc.

[0058] In this embodiment, the first step is to analyze the ops and latency statistics of the vdbench script on the client side, focusing on ops fluctuations and specific metadata operation latency (i.e., the script's open and close operations). When ops fluctuations are detected, or when a sudden increase in open or close latency is observed, the ops fluctuations actually indicate latency instability. The analysis then shifts to the server-side metadata operation performance statistics. If client-side open performance fluctuates, but ops performance statistics are not printed every few seconds, it indicates that the server-side open latency will not fluctuate accordingly, confirming a performance bottleneck in metadata operations. Otherwise, if corresponding OPS or time fluctuations are found on the server side, a performance bottleneck in data operations is confirmed.

[0059] The hybrid read / write performance diagnostic method provided in this application is applied to the client side. By running test scripts and analyzing OPS statistics and metadata operation latency statistics, when OPS fluctuations or sudden changes in metadata operation latency are detected, the server side determines whether corresponding OPS fluctuations or time fluctuations exist. If no corresponding OPS fluctuations or time fluctuations are detected on the server side, a performance bottleneck in metadata operations is confirmed; if corresponding OPS fluctuations or time fluctuations exist on the server side, a performance bottleneck in data operations is confirmed. This technical solution performs OPS statistics and metadata operation latency statistics on both the client and server sides. When fluctuations are detected on the client side, the server side checks for corresponding regular fluctuations to identify the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0060] Furthermore, when there are no corresponding fluctuations on the server side, it is also necessary to determine whether the long processing time is caused by client locks. In specific implementations, in a clustered environment, when a user makes multiple requests to the interface, the stored content may be duplicated. In this case, distributed locks can be used to restrict this. However, client locks can lead to long processing times and impact business performance. Therefore, based on the above embodiments, this embodiment, when a performance bottleneck in metadata operations is confirmed, also includes:

[0061] The performance bottleneck is identified as client lock contention if the client lock contention reaches the first threshold and the percentage reaches the second threshold.

[0062] In practice, if the performance bottleneck is identified as client lock contention, the number of clients can be increased to reduce client lock contention.

[0063] This embodiment determines whether client lock conflicts are a performance bottleneck by statistically analyzing the proportion of client lock time consumed when a performance bottleneck is confirmed in the metadata operation.

[0064] Based on the above embodiments, this embodiment further includes, when it is confirmed that there is a performance bottleneck in the data operation:

[0065] By analyzing network time and underlying processing time, if the total time percentage reaches the third threshold, it is confirmed that there is a performance bottleneck in the underlying disk.

[0066] The hybrid read / write performance diagnosis method provided in this embodiment statistically analyzes the fluctuation patterns of the underlying module disk. When the fluctuation meets the third threshold, it is necessary to continue analyzing the cause of the fluctuation and finally locate the fluctuation problem as being caused by the disk FC mechanism, thus realizing the judgment of the underlying module.

[0067] Figure 2 A flowchart of another hybrid read / write performance diagnosis method provided in this application embodiment, applied to the server, includes:

[0068] S20: Run the test script on the client and analyze the OPS statistics and metadata operation latency statistics. If OPS fluctuations or metadata operation latency changes are detected, determine whether there are corresponding OPS fluctuations or time fluctuations.

[0069] S21: If no corresponding OPS fluctuation or time fluctuation is detected, it is confirmed that there is a performance bottleneck in the metadata operation;

[0070] S22: If there are corresponding OPS fluctuations or time fluctuations, it is confirmed that there is a performance bottleneck in data operations.

[0071] The hybrid read / write performance diagnosis method for the server provided in this embodiment corresponds to the hybrid read / write performance diagnosis method for the client described above. The above embodiment describes the hybrid read / write performance diagnosis method for the server in detail, so this embodiment will not repeat it.

[0072] The hybrid read / write performance diagnostic method provided in this application is applied to the server side. By running test scripts on the client side and analyzing OPS statistics and metadata operation latency statistics, it determines whether corresponding OPS fluctuations or time fluctuations exist when OPS fluctuations or sudden changes in metadata operation latency are detected. If no corresponding OPS fluctuations or time fluctuations are detected, a performance bottleneck in metadata operations is confirmed; if corresponding OPS fluctuations or time fluctuations exist, a performance bottleneck in data operations is confirmed. This technical solution performs OPS statistics and metadata operation latency statistics on both the client and server sides. When fluctuations are detected on the client side, the server side checks for corresponding regular fluctuations to identify the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0073] The above embodiments have described the hybrid read / write performance diagnosis method in detail. This application also provides embodiments corresponding to the hybrid read / write performance diagnosis device. It should be noted that this application describes the device embodiments from two perspectives: one based on functional modules and the other based on hardware.

[0074] Figure 3This application provides a structural diagram of a hybrid read / write performance diagnostic device, applied to a client, such as... Figure 3 As shown, the device includes:

[0075] Analysis module 01 is used to run test scripts and analyze OPS statistics and metadata operation latency statistics;

[0076] The detection module 02 is used to determine on the server whether there is a corresponding OPS fluctuation or time fluctuation if an OPS fluctuation or metadata operation delay change is detected.

[0077] Processing module 03 is used to confirm that there is a performance bottleneck in metadata operations if no corresponding OPS fluctuation or time fluctuation is detected on the server side; and to confirm that there is a performance bottleneck in data operations if there is corresponding OPS fluctuation or time fluctuation on the server side.

[0078] It should be noted that, Figure 3 The structure shown does not constitute a limitation on the hybrid read / write performance diagnostic device. Preferably, the hybrid read / write performance diagnostic device may further include: a first statistics module, used to count the proportion of client lock time consumption. If the client lock time consumption reaches a first threshold and the proportion reaches a second threshold, then the performance bottleneck is confirmed as client lock conflict.

[0079] Preferably, it also includes: a second statistics module, used to count network time and underlying processing time. If the total time reaches a third threshold, it is confirmed that there is a performance bottleneck in the underlying disk.

[0080] Preferably, it also includes: a statistics module, used to analyze the latency of each operation word based on performance statistics, the operation words including metadata operation words and data operation words.

[0081] Since the embodiments of the apparatus and the embodiments of the method correspond to each other, please refer to the description of the embodiments of the method for the embodiments of the apparatus, which will not be repeated here.

[0082] The hybrid read / write performance diagnostic device provided in this application is applied to the client side. By running test scripts and analyzing OPS statistics and metadata operation latency statistics, when OPS fluctuations or sudden changes in metadata operation latency are detected, the device determines on the server side whether corresponding OPS fluctuations or time fluctuations exist. If no corresponding OPS fluctuations or time fluctuations are detected on the server side, a performance bottleneck in metadata operations is confirmed; if corresponding OPS fluctuations or time fluctuations exist on the server side, a performance bottleneck in data operations is confirmed. This technical solution performs OPS statistics and metadata operation latency statistics on both the client and server sides. When fluctuations are detected on the client side, the device checks on the server side for corresponding regular fluctuations to identify the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0083] Figure 4 A structural diagram of another hybrid read / write performance diagnostic device provided in this application embodiment, applied to a server, the device includes:

[0084] The judgment module 10 is used to run test scripts on the client and analyze OPS statistics and metadata operation latency statistics. If OPS fluctuations or metadata operation latency changes are detected, it is determined whether there are corresponding OPS fluctuations or time fluctuations.

[0085] The confirmation module 11 is used to confirm that there is a performance bottleneck in the metadata operation if no corresponding OPS fluctuation or time fluctuation is detected; and to confirm that there is a performance bottleneck in the data operation if there is a corresponding OPS fluctuation or time fluctuation.

[0086] Since the embodiments of the apparatus and the embodiments of the method correspond to each other, please refer to the description of the embodiments of the method for the embodiments of the apparatus, which will not be repeated here.

[0087] The hybrid read / write performance diagnostic device provided in this application is applied to the server side. By running test scripts on the client side and analyzing OPS statistics and metadata operation latency statistics, it detects OPS fluctuations or sudden changes in metadata operation latency. If no corresponding OPS or time fluctuations are detected, a performance bottleneck in metadata operations is confirmed; if corresponding OPS or time fluctuations exist, a performance bottleneck in data operations is confirmed. This technical solution performs OPS statistics and metadata operation latency statistics on both the client and server sides. When fluctuations are detected on the client side, the server side checks for corresponding regular fluctuations to identify the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0088] Figure 5 A structural diagram of another hybrid read / write performance diagnostic device provided in this application embodiment is shown below. Figure 5 As shown, the device includes: a memory 20 for storing computer programs;

[0089] The processor 21 is used to implement the steps of the hybrid read / write performance diagnosis method as described in the above embodiments when executing a computer program.

[0090] The hybrid read / write performance diagnostic device provided in this embodiment may include, but is not limited to, smartphones, tablets, laptops, or desktop computers.

[0091] The processor 21 may include one or more processing cores, such as a quad-core processor or an octa-core processor. The processor 21 may be implemented using at least one of the following hardware forms: Digital Signal Processor (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 21 may also include a main processor and a coprocessor. The main processor, also known as the Central Processing Unit (CPU), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, the processor 21 may integrate a Graphics Processing Unit (GPU), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, the processor 21 may also include an Artificial Intelligence (AI) processor, which is used to handle computational operations related to machine learning.

[0092] The memory 20 may include one or more computer-readable storage media, which may be non-transitory. The memory 20 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In this embodiment, the memory 20 is used to store at least the following computer program 201, which, after being loaded and executed by the processor 21, is capable of implementing the relevant steps of the read / write performance diagnosis method disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 20 may also include an operating system 202 and data 203, and the storage method may be temporary or permanent storage. The operating system 202 may include Windows, Unix, Linux, etc. The data 203 may include, but is not limited to, OPS statistics and metadata operation latency statistics.

[0093] In some embodiments, the read / write performance diagnostic device may further include a display screen 22, an input / output interface 23, a communication interface 24, a power supply 25, and a communication bus 26.

[0094] Those skilled in the art will understand that Figure 5 The structure shown does not constitute a limitation on the read / write performance diagnostic device and may include more or fewer components than shown.

[0095] The hybrid read / write performance diagnostic device provided in this application includes a memory and a processor. When the processor executes a program stored in the memory, it can implement the following method: hybrid read / write performance diagnostic method.

[0096] The hybrid read / write performance diagnostic device provided in this application runs a test script on the client and analyzes OPS statistics and metadata operation latency statistics. When OPS fluctuations or sudden changes in metadata operation latency are detected, the device determines on the server whether corresponding OPS or time fluctuations exist. If no corresponding OPS or time fluctuations are detected on the server, a performance bottleneck in metadata operations is confirmed; if corresponding OPS or time fluctuations exist on the server, a performance bottleneck in data operations is confirmed. This technical solution performs OPS statistics and metadata operation latency statistics on both the client and server sides. When fluctuations are detected on the client side, the device checks on the server side for corresponding regular fluctuations to identify the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0097] Finally, this application also provides an embodiment corresponding to a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps described in the above method embodiments (which may be a method corresponding to the client side, a method corresponding to the server side, or a method corresponding to both the client side and the server side).

[0098] It is understood that if the methods in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and executes all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0099] The computer-readable storage medium provided in this application embodiment, by running test scripts on the client and analyzing OPS statistics and metadata operation latency statistics, determines on the server whether there are corresponding OPS fluctuations or time fluctuations when OPS fluctuations or sudden changes in metadata operation latency are detected; if no corresponding OPS fluctuations or time fluctuations are detected on the server, a performance bottleneck in metadata operations is confirmed; if corresponding OPS fluctuations or time fluctuations are detected on the server, a performance bottleneck in data operations is confirmed. This technical solution, by performing OPS statistics and metadata operation latency statistics on both the client and server sides, detects fluctuations on the client side and checks on the server side for corresponding regular fluctuations to confirm the location of the performance bottleneck. When corresponding fluctuations exist, a performance bottleneck in metadata operations is confirmed; if not, a performance bottleneck in data operations is confirmed.

[0100] The above provides a detailed description of the hybrid read / write performance diagnostic method, apparatus, and medium provided in this application. The various embodiments in the specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of the claims of this application.

[0101] It should also be noted that, in this specification, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

Claims

1. A method for diagnosing hybrid read / write performance, characterized in that, Applied to the client side, including: Run the test script and analyze OPS statistics and metadata operation latency statistics; If OPS fluctuations or sudden changes in metadata operation latency are detected, the server determines whether there are corresponding OPS fluctuations or time fluctuations. If no corresponding OPS fluctuation or time fluctuation is detected on the server side, it is confirmed that there is a performance bottleneck in the metadata operation; the client lock time consumption ratio is counted. If the client lock time consumption reaches the first threshold and the ratio reaches the second threshold, the performance bottleneck is confirmed to be client lock conflict. If there are corresponding OPS fluctuations or time fluctuations on the server side, it is confirmed that there is a performance bottleneck in data operations; count the network time and underlying processing time, and if the total time reaches the third threshold, it is confirmed that there is a performance bottleneck in the underlying disk.

2. The hybrid read / write performance diagnostic method according to claim 1, characterized in that, If the performance bottleneck is identified as client lock contention, the number of clients can be increased to reduce client lock contention.

3. The hybrid read / write performance diagnostic method according to claim 1, characterized in that, Before the steps of running the test script and analyzing OPS statistics and metadata operation latency statistics, the following is also included: Based on performance statistics analysis of the latency of each operation word, the operation words include metadata operation words and data operation words.

4. A method for diagnosing hybrid read / write performance, characterized in that, Applied to the server side, including: Run test scripts on the client and analyze OPS statistics and metadata operation latency statistics. If OPS fluctuations or metadata operation latency changes are detected, determine whether there are corresponding OPS fluctuations or time fluctuations. If no corresponding OPS fluctuation or time fluctuation is detected, it is confirmed that there is a performance bottleneck in the metadata operation; the client lock time consumption ratio is counted on the client side. If the client lock time consumption reaches the first threshold and the ratio reaches the second threshold, it is confirmed that the performance bottleneck is client lock conflict. If there are corresponding OPS fluctuations or time fluctuations, it confirms that there is a performance bottleneck in data operations; if the total time spent on network and underlying processing is counted on the client side and the percentage of total time spent reaches the third threshold, it confirms that there is a performance bottleneck in the underlying disk.

5. A hybrid read / write performance diagnostic device, characterized in that, Applied to the client side, including: The analysis module is used to run test scripts and analyze OPS statistics and metadata operation latency statistics; The detection module is used to determine on the server side whether there is a corresponding OPS fluctuation or time fluctuation when an OPS fluctuation or metadata operation latency change is detected. The first statistics module is used to count the percentage of client lock time. If the client lock time reaches the first threshold and the percentage reaches the second threshold, then the performance bottleneck is confirmed to be client lock conflict. The processing module is used to confirm that there is a performance bottleneck in metadata operations if no corresponding OPS fluctuation or time fluctuation is detected on the server side; and to confirm that there is a performance bottleneck in data operations if there is corresponding OPS fluctuation or time fluctuation on the server side. The second statistics module is used to count network time and underlying processing time. If the total time reaches the third threshold, it is confirmed that there is a performance bottleneck in the underlying disk.

6. A hybrid read / write performance diagnostic device, characterized in that, Applied to the server side, including: The judgment module is used to run test scripts on the client and analyze OPS statistics and metadata operation latency statistics. When OPS fluctuations or metadata operation latency changes are detected, it determines whether there are corresponding OPS fluctuations or time fluctuations. The confirmation module is used to confirm that there is a performance bottleneck in metadata operations if no corresponding OPS fluctuation or time fluctuation is detected; to count the proportion of client lock consumption on the client side, and if the client lock consumption reaches a first threshold and the proportion reaches a second threshold, then the performance bottleneck is confirmed to be client lock conflict; if there is a corresponding OPS fluctuation or time fluctuation, then the data operation has a performance bottleneck; and to count the network consumption time and underlying processing time on the client side, and if the total consumption time proportion reaches a third threshold, then the underlying disk has a performance bottleneck.

7. A hybrid read / write performance diagnostic device, characterized in that, Includes memory used to store computer programs; A processor, configured to implement the steps of the hybrid read / write performance diagnostic method as described in any one of claims 1 to 4 when executing the computer program.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the hybrid read / write performance diagnostic method as described in any one of claims 1 to 4.

Images