Centralized storage performance evaluation method, device and electronic equipment

Abstract

The present disclosure provides a centralized storage performance evaluation method, device and electronic device, which relate to the field of computer technology. By first analyzing the performance impact factor set of the centralized storage system to be evaluated in the target deployment environment and the performance impact value corresponding to at least one performance impact factor in the performance impact factor set, the performance impact data of the centralized storage system to be evaluated in the target deployment environment is obtained; then, when the target deployment environment sample corresponding to the performance impact data is matched in the performance sample library, the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment is determined according to the target system operation performance data corresponding to the target deployment environment sample. Therefore, it can solve the technical problem that there may be a large difference between the performance baseline of the storage product and the actual deployment environment of the customer, resulting in a large error in the performance prediction and the inability to provide an accurate performance reference, and can improve the accuracy of the centralized storage performance evaluation.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a centralized storage performance evaluation method, device, and electronic device. Background Art

[0002] Performance evaluation of centralized storage systems is crucial for optimizing system performance, ensuring business continuity, ensuring data security, and supporting technical maintenance. An efficient and reliable performance evaluation method can not only further confirm whether a centralized storage system meets customer business requirements, but also serve as a key indicator for storage solution project delivery, assisting in troubleshooting delivery and implementation issues. Related technologies provide performance predictions based on the performance baseline of centralized storage systems' storage products. However, this performance baseline can differ significantly from the customer's actual deployment environment, resulting in significant errors in performance predictions and an inability to provide an accurate performance reference. Summary of the Invention

[0003] The present disclosure provides a centralized storage performance evaluation method, device, and electronic device, the main purpose of which is to solve the problem of low accuracy of centralized storage performance evaluation.

[0004] According to a first aspect of the present disclosure, a centralized storage performance evaluation method is provided, comprising:

[0005] Determining performance impact data of the centralized storage system to be evaluated in the target deployment environment, wherein the performance impact data includes a set of performance impact factors and a performance impact value corresponding to at least one performance impact factor in the set of performance impact factors;

[0006] Matching the performance impact data in a performance sample library, wherein the performance sample library includes a plurality of deployment environment samples, and sample performance impact data and system operation performance data corresponding to at least one deployment environment sample among the plurality of deployment environment samples;

[0007] When there is a target deployment environment sample matching the performance impact data in the performance sample library, the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment is determined based on the target system operation performance data corresponding to the target deployment environment sample.

[0008] According to a second aspect of the present disclosure, a centralized storage performance evaluation device is provided, comprising:

[0009] a data determining unit, configured to determine performance impact data of the centralized storage system to be evaluated in a target deployment environment, wherein the performance impact data includes a set of performance impact factors and a performance impact value corresponding to at least one performance impact factor in the set of performance impact factors;

[0010] a sample matching unit, configured to match performance impact data in a performance sample library, wherein the performance sample library includes a plurality of deployment environment samples, and sample performance impact data and system operation performance data corresponding to at least one deployment environment sample among the plurality of deployment environment samples;

[0011] The performance evaluation unit is used to determine the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment according to the target system operation performance data corresponding to the target deployment environment sample when there is a target deployment environment sample matching the performance impact data in the performance sample library.

[0012] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect.

[0016] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute the method described in the first aspect.

[0017] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method as described in the first aspect above.

[0018] The present invention obtains the performance impact data of the centralized storage system to be evaluated in the target deployment environment by first analyzing the performance impact factor set of the centralized storage system to be evaluated in the target deployment environment and the performance impact value corresponding to at least one performance impact factor in the performance impact factor set; then, when a target deployment environment sample corresponding to the performance impact data is matched in the performance sample library, the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment is determined according to the target system operation performance data corresponding to the target deployment environment sample. Therefore, it can solve the technical problem that there may be a large difference between the performance baseline of the storage product and the actual deployment environment of the customer, resulting in a large error in the performance prediction and the inability to provide an accurate performance reference, and can improve the accuracy of the centralized storage performance evaluation.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are provided to facilitate a better understanding of the present invention and do not constitute a limitation of the present disclosure.

[0021] Figure 1 A flowchart of a centralized storage performance evaluation method provided by an embodiment of the present disclosure;

[0022] Figure 2 A flowchart of another centralized storage performance evaluation method provided by an embodiment of the present disclosure;

[0023] Figure 3 A schematic diagram of the structure of a centralized storage performance evaluation architecture provided by an embodiment of the present disclosure;

[0024] Figure 4 A schematic diagram of the structure of a centralized storage performance evaluation device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0025] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0026] It's important to note that during the project delivery phase of a centralized storage system, it's necessary to evaluate whether the centralized storage system meets customer requirements, the theoretical performance of the current networking and storage configuration, and whether the current testing meets the theoretical estimate. However, evaluating the performance of a centralized storage system is a complex, systemic issue, influenced by a combination of numerous factors. This multitude of factors complicates the accuracy of accurately evaluating storage performance. For example, significant differences between the centralized storage system's shipping configuration and the baseline performance configuration in the lab, significant differences between the customer's site network topology and the lab's, and incomplete alignment between the ever-changing upper-layer business model and the performance baseline model can all affect the accuracy of the performance evaluation.

[0027] The performance evaluation of a centralized storage system can be achieved by constructing a sample library that includes multiple deployment environment samples and the performance data corresponding to each deployment environment sample, and matching the deployment environment of the application required by the centralized storage system within the sample library. However, due to the large number of factors involved in the deployment environment and the different combinations, it is difficult to achieve 100% coverage of all deployment environments in the sample library, and it is difficult to match deployment environment samples that are completely consistent with the deployment environment of the required application. Even if similar deployment environment samples are matched, the final evaluation results may be inaccurate because most of the common factors do not affect performance.

[0028] The centralized storage performance evaluation method, device, and electronic device according to embodiments of the present disclosure will be described below with reference to the accompanying drawings.

[0029] Figure 1 A flowchart of a centralized storage performance evaluation method provided by an embodiment of the present disclosure.

[0030] like Figure 1 As shown, the method comprises the following steps:

[0031] Step 101: Determine performance impact data of a centralized storage system to be evaluated in a target deployment environment.

[0032] The centralized storage system to be evaluated refers to a centralized storage system that requires performance evaluation.

[0033] The deployment environment refers to the actual operating environment in which the software or system runs, including hardware, operating system, network configuration, dependent software, and other components. The target deployment environment refers to the environment in which the centralized storage system to be evaluated will be deployed.

[0034] The performance impact data refers to the relevant data in the target deployment environment that affects the operational performance of the centralized storage system to be evaluated, and includes a set of performance impact factors and a performance impact value corresponding to at least one performance impact factor in the set of performance impact factors.

[0035] The performance impact factor refers to the factor that affects the operating performance of the centralized storage system to be evaluated in the target deployment environment. The performance impact value refers to the impact value of the performance impact factor on the operating performance of the centralized storage system to be evaluated.

[0036] Step 102: Match performance impact data in the performance sample library.

[0037] According to some embodiments, the performance sample library includes multiple deployment environment samples, and sample performance impact data and system operation performance data corresponding to at least one deployment environment sample among the multiple deployment environment samples.

[0038] The deployment environment sample refers to a sample of the deployment environment. The sample performance impact data refers to the performance impact data corresponding to the deployment environment sample. The system operation performance data refers to the operation performance data of the centralized storage system to be evaluated under the deployment environment sample.

[0039] Step 103 : if there is a target deployment environment sample matching the performance impact data in the performance sample library, determine the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment according to the target system operation performance data corresponding to the target deployment environment sample.

[0040] The target deployment environment sample refers to the deployment environment sample in the performance sample library that matches the performance impact data, and the target system operation performance data refers to the system operation performance data corresponding to the target deployment environment sample.

[0041] The performance evaluation result refers to the operational performance data of the centralized storage system to be evaluated in the target deployment environment.

[0042] In summary, the method provided by the embodiments of the present disclosure obtains the performance impact data of the centralized storage system to be evaluated in the target deployment environment by first analyzing the performance impact factor set of the centralized storage system to be evaluated in the target deployment environment and the performance impact value corresponding to at least one performance impact factor in the performance impact factor set; then, when the target deployment environment sample corresponding to the performance impact data is matched in the performance sample library, the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment is determined according to the target system operation performance data corresponding to the target deployment environment sample. Therefore, it can solve the technical problem that there may be a large difference between the performance baseline of the storage product and the actual deployment environment of the customer, resulting in large errors in performance prediction and inability to provide accurate performance references, and can improve the accuracy of centralized storage performance evaluation; secondly, since each deployment environment has many components and different combinations, compared with directly matching the target deployment environment in the performance sample library, since the factors in the performance impact data are only performance impact factors, their number is less than the number of factors in the target deployment environment, therefore, the matching efficiency and matching accuracy can be improved.

[0043] It should be noted that the embodiments of the present disclosure may include multiple steps. For the convenience of description, these steps are numbered, but these numbers do not limit the execution time slots or execution order between the steps; these steps can be implemented in any order, and the embodiments of the present disclosure do not limit this.

[0044] Furthermore, in a possible implementation of this embodiment, a centralized storage performance evaluation method is provided, which includes the following steps:

[0045] Step 201: Determine the architecture-level performance influencing factors of the centralized storage system to be evaluated in the target deployment environment, and obtain a set of architecture-level performance influencing factors.

[0046] According to some embodiments, the set of architecture-level performance influencing factors includes at least one architecture-level performance influencing factor, which refers to a factor at the architecture level of the target deployment environment that affects the operational performance of the centralized storage system to be evaluated.

[0047] In some embodiments, the architecture layer includes but is not limited to the application layer, the host layer, the network layer, and the storage layer.

[0048] The Application Layer (AL) refers to the software layer directly oriented toward business logic and user interaction. Its impact on operational performance can be reflected in the behavioral characteristics of application-layer services, such as the input / output (IO) business model and characteristics. These characteristics include, but are not limited to, block size, read / write ratio, cache hit rate, burstiness, and duration. This requires providing a storage product performance baseline for centralized storage systems based on the performance indicators of different IO business models, and analyzing and summarizing the centralized storage system applications and IO business models at customer sites. For example, the basic IO business model for a centralized storage system is an online transaction processing (OLTP) model, with a read / write ratio of 7:3 and a cache hit rate of 70%. Only by establishing a laboratory test to correlate this basic IO business model with the IO business model of actual customer applications can we provide highly accurate performance predictions without actual use.

[0049] The host layer (HL) refers to the physical or virtual computing resource layer that hosts applications. Configurations within the host layer, such as the physical machine type, operating system type, host kernel parameter settings, and multipath configuration, all impact performance. Different applications may perform differently under different configurations, or in other words, each application has an optimal configuration. The host layer provides the optimal parameter configuration for the centralized storage system being evaluated, which may differ from the target deployment environment.

[0050] The Network Layer (NL) is responsible for data transmission and communication between components in the deployment environment. Network parameters for the NL include, but are limited to, network type (such as Fibre Channel Storage Area Network (FC SAN), Internet Protocol Storage Area Network (IP SAN), etc.), interface type, and number of interfaces. As an intermediate link, link anomalies, interface types, and network types can all affect operational performance.

[0051] The Storage Layer (SL) is responsible for persistent data storage and access. It is the core node used in the centralized storage system being evaluated. Because the storage layer itself is a complex operating system, its performance is influenced by numerous factors, including but not limited to backend disk type, storage topology, pool type, number of volume types, and advanced features.

[0052] In some embodiments, to simplify the set of architecture-level performance influencing factors, a target number of factors that significantly impact the operational performance of the centralized storage system to be evaluated may be selected from the application, host, network, and storage planes as the architecture-level performance influencing factors. For example, for a centralized storage system to be evaluated, the ultimately selected architecture-level performance influencing factors may include the storage platform, storage version, storage model, networking architecture, and upper-layer IO model.

[0053] Step 202 : Determine a configuration-level performance influencing factor subset corresponding to at least one architecture-level performance influencing factor in the architecture-level performance influencing factor set under the target deployment environment, and obtain a configuration-level performance influencing factor set.

[0054] According to some embodiments, the configuration level of architectural performance influencing factors may include multiple configuration factors. For example, the configuration level of the networking architecture may include configuration factors such as network type and connection topology; the configuration level of the storage platform may include Redundant Array of Independent Disks (RAID) type, RAID number, disk type, disk number, volume type, and volume number.

[0055] In some embodiments, factors affecting the operational performance of the centralized storage system to be evaluated among the multiple configuration factors at the configuration level may constitute a configuration level performance impact factor subset corresponding to the architecture level performance impact factor. The configuration level performance impact factor set may include multiple configuration level performance impact factor subsets.

[0056] Step 203: Quantitatively analyze the configuration-level performance influencing factor set to obtain a performance impact value corresponding to at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set.

[0057] According to some embodiments, a correlation between at least one configuration-level performance influencing factor in a set of configuration-level performance influencing factors and system operational performance can be determined. Based on the correlation, the set of configuration-level performance influencing factors is quantitatively analyzed to obtain a performance impact value corresponding to the at least one configuration-level performance influencing factor in the set of configuration-level performance influencing factors. This improves the efficiency and accuracy of obtaining performance impact values.

[0058] In some embodiments, the correlation between the configuration-level performance influencing factors and system performance can be obtained through pre-testing. For example, the parameters of a certain configuration-level performance influencing factor in the deployment environment can be adjusted while the parameters of other configuration-level performance influencing factors remain unchanged, and the corresponding system performance data can be recorded to obtain the correlation between the certain configuration-level performance influencing factor and system performance. Alternatively, the configuration-level performance influencing factors can be quantified and linearized by establishing a performance impact analysis matrix.

[0059] Configuration-level performance factors are linearly additive. Once a parameter for a performance factor reaches a bottleneck, increasing that parameter will no longer improve the corresponding system performance. For example, when the number of mechanical disks is small, adding more mechanical disks can linearly increase performance. However, once the number of mechanical disks reaches a certain level, further additions will not improve performance.

[0060] For example, RAID type parameters include RAID10, RAID5, and RAID6. The correlation between RAID type and system performance is that the system performance data corresponding to RAID10 is greater than the system performance data corresponding to RAID5, and there is a 10% linear growth value between the system performance data corresponding to RAID5 and the system performance data corresponding to RAID6.

[0061] For example, the correlation between the number of RAIDs and the system performance is that the linear growth value of the system performance data is 1 / the current number of RAIDs; until the number of RAIDs reaches 6, the system performance data reaches the upper limit and remains unchanged.

[0062] For example, Non-Volatile Memory Express (NVMe) disks offer the highest performance, 5% higher than Serial Attached Small Computer System Interface (SAS) disks. The performance between the two types of disks increases linearly by 3%.

[0063] For example, volume types may include ordinary volumes, thin volumes, and compressed volumes. The volume performance of thin volumes may be 90% of that of ordinary volumes, and the volume performance of compressed volumes may be 70% of that of ordinary volumes, and so on. There is a linear growth value of 20% in the system operation performance data between them.

[0064] For example, for the number of volumes, the linear growth value of the corresponding system performance data is 1 / the current number of volumes; until the number of volumes reaches 16, the system performance data reaches the upper limit and remains unchanged.

[0065] Among them, when determining the performance impact value corresponding to the configuration-level performance impact factor, it can be determined based on the linear growth value indicated in the association relationship.

[0066] It should be noted that the performance impact data can be expressed in the form of a performance evaluation model. The performance evaluation model includes a basic architecture layer and a storage configuration layer. The basic architecture layer includes a set of architecture-level performance impact factors, and the storage configuration layer includes a set of configuration-level performance impact factors and performance impact values.

[0067] The sample performance impact data corresponding to a deployment environment sample in the performance sample library can also be expressed in the form of a performance evaluation model. For example, the sample performance impact data and system operation performance data corresponding to a deployment environment sample in the performance sample library can be exemplified as follows:

[0068] The basic architecture layer includes FC architecture, Hygon CPU platform, HF8000 storage model, and storage system version 6.X;

[0069] The storage configuration layer includes 128 NVMEs, 4 RAID5s, 10 thin volumes, and an upper-layer IO model of 8K with a 7:3 read-write ratio;

[0070] The system performance data includes X million IOPS, 0.5 ms latency, and 4 GB bandwidth.

[0071] In some embodiments, by executing steps S201 to S203, the performance impact factors in the target deployment environment can be hierarchically stratified, and the architecture-level performance impact factor set and the configuration-level performance impact factor set can be determined hierarchically. The architecture level serves as the framework or foundation for storage performance evaluation. First, the architecture-level performance impact factor set is determined, and then the configuration-level performance impact factor set is determined. This can improve the accuracy of performance impact data determination. Secondly, the configuration-level performance impact factors are quantified and linearized based on the hierarchical stratification to obtain a performance impact value corresponding to at least one configuration-level performance impact factor. This can improve the efficiency and accuracy of obtaining performance impact values.

[0072] Step 204: Match the architecture-level performance impact factor set in the performance sample library, and if there is a sample performance impact data set that matches the architecture-level performance impact factor set in the performance sample library, match the sample performance impact data set with the configuration-level performance impact factor set and the performance impact value of at least one configuration-level performance impact factor in the configuration-level performance impact factor set.

[0073] According to some embodiments, since the architecture level serves as the architecture or basis for storage performance evaluation, further evaluation can only be performed if there is sample performance impact data including a set of architecture-level performance impact factors in the performance sample library. Therefore, when matching the set of architecture-level performance impact factors in the performance sample library, a full match search is required in the performance sample library, that is, it is necessary to query the performance sample library for sample performance impact data including a set of architecture-level performance impact factors. If there is a set of sample performance impact data that matches the set of architecture-level performance impact factors, the basic architecture-level query is completed; if the query result is none, it means that there is no matching basic architecture-level data in the performance sample library and further evaluation is impossible.

[0074] In some embodiments, since there are many and ever-changing performance influencing factors at the storage level, the performance sample library may not be able to achieve 100% coverage, and the influence of the performance influencing factors at the storage level can be quantified as a performance impact value. Therefore, for configuration-level matching, an inexact match can be used. For example, when performing an inexact match, the total performance impact value corresponding to the configuration-level performance influencing factor set can be determined; the sample performance impact data set is searched for, and the difference between the sample performance impact total value and the performance impact total value is less than the difference threshold, and the sample performance impact data includes a preset number of configuration-level performance influencing factors in the configuration-level performance influencing factor set. Therefore, the matching efficiency and matching effect for the sample performance impact data set can be improved.

[0075] The total performance impact value refers to the sum of the performance impact values of multiple configuration-level performance impact factors in the configuration-level performance impact factor set.

[0076] The difference threshold and the preset number do not refer to a fixed value and can be determined according to the actual application scenario.

[0077] In some embodiments, when querying the sample performance impact data set for sample performance impact data in which the difference between the total sample performance impact value and the total performance impact value is less than the difference threshold, and includes a preset number of configuration-level performance impact factors in the configuration-level performance impact factor set, the sample performance impact data in which the difference between the total sample performance impact value and the total performance impact value is less than the difference threshold can be queried first; if relevant sample performance impact data is queried, then the sample performance impact data including the preset number of configuration-level performance impact factors in the configuration-level performance impact factor set are queried from these sample performance impact data.

[0078] The query order for the total performance impact value and the configuration-level performance impact factor set is not limited to the above order. The configuration-level performance impact factor set may be queried first, and then the total performance impact value may be queried.

[0079] It should be noted that, when the query result is none, the feedback mechanism can be activated. That is, when there is no sample performance impact data set that matches the architecture-level performance impact factor set in the performance sample library, a first match failure feedback message can be issued; wherein, the first match failure feedback message is used to indicate that a performance test is to be performed on the centralized storage system to be evaluated based on the architecture-level performance impact factor set. When there is no subset of sample performance impact data that matches the configuration-level performance impact factor set and the performance impact value in the deployment environment sample set, a second match failure feedback message is issued; wherein, the second match failure feedback message is used to indicate that a performance test is to be performed on the centralized storage system to be evaluated based on the configuration-level performance impact factor set and the performance impact value. Therefore, relevant personnel can be reminded to conduct on-site actual performance tests on the centralized storage system to be evaluated based on the configuration-level performance impact factor set or the architecture-level performance impact factor set in a timely manner.

[0080] In some embodiments, in response to receiving first performance test data uploaded in response to first match failure feedback information, the first performance test data can be placed in a performance sample library; wherein the first performance test data includes a deployment environment sample constructed based on a set of performance influencing factors at the architecture level, and sample performance impact data and system operation performance data corresponding to the deployment environment sample. In response to receiving second performance test data uploaded in response to second match failure feedback information, the second performance test data is placed in a performance sample library; wherein the second performance test data includes a deployment environment sample constructed based on a set of performance influencing factors at the configuration level, and sample performance impact data and system operation performance data corresponding to the deployment environment sample. Therefore, the performance use case supplement of the performance sample library can be achieved for use in subsequent performance evaluations.

[0081] Step 205: When there is a sample performance impact data subset in the sample performance impact data set that matches the configuration level performance impact factor set and the performance impact value, the sample performance impact data with the lowest difference between the configuration level performance impact factor set and the performance impact value is selected from the sample performance impact data subset as the target sample performance impact data.

[0082] According to some embodiments, the difference value may be, for example, the sum of a first similarity corresponding to a set of configuration-level performance impact factors and a second similarity corresponding to a total performance impact value.

[0083] Step 206: Use the deployment environment sample corresponding to the target sample performance impact data as the target deployment environment sample that matches the performance impact data.

[0084] It should be noted that, by executing steps 204 to 206 , the performance impact data can be matched step by step in a hierarchical manner, thereby improving the matching efficiency and matching effect of the performance impact data.

[0085] Step 207 : if there is a target deployment environment sample matching the performance impact data in the performance sample library, determine the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment according to the target system operation performance data corresponding to the target deployment environment sample.

[0086] According to some embodiments, when a target deployment environment sample that matches the performance impact data exists in the performance sample library, if the performance impact data is the same as the target sample performance impact data, the target system operating performance data corresponding to the target deployment environment sample is used as the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment; if the performance impact data is different from the target sample performance impact data, the target system operating performance data can be adjusted based on the configuration-level performance impact factor set and the performance impact value, and the adjusted target system operating performance data can be used as the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment. Therefore, the data accuracy of the performance evaluation result can be improved.

[0087] In some embodiments, the difference between the set of configuration-level performance influencing factors and the performance impact values can be calculated, and linear calculations can be performed on each configuration item corresponding to the target sample performance impact data according to the performance impact value of the configuration-level performance influencing factors in turn, to finally obtain the adjusted target system operation performance data.

[0088] In some embodiments, for configuration-level performance impact factors with upper limits, the parameters corresponding to the upper limits are used for comparison. For example, if 100 hard drives can achieve full performance, and the hard drive configuration is 120, the configuration comparison is based on the saturation value of 100 hard drives.

[0089] Step 208: Acquire actual operating performance data of the centralized storage system to be evaluated in the target deployment environment.

[0090] According to some embodiments, the actual operation performance data can be obtained by setting up a target deployment environment in a laboratory to perform performance testing on the centralized storage system to be evaluated, or by delivering the centralized storage system to be evaluated to a customer who adopts the target deployment environment and obtaining actual operation feedback from the customer.

[0091] Step 209: When the performance data difference between the actual operation performance data and the performance evaluation result is greater than the performance difference threshold, an evaluation calibration message is issued.

[0092] According to some embodiments, the evaluation and calibration information is used to indicate the calibration strategy for adjusting the target system's operational performance data. A performance data difference greater than a performance difference threshold indicates a problem with the adjustment strategy itself and requires updating, which can be accomplished through intervention by relevant personnel.

[0093] In some embodiments, the performance difference threshold does not specifically refer to a fixed threshold, and the performance difference threshold can be determined according to actual application scenarios, for example.

[0094] Step 210: Upload the target deployment environment, performance impact data, and actual operation performance data to the performance sample library.

[0095] It should be noted that, by executing steps 208 to 210 , the feedback of the performance evaluation can be improved, and the accuracy of the performance evaluation can be gradually improved.

[0096] Take a scenario as an example, Figure 2 This is a flow chart of another centralized storage performance evaluation method provided by the embodiment of the present disclosure. Figure 2 As shown, first, the performance impact data of the centralized storage system to be evaluated in the target deployment environment can be obtained through performance evaluation input or actual environment information input. Then, performance samples (deployment environment samples) can be queried in the performance sample library based on the performance impact data. If the query result is a complete match, the corresponding evaluation result is output and the process ends. If the query result is an incomplete match, the performance evaluation algorithm is used to adjust the target system operation performance data to evaluate the performance; then, by obtaining the actual operation performance data of the centralized storage system to be evaluated in the target deployment environment, the performance data evaluation accuracy is evaluated to see whether it meets the requirements, that is, to determine whether the performance data difference is greater than the performance difference threshold; if it meets the requirements, the corresponding evaluation result is output and the process ends; if it does not meet the requirements, manual intervention is performed to evaluate the performance evaluation algorithm, and the performance sample library is supplemented and improved.

[0097] In some embodiments, Figure 3 This is a schematic diagram of the structure of a centralized storage performance evaluation architecture provided by an embodiment of the present disclosure. Figure 3 As shown, the centralized storage performance evaluation architecture has a mechanism and framework for automatic evaluation and error correction. The centralized storage performance evaluation architecture includes an automatic performance evaluation mechanism state machine, which is responsible for executing the centralized storage performance evaluation method, namely, establishing a performance evaluation model, matching performance sample libraries, executing core algorithms (performance evaluation algorithms), and executing feedback mechanisms.

[0098] The performance sample library can include the measured performance data of basic laboratory scenarios and the performance feedback results of actual customer on-site delivery. The goal of establishing the performance sample library is 100% coverage. When the performance sample library is 100% established, there is no need for the participation of subsequent performance evaluation algorithms, but instead the matching is completed directly to output the corresponding data. By executing the feedback mechanism, the various deployment environments of the centralized storage system to be evaluated can be continuously analyzed, summarized, and judged to give performance evaluation results, which are added to the performance sample library, and the core algorithm is improved and corrected. With the increase of samples in the performance sample library and the correction of the core algorithm, the accuracy of the performance evaluation can be continuously improved until it reaches the ideal state, so that the centralized storage performance evaluation architecture can gradually cope with the performance evaluation of the customer's on-site pre-sales stage and the further accurate evaluation of the deployment environment in the after-sales handover stage, thereby improving the customer experience and solving the scenario where the performance cannot be supported after the business is launched in the later stage.

[0099] When obtaining measured performance data for basic laboratory scenarios, you can automatically import measured performance data from samples of different deployment environments. For example, after completing a set of system performance data for a particular version model in a network, you can begin traversing system performance data for different performance models. After traversing different performance models, you can then traverse system performance data for different versions. After traversing different versions, you can then traverse system performance data for different hardware models. After traversing different hardware models, you can then traverse system performance data for different network topologies.

[0100] The existence of the feedback mechanism enables the centralized storage performance evaluation architecture to learn, and it can operate in the following three scenarios:

[0101] There is a new deployment environment, that is, for the scenario where the query result of the performance sample library is none, in this scenario, you only need to upload the corresponding performance test data to the performance sample library;

[0102] If the difference between the actual operating performance data and the performance evaluation results exceeds the performance difference threshold, relevant personnel must intervene to update the core algorithm.

[0103] The scenario where the adjustment fails when using the performance evaluation algorithm to adjust the target system's operating performance data is mainly caused by the fact that the difference threshold and the preset number of values do not match the performance evaluation algorithm during incomplete matching, and the performance impact data that cannot be evaluated by the performance evaluation algorithm is matched. The difference threshold and the preset number can be adjusted through the intervention of relevant personnel, or the performance evaluation algorithm can be updated and improved.

[0104] 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 the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method.

[0105] According to an embodiment of the present disclosure, the present disclosure also provides a centralized storage performance evaluation device.

[0106] For example, Figure 4 This is a schematic diagram of the structure of a centralized storage performance evaluation device provided by an embodiment of the present disclosure. The centralized storage performance evaluation device 400 includes:

[0107] The data determining unit 401 is configured to determine performance impact data of the centralized storage system to be evaluated in a target deployment environment, wherein the performance impact data includes a set of performance impact factors and a performance impact value corresponding to at least one performance impact factor in the set of performance impact factors;

[0108] A sample matching unit 402 is configured to match performance impact data in a performance sample library, wherein the performance sample library includes multiple deployment environment samples, and sample performance impact data and system operation performance data corresponding to at least one deployment environment sample among the multiple deployment environment samples;

[0109] The performance evaluation unit 403 is used to determine the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment according to the target system operation performance data corresponding to the target deployment environment sample when there is a target deployment environment sample matching the performance impact data in the performance sample library.

[0110] Furthermore, when the data determining unit 401 is used to determine the performance impact data of the centralized storage system to be evaluated in the target deployment environment, it is specifically used to:

[0111] Determine the architecture-level performance influencing factors of the centralized storage system to be evaluated in the target deployment environment, and obtain a set of architecture-level performance influencing factors;

[0112] Determine a configuration-level performance influencing factor subset corresponding to at least one architecture-level performance influencing factor in the architecture-level performance influencing factor set in the target deployment environment to obtain a configuration-level performance influencing factor set;

[0113] A quantitative analysis is performed on the configuration-level performance influencing factor set to obtain a performance impact value corresponding to at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set.

[0114] Furthermore, the data determination unit 401 is configured to perform a quantitative analysis on the configuration-level performance influencing factor set and obtain a performance impact value corresponding to at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set, specifically for:

[0115] Determine a correlation relationship between at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set and system operation performance;

[0116] A quantitative analysis is performed on the configuration-level performance influencing factor set based on the correlation relationship to obtain a performance impact value corresponding to at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set.

[0117] Furthermore, when the sample matching unit 402 is used to match the performance impact data in the performance sample library, it is specifically used to:

[0118] Matching the architecture-level performance impact factor set in the performance sample library, and if there is a sample performance impact data set matching the architecture-level performance impact factor set in the performance sample library, matching the sample performance impact data set with the configuration-level performance impact factor set and the performance impact value of at least one configuration-level performance impact factor in the configuration-level performance impact factor set;

[0119] When there is a sample performance impact data subset that matches the configuration-level performance impact factor set and the performance impact value in the sample performance impact data set, select the sample performance impact data with the lowest difference between the sample performance impact data subset and the configuration-level performance impact factor set and the performance impact value as the target sample performance impact data;

[0120] The deployment environment sample corresponding to the target sample performance impact data is used as the target deployment environment sample that matches the performance impact data.

[0121] Furthermore, when the sample matching unit 402 is used to match the set of architecture-level performance influencing factors in the performance sample library, it is specifically used to:

[0122] Query the performance sample library for sample performance impact data including a set of architecture-level performance impact factors.

[0123] Furthermore, when the sample matching unit 402 is configured to match the sample performance impact data set with the configuration-level performance impact factor set and the performance impact value of at least one configuration-level performance impact factor in the configuration-level performance impact factor set, it is specifically configured to:

[0124] Determine the total performance impact corresponding to the set of configuration-level performance impact factors;

[0125] The sample performance impact data set is queried for sample performance impact data whose difference between the total sample performance impact value and the total performance impact value is less than the difference threshold, and includes a preset number of configuration-level performance impact factors in the configuration-level performance impact factor set.

[0126] Furthermore, the performance evaluation unit 403 is configured to determine the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment based on the target system operation performance data corresponding to the target deployment environment sample, specifically to:

[0127] If the performance impact data is the same as the target sample performance impact data, the target system operation performance data corresponding to the target deployment environment sample is used as the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment;

[0128] When the performance impact data is different from the target sample performance impact data, the target system operation performance data is adjusted according to the configuration level performance impact factor set and the performance impact value, and the adjusted target system operation performance data is used as the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment.

[0129] Furthermore, the centralized storage performance evaluation apparatus 400 further includes an evaluation feedback unit 404. After determining the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment, the evaluation feedback unit 404 is configured to:

[0130] Obtain actual operating performance data of the centralized storage system to be evaluated in the target deployment environment;

[0131] When the performance data difference between the actual operation performance data and the performance evaluation result is greater than the performance difference threshold, issuing evaluation calibration information, wherein the evaluation calibration information is used to indicate an adjustment strategy when the calibration adjusts the target system operation performance data;

[0132] Upload the target deployment environment, performance impact data, and actual operation performance data to the performance sample library.

[0133] Furthermore, after matching the set of architecture-level performance influencing factors in the performance sample library, the evaluation feedback unit 404 is used to:

[0134] In a case where there is no sample performance impact data set matching the architecture-level performance impact factor set in the performance sample library, issuing first match failure feedback information, wherein the first match failure feedback information is used to instruct to perform a performance test on the centralized storage system to be evaluated based on the architecture-level performance impact factor set;

[0135] When there is no sample performance impact data subset that matches the configuration-level performance impact factor set and the performance impact value in the deployment environment sample set, a second match failure feedback information is issued, wherein the second match failure feedback information is used to indicate that a performance test is to be performed on the centralized storage system to be evaluated based on the configuration-level performance impact factor set and the performance impact value.

[0136] Furthermore, the evaluation feedback unit 404 is further configured to:

[0137] In response to receiving the first performance test data uploaded in response to the first matching failure feedback information, placing the first performance test data into a performance sample library;

[0138] In response to receiving the second performance test data uploaded in response to the second matching failure feedback information, the second performance test data is placed in a performance sample library.

[0139] It should be noted that, for the description of the features in the embodiment corresponding to the centralized storage performance evaluation device, reference can be made to the relevant description of the embodiment corresponding to the centralized storage performance evaluation method, which will not be repeated here.

[0140] An embodiment of the present disclosure further provides an electronic device, 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 of the above-mentioned centralized storage performance evaluation method embodiments.

[0141] An embodiment of the present disclosure 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-mentioned centralized storage performance evaluation method embodiments when running.

[0142] 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.

[0143] An embodiment of the present disclosure further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any one of the above-mentioned centralized storage performance evaluation method embodiments are implemented.

[0144] An embodiment of the present disclosure also provides another computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in any of the above-mentioned centralized storage performance evaluation method embodiments are implemented.

[0145] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this disclosure.

[0146] The above is a detailed introduction to a centralized storage performance evaluation method provided by the present disclosure. This article uses specific examples to illustrate the principles and implementation methods of the present disclosure. The description of the above embodiments is only used to help understand the method of the present disclosure and its core ideas. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present disclosure, several improvements and modifications can be made to the present disclosure, and these improvements and modifications also fall within the scope of protection of the claims of the present disclosure.

Claims

1. A centralized storage performance evaluation method, characterized in that: include: Determine the performance impact data of the centralized storage system to be evaluated in the target deployment environment, wherein the performance impact data includes a set of performance impact factors and a performance impact value corresponding to at least one performance impact factor in the set of performance impact factors; including: determining the architecture-level performance impact factors of the centralized storage system to be evaluated in the target deployment environment to obtain the architecture-level performance impact factor set; determining a subset of configuration-level performance impact factors corresponding to at least one architecture-level performance impact factor in the architecture-level performance impact factor set in the target deployment environment to obtain the configuration-level performance impact factor set; performing quantitative analysis on the configuration-level performance impact factor set to obtain the performance impact value corresponding to at least one configuration-level performance impact factor in the configuration-level performance impact factor set; Matching the performance impact data in a performance sample library, wherein the performance sample library includes multiple deployment environment samples, and sample performance impact data and system operation performance data corresponding to at least one deployment environment sample among the multiple deployment environment samples; including: querying the performance sample library for sample performance impact data including the architecture-level performance impact factor set, and matching the sample performance impact data set with the configuration-level performance impact factor set and the performance impact value of at least one configuration-level performance impact factor in the configuration-level performance impact factor set when there is a sample performance impact data set matching the architecture-level performance impact factor set in the performance sample library; when there is a sample performance impact data subset matching the configuration-level performance impact factor set and the performance impact value in the sample performance impact data set, selecting the sample performance impact data with the lowest difference value between the sample performance impact data and the configuration-level performance impact factor set and the performance impact value from the sample performance impact data subset as the target sample performance impact data; and using the deployment environment sample corresponding to the target sample performance impact data as the target deployment environment sample matching the performance impact data; When a target deployment environment sample matching the performance impact data exists in the performance sample library, the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment is determined according to the target system operation performance data corresponding to the target deployment environment sample.

2. The method according to claim 1, characterized in that The quantitative analysis of the configuration-level performance influencing factor set to obtain a performance impact value corresponding to at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set includes: Determining a correlation relationship between at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set and system operation performance; Based on the association relationship, the configuration-level performance influencing factor set is quantitatively analyzed to obtain a performance impact value corresponding to at least one configuration-level performance influencing factor in the configuration-level performance influencing factor set.

3. The method according to claim 1, characterized in that Matching the sample performance impact data set with the configuration-level performance impact factor set and the performance impact value of at least one configuration-level performance impact factor in the configuration-level performance impact factor set includes: Determine a total performance impact value corresponding to the set of configuration-level performance impact factors; The difference between the total sample performance impact value and the total performance impact value in the sample performance impact data set is less than the difference threshold, and includes sample performance impact data of a preset number of configuration-level performance impact factors in the configuration-level performance impact factor set.

4. The method according to claim 3, characterized in that Determining, based on the target system operation performance data corresponding to the target deployment environment sample, a performance evaluation result of the centralized storage system to be evaluated under the target deployment environment includes: In the case where the performance impact data is the same as the target sample performance impact data, the target system operation performance data corresponding to the target deployment environment sample is used as the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment; When the performance impact data is different from the target sample performance impact data, the target system operation performance data is adjusted according to the configuration level performance impact factor set and the performance impact value, and the adjusted target system operation performance data is used as the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment.

5. The method according to claim 4, characterized in that After determining the performance evaluation result of the centralized storage system to be evaluated in the target deployment environment, the method further includes: Obtaining actual operating performance data of the centralized storage system to be evaluated in the target deployment environment; When a performance data difference between the actual operation performance data and the performance evaluation result is greater than a performance difference threshold, issuing evaluation calibration information, wherein the evaluation calibration information is used to indicate an adjustment strategy for adjusting the target system operation performance data; The target deployment environment, the performance impact data, and the actual operation performance data are uploaded to the performance sample library.

6. The method according to claim 1, characterized in that After matching the set of architecture-level performance influencing factors in the performance sample library, the method further includes: When there is no sample performance impact data set matching the architecture-level performance impact factor set in the performance sample library, issuing first match failure feedback information, wherein the first match failure feedback information is used to instruct to perform a performance test on the centralized storage system to be evaluated based on the architecture-level performance impact factor set; In response to receiving the first performance test data uploaded in response to the first matching failure feedback information, placing the first performance test data into the performance sample library; In a case where there is no sample performance impact data subset that matches the configuration-level performance impact factor set and the performance impact value in the deployment environment sample set, issuing second match failure feedback information, wherein the second match failure feedback information is used to instruct to perform a performance test on the centralized storage system to be evaluated based on the configuration-level performance impact factor set and the performance impact value; In response to receiving the second performance test data uploaded in response to the second matching failure feedback information, the second performance test data is placed in the performance sample library.

7. A centralized storage performance evaluation device, characterized in that: include: A data determination unit is used to determine performance impact data of the centralized storage system to be evaluated in a target deployment environment, wherein the performance impact data includes a set of performance impact factors and a performance impact value corresponding to at least one performance impact factor in the set of performance impact factors; specifically used to determine the architecture-level performance impact factors of the centralized storage system to be evaluated in the target deployment environment to obtain the architecture-level performance impact factor set; determine a subset of configuration-level performance impact factors corresponding to at least one architecture-level performance impact factor in the architecture-level performance impact factor set in the target deployment environment to obtain the configuration-level performance impact factor set; perform quantitative analysis on the configuration-level performance impact factor set to obtain a performance impact value corresponding to at least one configuration-level performance impact factor in the configuration-level performance impact factor set; a sample matching unit for matching the performance impact data in a performance sample library, wherein the performance sample library includes a plurality of deployment environment samples, and sample performance impact data and system operation performance data corresponding to at least one deployment environment sample among the plurality of deployment environment samples; specifically for querying the performance sample library for sample performance impact data including the architecture-level performance impact factor set, and, if a sample performance impact data set matching the architecture-level performance impact factor set exists in the performance sample library, matching the sample performance impact data set with the configuration-level performance impact factor set and the performance impact value of at least one configuration-level performance impact factor in the configuration-level performance impact factor set; if a sample performance impact data subset matching the configuration-level performance impact factor set and the performance impact value exists in the sample performance impact data set, selecting, from the sample performance impact data subset, the sample performance impact data having the lowest difference value with the configuration-level performance impact factor set and the performance impact value as the target sample performance impact data; and selecting the deployment environment sample corresponding to the target sample performance impact data as the target deployment environment sample matching the performance impact data; The performance evaluation unit is used to determine the performance evaluation result of the centralized storage system to be evaluated under the target deployment environment according to the target system operation performance data corresponding to the target deployment environment sample, when there is a target deployment environment sample matching the performance impact data in the performance sample library.

8. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

Images