A database parameter tuning method based on parameter type customized optimization

Abstract

The application discloses a database parameter tuning method based on parameter type customization optimization, and belongs to the field of database configuration optimization. The method comprises the following steps: dividing a database parameter space into a continuous parameter subspace and a discrete parameter subspace according to parameter types, and projecting the continuous parameter subspace and the discrete parameter subspace into a low-dimensional space; using a GP model and an SMAC model as proxy models to search for continuous parameter configurations and discrete parameter configurations in the low-dimensional continuous parameter subspace and the low-dimensional discrete parameter subspace respectively; realizing the interaction between the GP model and the SMAC model through iterative optimization and a context-based communication mechanism; and mapping the parameter configuration results back to the database parameter space for performance evaluation. The application can effectively improve the performance and efficiency of database parameter tuning, and allows users to search for database parameter configurations with higher performance in a shorter time.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of database configuration tuning, and more particularly, to a database parameter tuning method based on parameter type customized optimization. BACKGROUND

[0002] Modern databases usually have hundreds or thousands of parameters to be set. Only when these parameters are set to appropriate values can the database achieve high throughput or low latency. Automated database parameter tuning uses machine learning techniques to automatically search for the best performance parameters in the parameter configuration space, thereby reducing the manual cost of enterprises on database administrators. The parameters in the database can be generally divided into discrete type parameters and continuous type parameters. There is no obvious order relationship between the former parameters and the corresponding performance, while the values of the latter parameters and the level of performance have obvious order relationship. At the same time, subtle changes in continuous type parameters often do not cause great changes in performance, while changes in discrete type parameter values are more likely to cause significant changes in performance. Existing tuning methods often use the same surrogate model to optimize both discrete and continuous parameters, resulting in poor performance. At the same time, existing methods often need to spend a lot of time to search for relatively satisfactory parameters, and the tuning efficiency is not high. SUMMARY

[0003] In view of the above defects or improvement needs of the prior art, the present application provides a database parameter tuning method based on parameter type customized optimization, thereby solving the problems of poor tuning performance and low tuning efficiency of existing database parameter tuning methods.

[0004] To achieve the above-mentioned purpose, according to the first aspect of the present application, a database parameter tuning method based on parameter type customized optimization is provided, comprising:

[0005] S1, dividing the database parameter space into a continuous type parameter subspace and a discrete type parameter subspace according to the parameter value type, and respectively mapping the continuous type parameter subspace and the discrete type parameter subspace to a low-dimensional surrogate space to obtain a continuous type low-dimensional surrogate space and a discrete type low-dimensional surrogate space;

[0006] S2, using a GP model to search for the configuration of the continuous type parameters in the continuous type low-dimensional surrogate space, using a SMAC model to search for the configuration of the discrete type parameters in the discrete type low-dimensional surrogate space, and respectively mapping the configuration of the continuous type parameters and the configuration of the discrete type parameters back to the database parameter space for performance testing, and storing the test results in the tuning history until a preset tuning round is reached; wherein the search of the GP model and the SMAC model is alternately performed, and the optimal parameter configuration searched by each model in the respective space is transmitted as context features to the other model;

[0007] S3, the performance optimal discrete type parameter and continuous type parameter in the tuning history are taken as the optimal parameter configuration of the database.

[0008] According to a second aspect of the present application, a database parameter deployment method is provided, comprising:

[0009] The optimal parameter configuration of the database is obtained by using the database parameter tuning method according to the first aspect.

[0010] If all the tuned parameters are dynamic parameters, the corresponding parameter values are written into the database by using a SQL command; if the tuned parameters also include static parameters, the corresponding parameter values are written into the database configuration file, and the database is restarted to complete the parameter deployment.

[0011] According to a third aspect of the present application, an electronic device is provided, comprising: a computer readable storage medium and a processor.

[0012] The computer readable storage medium is configured to store executable instructions.

[0013] The processor is configured to read the executable instructions stored in the computer readable storage medium, and execute the database parameter tuning method according to the first aspect or the database parameter deployment method according to the second aspect.

[0014] According to a fourth aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for causing a processor to execute the database parameter tuning method according to the first aspect or the database parameter deployment method according to the second aspect.

[0015] According to a fifth aspect of the present application, a computer program product is provided, which comprises computer programs or instructions, and the computer programs or instructions are executed by a processor to execute the database parameter tuning method according to the first aspect or the database parameter deployment method according to the second aspect.

[0016] In general, the above technical solutions conceived by the present application can achieve the following beneficial effects compared with the prior art:

[0017] The method provided by the present invention, when using machine learning technology to perform database parameter tuning, divides the database parameter space into a continuous parameter subspace and a categorical parameter subspace according to parameter type, and projects the subspace into a low-dimensional space. The GP model (Gaussian process model) and the SMAC model (sequence model-based configuration algorithm model) are used as proxy models in the low-dimensional continuous parameter subspace and the categorical parameter subspace, respectively, to search for continuous parameter configuration and categorical parameter configuration. Through iterative optimization and a context-based communication mechanism, the interaction between the GP model and the SMAC model is achieved. The parameter configuration results are mapped back to the database parameter space for performance evaluation. By decomposing the parameter space and performing customized optimization for the continuous parameter space and the discrete parameter space, the ability of the GP model to model continuous relationships and the advantage of the SMAC model in processing categorical parameters can be fully utilized, thereby achieving better tuning results. By using iterative optimization and a communication mechanism based on context features, the two models are coordinated to jointly capture the dependency relationship between parameters, thereby searching for a better parameter configuration. By using parameter dimension mapping technology, the high-dimensional space is mapped to the low-dimensional proxy space, which can reduce the model search space and improve search efficiency while ensuring that tuning performance is not compromised. Using batch Bayesian optimization technology, a batch acquisition function is used to sample multiple configurations at a time, and parameter performance is evaluated in parallel, further improving the efficiency of the tuning method.

[0018] As a further advantage, the method provided by the present invention employs a locally penalized acquisition function for batch sampling of parameter configurations. The local penalizer employed assigns lower weights to configurations close to the expected optimal point. This behavior ensures a diverse sample of configurations, enabling efficient and comprehensive information collection. Specifically, batch configuration sampling involves alternating maximization and penalty steps. In the maximization step, configurations not previously seen in the tuning history are acquired by maximizing the penalized acquisition function. In the penalty step, the penalized acquisition function is updated to include the selected batch configuration information. This process is repeated until the target batch size is reached. Furthermore, the method provided by the present invention employs a parallel parameter evaluation strategy. The client distributes a batch of sampled parameter configurations to different cloned database instances, with each database instance responsible for performance evaluation of a single parameter configuration. All cloned DBMS instances replay the same workload during the evaluation process. Using a synchronous parallel evaluation mechanism, the client waits for all workloads to complete replay and receive performance data before saving these observations to the tuning history and using them to update the tuning model. This further improves tuning efficiency and reduces tuning time. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of a method for optimizing database parameters through parameter type customization provided by an embodiment of the present invention;

[0020] Figure 2 Fig. 2 is a flowchart of a database parameter tuning method based on parameter type customization optimization according to an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. In addition, the technical features involved in the various embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0022] The existing method uses a unified optimizer (such as GP or DDPG, etc.) to simultaneously optimize the classification parameters and the discrete parameters without distinction, resulting in suboptimal tuning performance. At the same time, the tuning process of the existing method often requires a large number of tuning rounds, consuming a large amount of tuning time.

[0023] Based on this, the present application provides a database parameter tuning method based on parameter customization optimization, comprising:

[0024] S1, decomposing the database parameter space into a continuous parameter subspace and a discrete parameter subspace according to the parameter value type, and respectively mapping the continuous parameter subspace and the discrete parameter subspace to a low-dimensional agent space to obtain a continuous low-dimensional agent space and a discrete low-dimensional agent space;

[0025] S2, using a GP model to search for a configuration of continuous parameters in the continuous low-dimensional agent space, using a SMAC model to search for a configuration of discrete parameters in the discrete low-dimensional agent space, and respectively mapping the configuration of continuous parameters and the configuration of discrete parameters back to the database parameter space for performance testing, and storing the test results in a tuning history until a preset number of tuning rounds is reached; wherein the search of the GP model and the SMAC model is alternately performed, and the optimal parameter configuration searched by each model in the respective space is transmitted as a context feature to the other model;

[0026] S3, taking the discrete parameters and the continuous parameters with the best performance in the tuning history as the optimal parameter configuration of the database.

[0027] Specifically, as shown in FIG. 1, the method provided by the embodiment of the present application comprises: database parameter space decomposition, multi-agent model collaborative optimization, and parameter dimension mapping. Figure 1

[0028] The parameter space decomposition comprises:

[0029] ​The parameter space is divided into a continuous parameter subspace and a discrete parameter subspace according to the parameter value type, and the two subspaces are orthogonal to each other.

[0030] In the spatial decomposition, the value type and meaning of each parameter are checked. If the value type is an integer type or a floating point type, and the value size corresponds to a clear order relationship, the parameter is added to the continuous parameter set; if the parameter type is an enumeration type, or the parameter value has no order relationship, and each value only corresponds to a selection of the database configuration, the parameter is added to the discrete parameter set. The continuous parameter configuration space is defined as the Cartesian product of the possible value set of all continuous parameters, and the classification parameter subspace is also defined as the Cartesian product of the value set of all classification parameters.

[0031] The multi-agent model cooperative optimization includes:

[0032] In the classification subspace, a sequence model based configuration algorithm model (SMAC) is used to search for the optimal value of the discrete parameter, and in the continuous subspace, a Gaussian process model (GP) is used to search for the optimal value of the continuous parameter. The GP and SMAC models cooperate in an iterative optimization manner to jointly explore the entire parameter space. At the same time, the optimal parameter configuration searched by each model in the respective subspace is transmitted as context features to the other model, realizing context-based model communication to capture the dependency between different types of parameters.

[0033] In order to enable the agent model to support context features as input to obtain the global optimal parameter configuration, the agent model (GP model and SMAC model) needs to be modified.

[0034] The features of the GP model are defined by a prior mean function μ(·) and a variance function σ2(·) as follows:

[0035]

[0036] where Φ is a set of continuous parameter configurations θ l in the adjustment history, K is a covariance matrix, the (i,j) item of which is k(θ con ,θ con '), and k(θ con ,θ con ') is a kernel function in the GP, which is used to measure the similarity between continuous parameter configurations. In the embodiment of the application, a new kernel k((θ con ,F),(θ con ′ ,F′)) is defined as k Θ (θ con ,θ con ′)+k C(F, F ′ ) as the kernel function of GP, where k Θ is the Mahalanobis kernel used to measure the distance between two continuous parameter configurations θ con and θ con '. And k C is the kernel function used to measure the similarity between two context features. k C can be specified as the Hamming kernel to measure the similarity between two discrete context features or the linear kernel to measure the similarity between two continuous context features. It can be understood that the context feature encapsulated by the categorical parameter configuration is a discrete context feature, and the context feature encapsulated by the continuous parameter configuration is a continuous context feature.

[0037] In order to accurately measure the similarity between different context features F input to the SMAC, for the SMAC model, the configuration features (i.e. the historical configuration of the categorical parameters) and the context features are combined into a connected input for the surrogate model. This enables the random forest to flexibly capture changes in the continuous adjustment configuration.

[0038] The history of each tuning is recorded to a file, including the parameter configuration (continuous parameters and categorical parameters) of each step of tuning, the performance indicators obtained by each set of parameters, and the context features used by the model in the tuning step.

[0039] The parameter dimension mapping includes:

[0040] The high-dimensional parameter space is mapped to the low-dimensional surrogate space by using a random mapping matrix. The above optimization algorithm will run in the low-dimensional surrogate space, and the parameter configuration explored in the low-dimensional surrogate space will be mapped back to the high-dimensional original space, so as to correctly obtain the performance of the parameters.

[0041] In the parameter dimension mapping process, the HeSBO algorithm defines the Project_Down() interface and the Project_Up() interface, which are respectively used to map the original space into the low-dimensional surrogate space and to map the points in the low-dimensional surrogate space back to the high-dimensional original space.

[0042] In the Project_Down() interface, a random mapping matrix S is first initialized, each column of which has only one non-zero value (-1 or 1), and each parameter configuration θ in the original space will be mapped to the low-dimensional surrogate space by It can be understood that the size of S is the number of input parameters x the dimension of the surrogate space to be mapped, such as the size of the random mapping matrix S in the Project_Down() interface shown in Figure 1

[0043] ​In the Project_Up interface, based on the hash function defined in the HeSBO algorithm, high-dimensional mapping can be realized in a manner similar to However, since the range of the parameters (continuous parameters and discrete parameters) mapped back to the high-dimensional space (i.e., mapped to the high-dimensional space) is [-1, 1], it is necessary to further map it back to the original parameter range.

[0044] For continuous parameters, in order to map the continuous parameters θ con `∈[-1,1] mapped back to the original parameters The following formula is used:

[0045]

[0046] For classification parameters, [-1, 1] is uniformly divided into a corresponding number of intervals according to the number of categories of the original parameter values (i.e., the number of options for the values), and the final parameter value is determined according to the interval in which the mapping value is located. For example, for a certain discrete original parameter, if the value category is 0 or 1, i.e., the number of value categories is 2, then [-1, 1] is divided into [-1, 0] and [0, 1], corresponding to values of 0 and 1, respectively. If the mapping value of the classification parameter mapped back to the high-dimensional space is in [-1, 0], the value is 0, otherwise it is 1.

[0047] Before the start of each proxy model optimization interval, the optimal value (current optimal parameter configuration) currently searched by the other model is applied to the database. Then switch to another model to continue optimizing another subspace. In each optimization interval of the model, the model will first sample the parameter configuration by the acquisition function, and map it back to the high-dimensional original space to form the real parameter value, which is applied to the database to realize parameter performance evaluation, and the parameter and its performance value are used to update the model.

[0048] In the process of implementing context-based model communication, when each proxy model (GP or SMAC) ends the current optimization interval, the performance test results of each parameter configuration can be obtained, and the parameter configuration with the optimal performance searched by the current model is encapsulated into a context feature. The context feature records the latest optimization decision of the current model, which will be passed to another optimization model. The other optimization model takes the context feature as input to train the proxy model to capture the correspondence between the parameter value, the context feature and the parameter performance.

[0049] As a further preferred embodiment of the present application, each model searches for parameters in the respective subspace by using batch Bayesian optimization technology, using a local penalty-based acquisition function, sampling multiple parameter configurations at a time, and broadcasting them to multiple test environments for parallelized parameter performance testing, and collecting performance indicators to update the model.

[0050] That is, in the use of batch sampling and parallel parameter evaluation process, the use of local penalty-based batch acquisition function. The method in the process of parameter configuration sampling, give close to the expected optimal point of configuration imposed penalty, that is multiplied by a lower penalty factor, and through the maximization step and penalization step each other iteration until the sampling to the expected batch size of parameter configuration.

[0051] After sampling a batch of configurations, perform distributed parallel evaluation. The client first clones the target database management system (DBMS) instance to provide an evaluation environment, and the number of cloned DBMS instances is equal to the batch size of the sampled configuration. At the same time, a batch of sampled parameter configurations is distributed to different database instances, and each database instance is responsible for the performance evaluation of a parameter configuration. All cloned DBMS instances replay the same workload during the evaluation process, use a synchronous parallel evaluation mechanism, and the client waits for all workloads to be replayed and receives performance data, saves the batch of observations to the tuning history, and updates the tuning model.

[0052] It can be understood that between the parameter space decomposition, there is also a process of client environment preparation and preprocessing: using the client to interact with the user, accepting the optimization target (tuning target) and tuning load related information required by the user. At the same time, the client manages the database instance, performs load replay and collects database performance indicators to pass to the tuning model. The defined client needs to complete parameter allocation in parallel tuning mode, database instance cloning, and synchronous parallel parameter verification and other tasks.

[0053] That is, the method provided by the present application comprises:

[0054] 1. Client environment preparation and preprocessing

[0055] 1.1 The client server acts as the central controller of the tuning method, responsible for interacting with the user and managing the database instance, managing the load tool, etc. It first reads the database related information, including the database startup location, parameter configuration file location, parameter list and parameter detailed value range, etc.

[0056] 1.2 The client simultaneously accepts user tuning requests, including target tuning load, tuning indicators (transactions per second TPS or delay, etc.), and initializes the parameter evaluation environment and related configurations according to user requests, and replicates the database instance according to the user set tuning parallelism.

[0057] 2. Search space optimization

[0058] 2.1 Determine whether the parameter is continuous or categorical according to the order relationship between the parameter type and the parameter value and the performance, and construct a continuous parameter subspace from continuous parameters, and a categorical parameter subspace from categorical parameters.

[0059] 2.2 Map the high-dimensional parameter space to the low-dimensional agent space using a random mapping matrix. The above optimization algorithm will run in the low-dimensional agent space, and the parameter configuration explored in the low-dimensional agent space will be mapped back to the high-dimensional original space, so as to correctly obtain the performance of the parameter.

[0060] 2.3 Initialize the GP-based Bayesian optimization model in the low-dimensional continuous parameter subspace, and initialize the SMAC-based Bayesian optimization model in the low-dimensional classification parameter subspace, including defining the corresponding search space, and setting the parameters of the GP and SMAC models, etc.

[0061] 3. Communication mechanism based on context features

[0062] The GP model in the low-dimensional continuous space and the SMAC model in the low-dimensional classification space work through iterative optimization (see step 4 for details), that is, GP and SMAC search the low-dimensional continuous space and classification space alternately, each model runs for a fixed number of rounds (tuning interval), and when the tuning interval of one model ends, the optimal parameter value is passed to the next model as input as context features.

[0063] 4. Collaborative optimization process

[0064] 4.1 In the two low-dimensional agent spaces, 10 configurations are sampled using Latin hypercube sampling, and the performance of the 10 parameter configurations is tested by load replay, and the observation results (parameter configuration, context feature, and corresponding parameter performance) are stored in the tuning history.

[0065] 4.2 Train two agent models GP and SMAC using the initial observation data sampled from the continuous space and the discrete space respectively.

[0066] 4.3 Apply the optimal categorical parameters of the SMAC model in the initialization stage Map back to the original parameter range through the Project_Up interface And apply it to the database for performance testing. At the same time, the Encapsulated into context features F s , that is, the context features from the SMAC model.

[0067] 4.4 Start the tuning interval of GP surrogate model. Include: 1) First sample a batch of parameter configurations in the classification space using acquisition function based on local penalty. 2) Map all the configurations back to the original parameter range in the original space using Project_Up interface. 3) Perform load replay on different test instances in parallel, collect the performance corresponding to this batch of parameter configurations. 4) Store the test observation (parameter configuration, parameter performance, context feature) in this round into the tuning history. 5) Update the GP surrogate model using the data in the tuning history. 6) Repeat 1)-5) until the round of this tuning interval is reached, and return the optimal parameter configuration in the continuous subspace and the parameter configuration mapped back to the original space

[0068] 4.5 Apply the optimal parameter configuration searched to the database, and store into the context feature F g .

[0069] 4.6 Start the tuning interval of SMAC surrogate model. The SMAC tuning interval is consistent with the steps in 4-4, but uses SMAC as the surrogate model to search the discrete parameter subspace, and uses the context feature from GP to update the model.

[0070] 4.7 Repeat the tuning interval of GP and SMAC iteration until the preset tuning resource (reach the preset time or the preset number of tuning rounds) is exhausted.

[0071] It can be understood that the above process takes the SMAC model as an example to start the first search, and the GP model can also be used to start the first search, as shown in Figure 2 .

[0072] 5, Return the combination of the classification parameters and the continuous parameters corresponding to the round of optimal parameter performance in the tuning history as the final parameter configuration.

[0073] The embodiment of the present application provides a database parameter deployment method, comprising:

[0074] The optimal parameter configuration of the database is obtained by using the database parameter tuning method in any of the above embodiments.

[0075] If all the tuned parameters are dynamic parameters, the corresponding parameter values are written into the database through a SQL command, and if the tuned parameters also include static parameters, the corresponding parameter values are written into the database configuration file, and the database is restarted to complete parameter deployment.

[0076] The embodiment of the present application provides an electronic device, comprising: a computer readable storage medium and a processor. ​

[0077] The computer readable storage medium is configured to store executable instructions.

[0078] The processor is configured to read the executable instructions stored in the computer readable storage medium, and execute the database parameter tuning method according to any one of the above embodiments or the database parameter deployment method according to any one of the above embodiments.

[0079] The computer readable storage medium is configured to store executable instructions.

[0080] The computer readable storage medium is configured to store executable instructions.

[0081] Those skilled in the art can easily understand that the above description is only the preferred embodiment of the present application, and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A database parameter tuning method based on parameter type customization optimization, characterized in that: include: S1, decomposing the database parameter space into a continuous parameter subspace and a categorical parameter subspace according to the parameter value type, and mapping the continuous parameter subspace and the categorical parameter subspace to a low-dimensional proxy space, respectively, to obtain a continuous low-dimensional proxy space and a categorical low-dimensional proxy space; S2: Use the GP model to search for the configuration of continuous parameters in the continuous low-dimensional proxy space, and use the SMAC model to search for the configuration of categorical parameters in the categorical low-dimensional proxy space. The configurations of the continuous parameters and the categorical parameters are mapped back to the database parameter space for performance testing. The test results are stored in the tuning history until the preset number of tuning rounds is reached. The GP model and the SMAC model are searched alternately, and the optimal parameter configuration searched by each model in its own space is passed to the other model as a context feature. Before each model tuning interval begins, the optimal value currently searched by the other model is first applied to the database, and then the other model is switched to continue optimizing another subspace. Within each model tuning interval, the model will first obtain the function sampling parameter configuration, map it back to the high-dimensional original space to form the actual parameter value, apply it to the database to implement parameter performance evaluation, and use the parameters and their performance values ​​to update the model. In the process of implementing context-based model communication, when each model ends the current tuning interval, it obtains the performance test results of each parameter configuration and encapsulates the parameter configuration with the best performance currently searched as a context feature. This context feature records the latest tuning decision of the current model and will be passed to another model; the other model uses the context feature as input and trains to capture the correspondence between parameter values, context features, and parameter performance; S3: Use the categorical and continuous parameters with the best performance in the tuning history as the optimal parameter configuration for the database.

2. The method according to claim 1, wherein The kernel function of the GP model is: in, is the Martin kernel, used to measure continuous parameter configuration and The distance between is a Hamming kernel or linear kernel used to measure contextual features and The similarity between In the SMAC model, the historical configuration of classification parameters and context features are combined into one connection input.

3. The method according to claim 1, wherein In step S1, the continuous parameter subspace and the categorical parameter subspace are respectively mapped to the low-dimensional proxy space through the Project_Down() interface; In the Project_Down() interface, through Will is mapped to a low-dimensional proxy space, where is a continuous parameter in a continuous parameter subspace or a categorical parameter in a categorical parameter subspace, is a continuous parameter in a continuous low-dimensional proxy space or a categorical parameter in a categorical low-dimensional proxy space, S is a random mapping matrix, each column of which has only one non-zero value, which is 1 or -1; In step S2, the optimal configurations of the continuous parameters and the categorical parameters are mapped back to the database parameter space through the Project_Up() interface; In the Project_Up() interface, through Will Mapped back to the database parameter space; where, for the continuous parameters mapped back to the database parameter space ,pass Will Mapping back to the original parameter range yields , , They are The maximum and minimum values ​​of the original parameter range, `For The minimum value of the database parameter space; for the classification parameter mapped back to the database parameter space ,according to The original value category number of [-1,1] is evenly divided into the corresponding number of intervals, and according to The interval where the value of determines the corresponding original value category, thereby determining The original value of .

4. The method according to claim 1, wherein In step S2, a local penalty-based acquisition function is used to sample a batch of parameter configurations in a continuous low-dimensional proxy space or a categorical low-dimensional proxy space, and the batches are distributed to different clone database instances for parallel performance evaluation.

5. A database parameter deployment method, characterized in that: include: Obtaining the optimal parameter configuration of the database using the database parameter tuning method according to any one of claims 1 to 4; If all the adjusted parameters are dynamic parameters, the corresponding parameter values ​​are written to the database through SQL commands. If the adjusted parameters also include static parameters, the corresponding parameter values ​​are written to the database configuration file, and the database is restarted to complete the parameter deployment.

6. An electronic device, characterized in that: include: Computer-readable storage medium and processor; The computer-readable storage medium is used to store executable instructions; The processor is configured to read the executable instructions stored in the computer-readable storage medium and execute the database parameter tuning method according to any one of claims 1 to 4 or the database parameter deployment method according to claim 5.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to execute the database parameter tuning method according to any one of claims 1 to 4 or the database parameter deployment method according to claim 5.

8. A computer program product comprising a computer program or instructions, characterized in that The computer program or instruction is executed by a processor to perform the database parameter tuning method according to any one of claims 1 to 4 or the database parameter deployment method according to claim 5.

Images