A method, device, equipment and storage medium for selecting mail storage nodes

Abstract

Embodiments of the present application provide a mail storage node selection method, apparatus, device and storage medium. The method includes using a storage prediction model to obtain a health prediction value of a first storage node, and screening a second storage node based on the health prediction value of the first storage node and a first threshold. Then, the number of storage nodes and the number of storage volumes of the second storage node, the sequence of incoming letter weights and the sequence of incoming letter times of the storage volumes in the second storage node are input into a floating-point calculation unit for a one-time calculation to obtain the storage exclusion degree of the second storage node. The second storage nodes are sorted according to their storage exclusion degrees, and the second storage node with the smallest storage exclusion degree is selected as the incoming letter node. In this way, the efficiency, stability and resource utilization of the mail storage system can be effectively improved.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of communications, and in particular, to a mail storage node selection method, apparatus, device, and storage medium. Background Art

[0002] With the widespread application of email systems and the growing demand for email storage, the research and optimization of email storage technology in large email systems has become increasingly important. At present, in large email systems, tens of thousands of new emails flow into the system every day, and each email needs to be assigned to a suitable storage node to ensure its security and reliability. However, when the number of storage nodes reaches hundreds and the number of storage volumes on each node reaches dozens, the traditional email storage node selection method relies on complex computing logic. Every time a new email arrives, it needs to traverse all storage nodes and evaluate its available space, load and other factors before selecting the most suitable node. This leads to a significant increase in the amount of calculation and a sharp increase in computational complexity, which seriously affects the operating efficiency of the entire system and may even cause email storage delays and affect user experience.

[0003] Therefore, when there are a large number of nodes and complex storage requirements, how to quickly and evenly select the storage nodes for new emails becomes an urgent problem to be solved. Summary of the invention

[0004] According to the embodiments of the present application, a mail storage node selection method, apparatus, device and storage medium are provided, which can effectively improve the efficiency, stability and resource utilization of the mail storage system, and realize the effective management and optimization of the mail storage nodes.

[0005] In a first aspect of the present application, a method for selecting a mail storage node is provided. The method comprises:

[0006] Acquire data information of the first storage node, input the data information of the first storage node into a storage prediction model, and acquire a health prediction value of the first storage node;

[0007] Selecting a second storage node according to the health prediction value of the first storage node and a first threshold;

[0008] Obtain the number of storage nodes and storage volumes of the second storage node, and the credit weight sequence and credit number sequence of the storage volumes in the second storage node;

[0009] The number of storage nodes and storage volumes of the second storage node, the sequence of credit weights and the sequence of credit times of the storage volumes in the second storage node are input into the floating point calculation unit for one-time calculation to obtain the storage exclusion degree of the second storage node;

[0010] The second storage nodes are sorted according to their storage exclusion degrees, and the second storage node with the smallest storage exclusion degree is selected as the incoming node.

[0011] In a possible implementation, the storage prediction model is a time series model pre-trained by historical data information of the first storage node;

[0012] The historical data information of the first storage node includes the storage load, network delay and error rate of the first storage node.

[0013] In a possible implementation, the calculation formula of the floating-point calculation unit is:

[0014] ,

[0015] in, is the storage exclusion degree of the i-th storage node, is the credit weight sequence of the storage volume in the storage node, The number of times a storage volume in a storage node is received.

[0016] Optionally, the trust weight sequence of the storage volume is a set of inverses of the trust weights of all storage volumes in the storage node, which can be expressed as:

[0017] ,

[0018] in, is the credit weight of the jth storage volume in the ith storage node, and N is the number of storage volumes in the storage node.

[0019] In a possible implementation, the initial credit weight sequence of the storage volume is generated randomly;

[0020] The incoming letter weight sequence is adaptively adjusted based on the initial incoming letter weight according to the read and write frequency of the storage volume, the matching degree between the storage volume type and the email.

[0021] In a possible implementation, the floating-point computing unit is composed of an AVX instruction set dedicated to floating-point sequence calculations of the x86-64 architecture.

[0022] In a possible implementation, the method further includes:

[0023] Select the storage node with the second smallest storage exclusion as the backup node;

[0024] When an incoming node fails, the mail is automatically stored in a backup node.

[0025] In a second aspect of the present application, a mail storage node selection device is provided. The device comprises:

[0026] A prediction module, used to obtain data information of the first storage node, input the data information of the first storage node into a storage prediction model, and obtain a health prediction value of the first storage node;

[0027] A screening module, configured to screen a second storage node according to a health prediction value of the first storage node and a first threshold;

[0028] An acquisition module, used to acquire the number of storage nodes and storage volumes of the second storage node, and a sequence of credit weights and a sequence of credit times of the storage volumes in the second storage node;

[0029] A calculation module, used for inputting the number of storage nodes and storage volumes of the second storage node, the sequence of credit weights and the sequence of credit times of the storage volumes in the second storage node into a floating point calculation unit for one-time calculation to obtain the storage exclusion degree of the second storage node;

[0030] The selection module is used to sort the second storage nodes according to their storage exclusion degrees, and select the second storage node with the smallest storage exclusion degree as the incoming node.

[0031] In a third aspect of the present application, an electronic device is provided, which includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the program, the method described above is implemented.

[0032] In a fourth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the method according to the first aspect of the present application is implemented.

[0033] The embodiment of the present application provides a mail storage node selection method, which obtains the data information of the first storage node, inputs the data information of the first storage node into the storage prediction model, and obtains the health prediction value of the first storage node. The second storage node is screened according to the health prediction value of the first storage node and the first threshold. The number of storage nodes and the number of storage volumes of the second storage node, the letter weight sequence and the letter number sequence of the storage volume in the second storage node are obtained. The number of storage nodes and the number of storage volumes of the second storage node, the letter weight sequence and the letter number sequence of the storage volume in the second storage node are input into the floating point calculation unit for a one-time calculation to obtain the storage exclusion degree of the second storage node. The second storage nodes are sorted according to the storage exclusion degree of the second storage node, and the second storage node with the smallest storage exclusion degree is selected as the letter node. In other words, the health prediction value of the first storage node is obtained by using the storage prediction model, and the second storage node is screened according to the health prediction value of the first storage node and the first threshold, so that the storage nodes with poor status can be effectively and accurately screened out. Then, based on the number of storage nodes and storage volumes of the second storage node, the credit weight sequence and credit number sequence of the storage volumes in the second storage node, the storage exclusion degree of the storage node is calculated in real time and once through the floating-point computing unit, which effectively balances the load of each storage node and avoids the situation where some storage nodes are overloaded while other nodes are idle. In addition, the use of the floating-point computing unit can calculate the storage exclusion degree of the storage node at one time, without the need to repeatedly perform multiplication and then addition in sequence, which cleverly improves the calculation speed of the storage exclusion degree.

[0034] It should be understood that the contents described in the Summary of the Invention are not intended to limit the key or important features of the embodiments of the present application, nor are they intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0036] Figure 1 A flowchart of a mail storage node selection method provided in an embodiment of the present application;

[0037] Figure 2 is a schematic diagram of the structure of a storage prediction model according to an embodiment of the present application;

[0038] Figure 3 A block diagram of a mail storage node selection device according to an embodiment of the present application;

[0039] Figure 4It is a schematic diagram of the structure of a terminal device or a server suitable for implementing the embodiments of the present application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0041] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0042] Figure 1 FIG. 1 is a flowchart of a mail storage node selection method according to an embodiment of the present disclosure. Figure 1 , the method comprising:

[0043] S101, obtaining data information of a first storage node, inputting the data information of the first storage node into a storage prediction model, and obtaining a health prediction value of the first storage node.

[0044] Among them, in order to effectively avoid the delays and system burdens caused by the real-time calculation of the storage prediction model, the acquisition of the health prediction value of a storage node is not real-time, but the prediction period is customized according to the needs of the operation and maintenance personnel, so as to maximize the advantages of the storage prediction model in the intelligent selection of mail storage nodes.

[0045] In this embodiment, the storage prediction model is used to accurately predict the health of the first storage node, which provides a basis for the subsequent screening of the second storage node and enhances the reliability of the mail node selection.

[0046] Optionally, the storage prediction model is a time series model pre-trained by historical data information of the first storage node;

[0047] The historical data information of the first storage node includes the storage load, network delay and error rate of the first storage node.

[0048] The historical data information of the first storage node may be as shown in Table 1:

[0049] Table 1

[0050] Timestamp Storage Load (GB) Network latency (ms) Error rate (%) Health Prediction Value Label 2024.2.15 12:00 50 10 0.2 65 2024.2.15 14:30 52 12 0 85 2024.2.15 15:00 48 9 0.1 70 ... ... ... ... ...

[0051] Figure 2 is a schematic diagram of the structure of a storage prediction model according to an embodiment of the present application, such as Figure 2 As shown:

[0052] The storage prediction model includes an input layer, an LSTM layer, a fully connected layer, and an output layer. The input layer is responsible for preprocessing the historical data information of the first storage node and standardizing the historical data information. The LSTM layer (Long Short-term Memory Networks) is responsible for capturing and learning the preprocessed historical data information and obtaining the feature data in the historical data information. The fully connected layer is responsible for mapping the feature data output by the LSTM layer to a lower-dimensional space and performing nonlinear transformations to prepare for the final prediction results. The output layer uses an activation function to predict the mapped data output by the fully connected layer to obtain the final health prediction value of the first storage node. Among them, the batch size batch_size of the storage prediction model during training is set to 32.

[0053] In this embodiment, the storage prediction model is pre-trained using the historical data information of the first storage node, so that the storage prediction model can capture the operation rules and trends of the mail storage node, predict its future health status, and provide a reliable basis for the selection and management of subsequent storage nodes.

[0054] S102: Filter a second storage node according to the health prediction value of the first storage node and a first threshold.

[0055] The first threshold is a preset threshold that can be customized as needed.

[0056] For example, when the health prediction value of the first storage node is 85, the first threshold is set to 80, and the health prediction value of the first storage node exceeds the first threshold, then it can be used as the second storage node. When the health prediction value of the first storage node is 75, the first threshold is set to 80, and the health prediction value of the first storage node does not exceed the first threshold, there is a risk of storage failure, and it cannot be used as the second storage node.

[0057] In this embodiment, by preliminarily screening the first storage nodes to obtain second storage nodes with higher health prediction values, the health status of the first storage nodes is taken into consideration, thereby reducing the risk of storage failure.

[0058] S103, obtaining the number of storage nodes and storage volumes of the second storage node, and the credit weight sequence and credit number sequence of the storage volumes in the second storage node.

[0059] Among them, the storage volume's incoming mail weight can be customized and modified by the operation and maintenance engineer according to actual needs. When the email is successfully stored in the storage volume in the storage node, the number of incoming mails of the storage volume increases by one. In addition, the number of incoming mails of a storage volume is inversely proportional to the possibility that the storage node to which it belongs is selected as the incoming mail node, and the storage volume's incoming mail weight is proportional to the possibility that the storage node to which it belongs is selected as the incoming mail node. Combining the dual effects of the number of incoming mails and the incoming mail weight, it is possible to achieve a reasonable selection of the second storage node and improve the overall performance and stability of the system.

[0060] In this embodiment, the number of storage nodes and storage volumes of the second storage node are taken into consideration, and the incoming letter weight and the number of incoming letters of the storage volume are also combined. This comprehensive evaluation method helps to avoid selecting nodes that may be overloaded or have poor performance, thereby improving the stability and reliability of the overall mail system.

[0061] S104, input the number of storage nodes and storage volumes of the second storage node, the credit weight sequence and credit number sequence of the storage volumes in the second storage node into the floating point calculation unit for one-time calculation to obtain the storage exclusion degree of the second storage node.

[0062] In this embodiment, a floating point calculation unit is used to perform a one-time complex calculation, which improves the calculation efficiency, can quickly complete the selection process of the mail storage node, reduces the delay of mail storage, and improves the response speed of the mail system.

[0063] Optionally, the calculation formula of the floating point calculation unit is:

[0064] ,

[0065] in, is the storage exclusion degree of the i-th second storage node, is the credit weight sequence of the storage volume in the second storage node, It is a sequence of the number of times the storage volume in the second storage node is received.

[0066] For example, there are 100 storage nodes in the mail system, each storage node has 10 storage volumes, and the credit weight sequence of the storage volume in the 10th storage node is , the number of letters received by the storage volume in the 10th storage node is , then the one-time calculation formula of the storage exclusion degree of the 10th storage node in the floating-point computing unit is, The AVX instruction set simultaneously loads the storage volume's entry weight sequence and entry number sequence, performs a one-time multiplication and cumulative summation calculation, and quickly calculates the storage exclusion degree of each storage node through parallel computing, providing a basis for the subsequent selection of the optimal entry node.

[0067] In this embodiment, a credit weight is defined for each storage volume in the second storage node, so that the floating-point computing unit can perform global optimization allocation when calculating storage exclusion.

[0068] Optionally, the trust weight sequence of the storage volume is a set of inverses of the trust weights of all storage volumes in the storage node, which can be expressed as:

[0069] ,

[0070] in, is the credit weight of the jth storage volume in the ith storage node, and N is the number of storage volumes in the storage node.

[0071] In this embodiment, the inverse of the storage volume's credit weight is used to form a storage volume credit weight sequence. For example, storage volumes with low credit weights are often used less, so their inverses are larger, which can increase the probability of the storage volume being selected in the calculation, thereby achieving balanced utilization of resources.

[0072] Optionally, the initial credit weight sequence of the storage volume is generated randomly;

[0073] The incoming letter weight sequence is adaptively adjusted based on the initial incoming letter weight according to the read and write frequency of the storage volume, the matching degree between the storage volume type and the email.

[0074] For example, the initial credit weight sequence of the four randomly generated storage volumes in storage volume A is [0.2, 0.4, 0.1, 0.3], and the read and write frequencies are [1, 2, 3, 4]. The read and write sequence obtained by normalizing the read and write frequencies is [0.1, 0.2, 0.3, 0.4]. The credit weight adjusted according to the read and write frequencies is [0.2, 0.4, 0.1, 0.3] + (1-[0.1, 0.2, 0.3, 0.4]) = [1.1, 1.2, 0.8, 0.9]. In addition, the types of storage volumes can be divided into three levels. The first-level storage volume is used to store frequently read email data, the second-level storage volume is used to store emails containing large files (such as video and audio), and the third-level storage volume is used to store transactional emails. If the storage volume types are [primary, secondary, secondary, tertiary], and the type of email to be received is an email containing large files (such as video or audio), then add 1 / N to the corresponding storage volume incoming weight, where N is the number of storage volumes, that is, 1 / 4=0.25. After adjusting the matching degree between the storage volume type and the email, the incoming weight is [1.1, 1.2, 0.8, 0.9] + [0, 0.25, 0.25, 0] = [1.1, 1.45, 1.05, 0.9].

[0075] In this embodiment, the initial weight of incoming mail is randomly generated to avoid the deviation caused by human setting. Then, the weight of incoming mail is adaptively adjusted according to the read and write frequency of the storage volume, the matching degree between the storage volume type and the mail, so that the weight of incoming mail more accurately reflects the real performance and applicability of the storage volume.

[0076] Optionally, the floating-point computing unit is composed of an AVX instruction set dedicated to floating-point sequence calculations of the x86-64 architecture.

[0077] Among them, AVX (Advanced Vector Extensions) is a set of advanced single instruction multiple data SIMD (Single Instruction Multiple Data) instruction set extensions introduced by Intel and AMD in the x86-64 architecture, which aims to improve the performance of processors in parallel computing, especially in the field of floating-point computing. AVX supports 128-bit, 256-bit and even 512-bit register operations, and can process multiple data at a time, thereby significantly accelerating matrix operations, vector operations and other tasks requiring high-performance computing. On the x86-64 architecture, the AVX instruction set can process multiple floating-point calculation tasks at a time, significantly improving the speed and efficiency of data processing. In non-x86-64 architectures or other heterogeneous computing platforms, the corresponding SIMD instruction set or parallel computing technology can also be used to achieve similar data parallel acceleration effects, which are widely used in data mining, image processing and other fields. However, in the present invention, the AVX instruction set is applied to the scenario of mail storage node selection, giving full play to its advantages in floating-point calculations and improving the selection efficiency of storage nodes.

[0078] In this embodiment, the AVX instruction set can simultaneously calculate the letter-entry weights and letter-entry times of multiple storage volumes, and use AVX to calculate the calculation results in one go, rather than multiplying first and then accumulating, thereby improving calculation efficiency.

[0079] S105 , sorting the second storage nodes according to their storage exclusion degrees, and selecting the second storage node with the smallest storage exclusion degree as the incoming node.

[0080] Furthermore, after selecting the incoming node, the incoming node can randomly select an available storage volume to store the email, thereby avoiding complex storage volume selection algorithms and simplifying the decision-making process. In addition, if a storage volume in a storage node fails, this random selection strategy can also quickly switch to other available storage volumes, improving the fault tolerance of the email storage node selection system. Alternatively, the selection can be further optimized based on the incoming weight and number of incoming letters of the storage volume. Unlike the storage exclusion calculation method of the storage node, the selection of the storage volume does not require the accumulation of the incoming weights of all storage volumes, but can be evaluated and selected based on the incoming weight and number of incoming letters of a single storage volume. Operation and maintenance personnel can make more targeted choices by flexibly controlling the incoming weight of the storage volume.

[0081] In this embodiment, by selecting the second storage node with the smallest storage exclusion degree as the incoming node, the load of each storage node can be effectively balanced, thereby improving resource utilization.

[0082] Optionally, the method further comprises:

[0083] Select the storage node with the second smallest storage exclusion as the backup node;

[0084] When an incoming node fails, the mail is automatically stored in a backup node.

[0085] In this embodiment, by selecting a suitable standby node to implement automatic failover, the failure of the mail storage node can be effectively dealt with, and the continuity of the mail storage service and the security of the data can be ensured.

[0086] According to the embodiments of the present disclosure, the following technical effects are achieved:

[0087] 1. Through the multi-layer selection mechanism, while improving the reliability of mail storage node selection, it realizes the selection of appropriate storage nodes more intelligently and efficiently, and improves the accuracy of decision-making.

[0088] 2. By calculating the storage exclusion degree of the second storage node, a reasonable estimate is made of the load of each storage node, which effectively balances the load of each storage node, effectively avoiding the situation where some storage nodes are overloaded while other storage nodes are idle, and improving resource utilization.

[0089] 3. The floating-point computing unit is used to perform parallel calculations on multiple floating-point number sequences at the same time, which reduces the number of loops and the overhead of step-by-step operations, greatly saves calculation time, and cleverly improves the calculation speed of storage exclusion.

[0090] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present application.

[0091] The above is an introduction to the method embodiment. The following is a further explanation of the scheme described in this application through an apparatus embodiment.

[0092] Figure 3 A block diagram of a mail storage node selection device according to an embodiment of the present application is shown. Figure 3 Shown include:

[0093] The prediction module 301 is used to obtain data information of the first storage node, input the data information of the first storage node into the storage prediction model, and obtain the health prediction value of the first storage node;

[0094] A screening module 302, configured to screen a second storage node according to a health prediction value of the first storage node and a first threshold;

[0095] The acquisition module 303 is used to acquire the number of storage nodes and storage volumes of the second storage node, and the credit weight sequence and credit number sequence of the storage volumes in the second storage node;

[0096] The calculation module 304 is used to input the number of storage nodes and the number of storage volumes of the second storage node, the letter weight sequence and the letter number sequence of the storage volumes in the second storage node into the floating point calculation unit for one-time calculation to obtain the storage exclusion degree of the second storage node;

[0097] The selection module 305 is used to sort the second storage nodes according to their storage exclusion degrees, and select the second storage node with the smallest storage exclusion degree as the incoming node.

[0098] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0099] Figure 4 A schematic diagram of the structure of a terminal device or server suitable for implementing an embodiment of the present application is shown.

[0100] like Figure 4As shown, the terminal device or server includes a central processing unit (CPU) 401, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 402 or the program loaded from the storage part 408 to the random access memory (RAM) 403. In the RAM 403, various programs and data required for the operation of the terminal device or server are also stored. The CPU 401, the ROM 402, and the RAM 403 are connected to each other through a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0101] The following components are connected to the I / O interface 405: an input section 406 including a keyboard, a mouse, etc.; an output section 407 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN card, a modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the I / O interface 405 as needed. A removable medium 411, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 410 as needed, so that a computer program read therefrom is installed into the storage section 408 as needed.

[0102] In particular, according to an embodiment of the present application, the above method flow steps can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a machine-readable medium, and the computer program includes a program code for executing the method shown in the flow chart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 409, and / or installed from the removable medium 411. When the computer program is executed by the central processing unit (CPU) 401, the above-mentioned functions defined in the system of the present application are executed.

[0103] It should be noted that the computer-readable medium shown in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0104] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of the code, and the aforementioned module, program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0105] The units or modules involved in the embodiments described in the present application may be implemented by software or hardware. The units or modules described may also be arranged in a processor. The names of these units or modules do not, in some cases, constitute limitations on the units or modules themselves.

[0106] As another aspect, the present application further provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable storage medium stores one or more programs, and when the above programs are used by one or more processors to execute the method described in the present application.

[0107] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of application involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the aforementioned application concept. For example, the above features are replaced with (but not limited to) technical features with similar functions applied in the present application.

Claims

1. A method for selecting a mail storage node, characterized in that: include: Acquire data information of a first storage node, input the data information of the first storage node into a storage prediction model, and acquire a health prediction value of the first storage node; Selecting a second storage node according to the health prediction value of the first storage node and a first threshold; Acquire the number of storage nodes and storage volumes of the second storage node, and the sequence of credit weights and credit times of the storage volumes in the second storage node; The number of storage nodes and storage volumes of the second storage node, the sequence of credit weights and the sequence of credit times of the storage volumes in the second storage node are input into a floating point calculation unit for one-time calculation to obtain the storage exclusion degree of the second storage node; The second storage nodes are sorted according to their storage exclusion degrees, and the second storage node with the smallest storage exclusion degree is selected as the incoming node.

2. The mail storage node selection method according to claim 1, characterized in that: The storage prediction model is a time series model pre-trained by historical data information of the first storage node; The historical data information of the first storage node includes the storage load, network delay and error rate of the first storage node.

3. The mail storage node selection method according to claim 1, characterized in that: The calculation formula of the floating point calculation unit is: ， in, is the storage exclusion degree of the i-th second storage node, is the trusted weight sequence of the storage volume in the i-th second storage node, is the sequence of the number of times the storage volume in the i-th second storage node is received.

4. The mail storage node selection method according to claim 3, characterized in that: The trust weight sequence of the storage volume is a set of inverses of the trust weights of all storage volumes in the second storage node, which can be expressed as: ， in, is the trust weight of the jth storage volume in the ith second storage node, and N is the number of storage volumes in the second storage node.

5. The mail storage node selection method according to claim 4, characterized in that: The initial credit weight sequence of the storage volume is generated randomly; The incoming letter weight sequence is adaptively adjusted based on the initial incoming letter weight according to the read and write frequency of the storage volume, the matching degree between the storage volume type and the mail.

6. The mail storage node selection method according to claim 1, characterized in that: The floating point calculation unit is composed of an AVX instruction set dedicated to floating point sequence calculation of the x86-64 architecture.

7. The mail storage node selection method according to claim 1, characterized in that: The method further comprises: Selecting a second storage node with the second smallest storage exclusion degree as a backup node; When the incoming node fails, the mail is automatically stored in the standby node.

8. A mail storage node selection device, characterized in that: include: A prediction module, used to obtain data information of a first storage node, input the data information of the first storage node into a storage prediction model, and obtain a health prediction value of the first storage node; A screening module, configured to screen a second storage node according to the health prediction value of the first storage node and a first threshold; An acquisition module, used to acquire the number of storage nodes and storage volumes of the second storage node, and a sequence of credit weights and a sequence of credit times of the storage volumes in the second storage node; A calculation module, used for inputting the number of storage nodes and storage volumes of the second storage node, the sequence of credit weights and the sequence of credit times of the storage volumes in the second storage node into a floating point calculation unit for one-time calculation to obtain the storage exclusion degree of the second storage node; The selection module is used to sort the second storage nodes according to their storage exclusion degrees, and select the second storage node with the smallest storage exclusion degree as the incoming node.

9. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images