WO2021142971A1 - 传输速率控制方法、装置、计算机***及可读存储介质 - Google Patents
传输速率控制方法、装置、计算机***及可读存储介质 Download PDFInfo
- Publication number
- WO2021142971A1 WO2021142971A1 PCT/CN2020/087614 CN2020087614W WO2021142971A1 WO 2021142971 A1 WO2021142971 A1 WO 2021142971A1 CN 2020087614 W CN2020087614 W CN 2020087614W WO 2021142971 A1 WO2021142971 A1 WO 2021142971A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- server
- rate
- network model
- parameters
- database
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/25—Flow control; Congestion control with rate being modified by the source upon detecting a change of network conditions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/147—Network analysis or design for predicting network behaviour
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0876—Network utilisation, e.g. volume of load or congestion level
- H04L43/0894—Packet rate
Definitions
- This application relates to the field of artificial intelligence technology, and in particular to a transmission rate control method, device, computer system, and readable storage medium.
- the current mechanism is to set a fixed threshold to specify the source The data transfer rate between the database and the target database.
- this method only limits a certain safe rate range to avoid the server from being unable to operate normally due to data transmission occupying too much resources; however, the transmission rate is not adjusted based on the server's operating conditions. , So it is easy to cause the problem that the current server operating condition is poor but the transmission rate is too high due to the high threshold and the server crashes, or the current server operating condition is good but the transmission rate is too small due to the low threshold and the server transmission efficiency is reduced. happen.
- the purpose of this application is to provide a transmission rate control method, device, computer system, and readable storage medium, which are used to solve the problem that the current server operating condition in the prior art is poor but the transmission rate is too high and the server crashes due to the high threshold. , Or the current server operating condition is good, but the transmission rate is too low due to the low threshold, resulting in a problem that the server transmission efficiency is reduced.
- this application provides a transmission rate control method, including:
- the basic parameters are predicted to obtain the best matching rate through a preset mature network model; wherein the mature network model is used to predict the maximum transmission rate allowed between the source database and the target database according to the basic parameters, and the transmission rate Is the best matching rate between the current source database and the target database;
- the transmission rate of data transmission between the source database and the target database is adjusted to the best matching rate.
- the present application also provides a transmission rate control device, including:
- the parameter extraction module is used to periodically obtain the basic parameters of the server where the source database is located and the server where the target database is located; wherein the basic parameters represent the hardware conditions, environmental conditions, and operating conditions of the server and its database;
- the rate calculation module is used to predict the basic parameters to obtain the best matching rate through a preset mature network model; wherein, the mature network model is used to predict the maximum allowable rate between the source database and the target database according to the basic parameters.
- Transmission rate which is the best matching rate between the current source database and the target database;
- the rate adjustment module is used to adjust the transmission rate of data transmission between the source database and the target database to the optimal matching rate.
- the present application also provides a computer system, which includes a plurality of computer devices, each computer device includes a memory, a processor, and a computer program stored in the memory and running on the processor, the multiple computers
- the processor of the device executes the computer program, the steps of the above transmission rate control method are jointly implemented.
- the present application also provides a computer-readable storage medium, which includes multiple storage media, each of which stores a computer program, and when the computer program stored in the multiple storage media is executed by a processor
- the steps of the above transmission rate control method are jointly implemented.
- the transmission rate control method, device, computer system and readable storage medium provided in this application comprehensively evaluate the maximum transmission rate that can be tolerated between the current source database and the target database from multiple dimensions through basic parameters, so as to avoid current threshold adjustments.
- the server is adjusted to the data transmission rate with the best matching rate as the goal, so that the data transmission is guaranteed to the server.
- the fastest transmission rate can be obtained; at the same time, the best matching rate is periodically obtained and the transmission rate of the server is adjusted according to the best matching rate, which realizes the technical effect of dynamically adjusting the transmission rate between two servers , Continuously ensuring that the data transmission can obtain the technical effect of the fastest transmission rate under the premise of ensuring the stability of the server, and continuously ensuring the high efficiency and stability of the data transmission operation.
- FIG. 1 is a flowchart of Embodiment 1 of a transmission rate control method according to this application;
- FIG. 2 is a schematic diagram of the environmental application of the transmission rate control method in Embodiment 2 of the transmission rate control method of this application;
- FIG. 3 is a flow chart of a specific method for transmission rate control provided in Embodiment 2 of the transmission rate control method of this application;
- Fig. 4 is a specific method flow chart of the mature network model provided in the second embodiment of the transmission rate control method of the present application through the following steps;
- FIG. 5 is a flowchart of a specific method for adjusting a transmission rate according to the optimal matching rate according to the second embodiment of the transmission rate control method of the present application;
- FIG. 6 is a schematic diagram of program modules of Embodiment 3 of the transmission rate control device of this application.
- FIG. 7 is a schematic diagram of the hardware structure of the computer equipment in the fourth embodiment of the computer system of this application.
- the transmission rate control method, device, computer system, and readable storage medium provided in this application are suitable for the field of big data and provide a transmission rate control method based on a parameter extraction module, a rate calculation module, a rate adjustment module, and a connection establishment module .
- This application periodically obtains the basic parameters of the server where the source database is located and the server where the target database is located; calculates the basic parameters through a preset mature network model to obtain the best matching rate; and adjusts the source according to the best matching rate.
- the data transfer rate between the database and the target database In order to avoid the current problem of server crash due to excessive transmission rate caused by threshold adjustment, or lower transmission efficiency of server due to excessive transmission rate, the data transmission can be the fastest under the premise of ensuring the stability of the server.
- the transmission rate, as well as the technical effect of dynamically adjusting the transmission rate between servers continuously ensure that the data transmission can obtain the fastest transmission rate technical effect under the premise of ensuring the stability of the server, and continue to ensure Efficient and stable data transmission operation.
- a transmission rate control method of this embodiment includes:
- S102 Periodically obtain the basic parameters of the server where the source database is located and the server where the target database is located; wherein the basic parameters represent the hardware conditions, environmental conditions, and operating conditions of the server and its database;
- S103 Predict the basic parameters to obtain the best matching rate through a preset mature network model; wherein, the mature network model is used to predict the maximum transmission rate allowed between the source database and the target database according to the basic parameters.
- the transmission rate is the best matching rate between the current source database and the target database;
- S104 Adjust the transmission rate of data transmission between the source database and the target database to the optimal matching rate.
- the process of transferring data from the source database to the target database is achieved by establishing a communication connection between the server where the source database is located and the server where the target database is located; in order to be able to dynamically and comprehensively understand the server where the source database is located and the target database
- the operating state of the server where it is located periodically obtains the basic parameters of the server.
- the data transmission rate of the two servers is adjusted with the best matching rate as the target, so that the data transmission can obtain the fastest transmission rate under the premise of ensuring the stability of the two servers.
- steps S102-S104 are performed periodically, it is ensured that the basic parameters of the server are cyclically detected and calculated to obtain the best matching rate, and the transmission rate between the two servers is adjusted according to the best matching rate, realizing dynamic
- the technical effect of adjusting the transmission rate between two servers continuously ensures that the data transmission can obtain the technical effect of the fastest transmission rate under the premise of ensuring the stability of the server, and continues to ensure the efficiency of data transmission operations stability.
- the mature network model is the DeepFM model with the ReLU function as the activation function. Because it learns both low-order feature combinations and high-order feature combinations at the same time, it can take into account the different data types in the basic parameters. The data is better adapted to the application scenarios of transmission rate matching; and the ReLU function is used as the activation function to create the sparsity of the neural network, reduce the interdependence between the data in the basic parameters, and alleviate the over-simulation The occurrence of the joint problem.
- This embodiment is a specific application scenario of the foregoing Embodiment 1. Through this embodiment, the method provided by this application can be described more clearly and specifically.
- the basic parameters of the server where the source database is located and the server where the target database is located can be periodically obtained, and the optimal matching rate obtained by calculation can be used to adjust the data between the source database and the target database.
- the transfer rate of the transfer can be used to adjust the data between the source database and the target database.
- Fig. 2 schematically shows an environmental application diagram of the transmission rate control method according to the second embodiment of the present application.
- a communication connection is established between the server 4 where the source database is located and the server 5 where the target database is located, and the server 2 where the transmission rate control method is located is connected to the server 4 where the source database is located and the server 5 where the target database is located through the network 3.
- the server 2 can provide services through one or more networks 3.
- the network 3 can include various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, and proxy devices. And/or wait.
- the network 3 may include physical links, such as coaxial cable links, twisted pair cable links, optical fiber links, combinations thereof, and/or the like.
- the network 3 may include wireless links, such as cellular links, satellite links, Wi-Fi links and/or the like;
- the server 2, the server 4 where the source database is located, and the server 5 where the target database is located may be composed of a single or multiple computer devices (e.g., servers).
- the single or multiple computing devices may include virtualized computing instances.
- Virtualized computing instances may include virtual machines, such as computer system simulations, operating systems, servers, and so on.
- the computing device may load the virtual machine based on a virtual image and/or other data defining specific software (eg, operating system, dedicated application, server) for simulation. As the demand for different types of processing services changes, different virtual machines can be loaded and/or terminated on one or more computing devices.
- a hypervisor can be implemented to manage the use of different virtual machines on the same computing device.
- FIG. 3 is a flowchart of a specific method for transmission rate control provided by an embodiment of the present application. The method specifically includes steps S201 to S204.
- S201 Establish a connection with the server where the source database is located and the server where the target database is located respectively; wherein, establish a communication connection with the server through TCP or UDP.
- this step by establishing a communication connection between the server where the source database is located and the server where the target database is located, the technical effect of transmitting the data in the source database to the target database is realized; respectively establish with the monitoring system of the server where the source database is located and the server where the target database is located. Connect in order to obtain the basic parameters of the server where the source database is located and the server where the target database is located, as well as the data transfer rate between the source database and the target database.
- TCP Transmission Control Protocol
- IETF RFC 793 is a connection-oriented, reliable, byte stream-based transport layer communication protocol, which is defined by IETF RFC 793.
- UDP User Datagram Protocol
- the TCP layer is an intermediate layer located above the IP layer and below the application layer.
- the application layers of different hosts often need reliable, pipe-like connections, but the IP layer does not provide such a flow mechanism, but provides unreliable packet exchange.
- UDP is the abbreviation of User Datagram Protocol.
- the Chinese name is User Datagram Protocol.
- It is a connectionless transport layer protocol in the OSI (Open System Interconnection) reference model. It provides transaction-oriented simple and unreliable information transmission services.
- IETF RFC 768 is the formal specification of UDP.
- the monitoring system is a computer system for real-time monitoring of the hardware conditions, environmental conditions, and operating conditions of the server and its database.
- S202 Periodically obtain the basic parameters of the server where the source database is located and the server where the target database is located; the basic parameters represent the hardware conditions, environmental conditions, and operating conditions of the server and its database.
- the process of transferring data from the source database to the target database is achieved by establishing a communication connection between the server where the source database is located and the server where the target database is located; in order to be able to dynamically and comprehensively understand the server where the source database is located and the target database
- the operating state of the server where it is located periodically obtains the basic parameters of the server.
- the basic parameters include at least hardware parameters, software parameters, and service parameters.
- the hardware parameters include at least the server's CPU, memory, network RTT, and ACK; and the software parameters include at least the database in the server.
- the service parameters include at least the service response time of the server; among them, the hardware parameters are used to characterize the hardware conditions of the server itself, and the software parameters are used to characterize the environment the server is currently facing Status, the service parameters are used to characterize the operating status of the database in the server; therefore, the current operating status of the server and its database can be obtained from the three dimensions of hardware conditions, environmental conditions, and business environments by acquiring hardware parameters, software parameters, and business parameters.
- the basic parameters may also include a version type
- the version type includes at least the disk type of the server, the server version, and the driver version of the connected data source.
- the basic parameters can be obtained through the monitoring system in the server. Therefore, the method of obtaining the basic parameters is not described here.
- S203 Predict the basic parameters through a preset mature network model to obtain the best matching rate; wherein the mature network model is used to predict the maximum transmission rate allowed between the source database and the target database according to the basic parameters.
- the transmission rate is the best matching rate between the current source database and the target database.
- the mature network model is the DeepFM model with the ReLU function as the activation function. Because it learns both low-order feature combinations and high-order feature combinations at the same time, it can learn the data of different data types in the basic parameters at the same time, so as to better Adapt to the application scenario of transmission rate matching; and, use the ReLU function as the activation function to create the sparsity of the neural network, reduce the interdependence between the data in the basic parameters, and alleviate the occurrence of over-fitting problems.
- the mature network model is obtained through training in the following steps:
- S203-1 Obtain an initial neural network model and at least one training sample, where the training sample includes training parameters and training rate.
- the training parameters refer to the historical operating conditions of the server where the source database is located and the server where the target database is located;
- the training rate refers to the historical operating conditions of the server where the source database is located and the server where the target database is located based on the training parameters.
- the maximum transmission rate that can be allowed under the premise of ensuring the stable operation of the server.
- the training parameters include at least hardware parameters, software parameters, and service parameters.
- the hardware parameters include at least CPU, memory, network RTT, and ACK of the server
- the software parameters include at least the number of connections to the database in the server. SQL response index, load pressure, service response time length; the service parameters include at least the service response time length of the server.
- the basic parameters may also include a version type, and the version type includes at least the disk type of the server, the server version, and the driver version of the connected data source.
- the initial neural network model adopts the DeepFM model to achieve feature crossover, avoid the problem of manual coding and combination features, so as to better adapt to the application scenario of transmission rate matching; adopt the ReLU function as the activation of the DeepFM model Function to create the sparsity of the neural network, provide a mechanism to shield irrelevant factors and then automatically correlate related influencing factors.
- the DeepFM model combines the advantages of the breadth and depth models, and jointly trains the FM model and the DNN model to simultaneously learn low-order feature combinations and high-order feature combinations.
- FM is a factorization machine that uses hidden vectors.
- the latent vector does the inner product to represent the combined features, which can model low-level feature interaction and high-level feature interaction, so feature crossover is realized without the need to avoid the problem of manually coding combined features for feature engineering.
- the ReLU function refers to a linear rectification function (Rectified Linear Unit, ReLU), which is an activation function (activation function) commonly used in artificial neural networks, and usually refers to a non-linear function represented by a ramp function and its variants.
- ReLU Rectified Linear Unit
- activation function activation function
- the use of Relu will cause some neurons to be 0, which causes the sparsity of the network, reduces the interdependence between parameters, and alleviates the occurrence of over-fitting problems.
- the hardware parameters, software parameters, and service parameters of the basic parameters are low-level features for the model; and the version type is high-level features for the model; therefore, the FM model and the DNN model will be At the same time, it learns the hardware parameters, software parameters and business parameters as low-level features, as well as the version type as high-level features, and the FM model uses latent vector as the inner product to express the combined features, which can interact and interact with low-level features.
- the high-level feature interaction is modeled, so feature intersection is realized, and the operating status of the server and its database is evaluated from multiple dimensions such as version type dimension, hardware condition dimension, environmental condition dimension, and business environment dimension to obtain more information. Accurate and reasonable technical effect of the best matching speed.
- S203-2 Enter the training parameters into the initial neural network model to calculate a prediction result, and iterate the initial neural network model based on the prediction result and the training rate.
- the training parameters are entered into the initial neural network model to obtain the prediction result by calculation, the difference between the prediction result and the training rate is calculated through a loss function, and the initial neural network is updated according to the difference through a backpropagation algorithm.
- the parameters in the neural network model so that the initial neural network model calculates the prediction result obtained by the training parameter, and the difference between the training rate and the training rate is within a preset training threshold, so that the initial neural network model is Of iterations.
- S203-3 Perform the iteration on the initial neural network model through each training sample in turn, and form a mature network model by performing the iteration according to the steps described in S203-2.
- S204 Adjust the transmission rate of data transmission between the source database and the target database to the best matching rate.
- the data transmission rate of the two servers is adjusted with the best matching rate as the goal, so that all The data transmission can obtain the fastest transmission rate under the premise of ensuring the stability of the two servers.
- steps S202-S204 are executed periodically according to the preset extraction period, it ensures the cyclic detection and calculation of the basic parameters of the server to obtain the best matching rate, and adjust the relationship between the two servers according to the best matching rate.
- the transmission rate achieves the technical effect of dynamically adjusting the transmission rate between the two servers, and continuously ensures that the data transmission can obtain the technical effect of the fastest transmission rate under the premise of ensuring the stability of the server. Ensure the high efficiency and stability of the data transmission operation.
- the transmission rate between the source database and the target database can be obtained through the monitoring system of the server, so it will not be repeated here.
- the step of adjusting the transmission rate of data transmission between the source database and the target database to the optimal matching rate includes:
- S204-1 Obtain the data transmission rate between the source database and the target database and set it as the current rate.
- the data transmission rate on the communication connection between the source database and the target database can be obtained by accessing the monitoring system of the server where the source database is located or the monitoring system of the server where the target database is located.
- S204-2 Determine whether the current rate is greater than the best matching rate; if so, reduce the current rate to the best matching rate; if not, calculate the difference between the current rate and the best matching rate Adjust the difference value, calculate the adjustment difference value through a preset adjustment time to obtain a unit adjustment amount, and execute S204-3.
- the server Since the current rate is greater than the best matching rate, the server will have the risk of crashing and crashing. Therefore, by directly reducing the current rate to the best matching rate, the server will be quickly and effectively kept away from the risk of crash and ensure its smooth operation .
- the adjustment difference is obtained by subtracting the current rate and the best matching rate, and the adjustment difference is divided by the preset adjustment time to obtain the unit adjustment amount.
- S204-3 Adjust the current rate by the unit adjustment amount until the transmission rate of data transmission between the source database and the target database reaches the best matching rate.
- the server's CPU and memory occupancy will increase suddenly, resulting in unstable server operation.
- the current rate is adjusted within a unit time until the current rate reaches the best matching rate.
- this application uses the TCP protocol to control the transmission rate between the source database and the target database.
- the TCP sender when the TCP sender sends data, it does not directly transmit to the receiver, but first copies the data to the send buffer. Similarly, after the TCP receiver gets the data from the TCP data segment, it puts it into the receiving buffer. This can prevent TCP connections from misusing memory, CPU, and network bandwidth, thereby preventing other connections from using these resources.
- the transmission rate control device 1 is installed in a server 2, and includes:
- the parameter extraction module 11 is used to periodically obtain the basic parameters of the server where the source database is located and the server where the target database is located; wherein the basic parameters represent the hardware conditions, environmental conditions, and operating conditions of the server and its database;
- the rate calculation module 12 is used to predict the basic parameters to obtain the best matching rate through a preset mature network model; wherein, the mature network model is used to predict the allowable rate between the source database and the target database according to the basic parameters.
- Maximum transmission rate which is the best matching rate between the current source database and the target database;
- the rate adjustment module 13 is configured to adjust the transmission rate of data transmission between the source database and the target database to the best matching rate.
- the transmission rate control device further includes:
- connection establishment module 10 is used to establish a connection with the server where the source database is located and the server where the target database is located respectively; wherein, a communication connection is established with the server via TCP or UDP.
- This technical solution is based on the data analysis technology in the big data field, and periodically obtains the basic parameters of the server where the source database is located and the server where the target database is located; and calculates the basic parameters by constructing a mature network model (ie, neural network) as a classification model Obtain the best matching rate to realize the relationship network analysis between the servers; adjust the data transmission rate between the source database and the target database according to the best matching rate to achieve better resource allocation.
- a mature network model ie, neural network
- the present application also provides a computer system that includes a plurality of computer devices 6.
- the components of the transmission rate control device 1 of the second embodiment can be dispersed in different computer devices, and the computer devices can be executed Program smart phones, tablet computers, notebook computers, desktop computers, rack servers, blade servers, tower servers or cabinet servers (including independent servers, or server clusters composed of multiple servers), etc.
- the computer equipment in this embodiment at least includes but is not limited to: a memory 61 and a processor 62 that can be communicatively connected to each other through a system bus, as shown in FIG. 7. It should be pointed out that FIG. 7 only shows a computer device with components, but it should be understood that it is not required to implement all the illustrated components, and more or fewer components may be implemented instead.
- the memory 61 (ie, readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), Read only memory (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory (PROM), magnetic memory, magnetic disk, optical disk, etc.
- the memory 61 may be an internal storage unit of a computer device, for example, a hard disk or a memory of the computer device.
- the memory 61 may also be an external storage device of the computer device, for example, a plug-in hard disk equipped on the computer device, a smart memory card (Smart Media Card, SMC), and a Secure Digital (SD) Card, Flash Card, etc.
- the memory 61 may also include both the internal storage unit of the computer device and its external storage device.
- the memory 61 is generally used to store an operating system and various application software installed in a computer device, such as the program code of the transmission rate control device in the first embodiment, and so on.
- the memory 61 may also be used to temporarily store various types of data that have been output or will be output.
- the processor 62 may be a central processing unit (Central Processing Unit, CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments.
- the processor 62 is generally used to control the overall operation of the computer equipment.
- the processor 62 is used to run program codes or process data stored in the memory 61, for example, to run a transmission rate control device, so as to implement the transmission rate control method of the first embodiment.
- this application also provides a computer-readable storage system, which includes multiple storage media.
- the storage media may be non-volatile or volatile, such as flash memory, hard disk, multimedia card, card-type memory ( For example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read only memory (ROM), electrically erasable programmable read only memory (EEPROM), programmable read only memory ( PROM), magnetic storage, magnetic disks, optical disks, servers, App application malls, etc., have computer programs stored thereon, and corresponding functions are realized when the programs are executed by the processor 62.
- the computer-readable storage medium of this embodiment is used to store the transmission rate control device, and when executed by the processor 62, the transmission rate control method of the first embodiment is implemented.
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Computer And Data Communications (AREA)
Abstract
本申请公开了传输速率控制方法、装置、计算机***及可读存储介质,基于大数据,包括:周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;通过预设的成熟网络模型计算所述基本参数获得最佳匹配速率;将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。本申请避免了当前因通过阈值调节导致传输速率过大导致服务器崩溃,或传输速率过小导致服务器传输效率降低的问题,以最佳匹配速率为目标对服务器进行数据传输速率的动态调整,持续的保证了所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率的技术效果,持续保证了数据传输操作的高效稳定性。
Description
本申请要求于2020年1月19日提交中国专利局、申请号为202010057739.1,申请名称为“传输速率控制方法、装置、计算机***及可读存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
本申请涉及人工智能技术领域,尤其涉及一种传输速率控制方法、装置、计算机***及可读存储介质。
在现今数据***时代,企业内部拥有众多***沉淀了大量的数据,但内部***之间缺乏统一的平台对数据进行关联、整合及联通,导致***之间无法协同工作,难以充分释放数据的真正价值,需要一个统一的数据集成平台处理这些孤岛数据。同时市面上有大量类型的数据库,关系型、非关系型、大数据等,产品种类也是多种多样,现有集成工具支持的数据库类型。集成工具在同步数据速率上普遍采用较固定的方式配置,但是实际应用中设备和网络环境比较复杂,数据库的主要性能需要满足业务***的频繁调用。
那么对于如何实现最大化利用作为源数据源的源数据库,和作为目标数据源的目标数据库所在服务器的剩余性能,充分利用网络线路剩余带宽,当前采用的机制是通过设定固定阈值,以规定源数据库和目标数据库之间数据传输的传输速率。
发明人意识到,这种方法只是通过限定某一安全的速率区间,以避免服务器因数据传输占用过大资源,而导致服务器无法正常运行的情况;但是没有基于服务器的运行情况对传输速率进行调整,因此很容易造成当前服务器运行状况较差但因阈值较高导致传输速率过大导致服务器崩溃,或当前服务器运行状况较好但因阈值较低而导致传输速率过小导致服务器传输效率降低的问题发生。
申请内容
本申请的目的是提供一种传输速率控制方法、装置、计算机***及可读存储介质,用于解决现有技术存在的当前服务器运行状况较差但因阈值较高导致传输速率过大导致服务器崩溃,或当前服务器运行状况较好但因阈值较低而导致传输速率过小导致服务器传输效率降低的问题。
为实现上述目的,本申请提供一种传输速率控制方法,包括:
周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;
通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络 模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;
将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
为实现上述目的,本申请还提供一种传输速率控制装置,包括:
参数提取模块,用于周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;
速率计算模块,用于通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;
速率调整模块,用于将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
为实现上述目的,本申请还提供一种计算机***,其包括多个计算机设备,各计算机设备包括存储器.处理器以及存储在存储器上并可在处理器上运行的计算机程序,所述多个计算机设备的处理器执行所述计算机程序时共同实现上述传输速率控制方法的步骤。
为实现上述目的,本申请还提供一种计算机可读存储介质,其包括多个存储介质,各存储介质上存储有计算机程序,所述多个存储介质存储的所述计算机程序被处理器执行时共同实现上述传输速率控制方法的步骤。
本申请提供的传输速率控制方法、装置、计算机***及可读存储介质,通过基本参数从多种维度综合评价当前源数据库和目标数据库之间能够承受的最大传输速率,以避免当前因通过阈值调节导致传输速率过大导致服务器崩溃,或传输速率过小导致服务器传输效率降低的问题,因此,以最佳匹配速率为目标对服务器进行数据传输速率的调整,使所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率;同时,周期性地获得最佳匹配速率并根据最佳匹配速率调整服务器的传输速率,实现了动态调整两个服务器之间传输速率的技术效果,持续的保证了所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率的技术效果,持续保证了数据传输操作的高效稳定性。
图1为本申请传输速率控制方法实施例一的流程图;
图2为本申请传输速率控制方法实施例二的传输速率控制方法的环境应用示意图;
图3是本申请本申请传输速率控制方法实施例二提供的一种传输速率控制的具体方法流程图;
图4是本申请本申请传输速率控制方法实施例二提供的所述成熟网络模型通过以下步 骤的具体方法流程图;
图5是本申请本申请传输速率控制方法实施例二提供的根据所述最佳匹配速率调整传输速率的具体方法流程图;
图6为本申请传输速率控制装置实施例三的程序模块示意图;
图7为本申请计算机***实施例四中计算机设备的硬件结构示意图。
为了使本申请的目的、技术方案及优点更加清楚明白,以下结合附图及实施例,对本申请进行进一步详细说明。应当理解,此处所描述的具体实施例仅用以解释本申请,并不用于限定本申请。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
本申请提供的传输速率控制方法、装置、计算机***及可读存储介质,适用于大数据领域,为提供一种基于参数提取模块、速率计算模块、速率调整模块和连接建立模块的传输速率控制方法。本申请通过周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;通过预设的成熟网络模型计算所述基本参数获得最佳匹配速率;根据所述最佳匹配速率调整所述源数据库与目标数据库之间数据传输的传输速率。以避免当前因通过阈值调节导致传输速率过大导致服务器崩溃,或传输速率过小导致服务器传输效率降低的问题,使所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率,以及实现了动态调整服务器之间传输速率的技术效果,持续的保证了所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率的技术效果,持续保证了数据传输操作的高效稳定性。
实施例一:
请参阅图1,本实施例的一种传输速率控制方法,包括:
S102:周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;
S103:通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;
S104:将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
在应用场景下,源数据库将数据传输至目标数据库的过程,是通过在源数据库所在服务器与目标数据库所在服务器建立通信连接实现的数据传输;为能够动态且全面的了解源数据库所在服务器和目标数据库所在服务器的运行状态,按照预设的提取周期,周期性地 获得所述服务器的基本参数。
为从多种维度综合评价当前源数据库和目标数据库之间能够承受的最大传输速率,以避免当前因通过阈值调节导致传输速率过大导致服务器崩溃,或传输速率过小导致服务器传输效率降低的问题,因此,以最佳匹配速率为目标对两个服务器进行数据传输速率的调整,使所述数据传输在保证两个所述服务器稳定性的前提下,能够获得最快的传输速率。
由于周期性地执行步骤S102-S104,因此,保证了对服务器的基本参数的循环检测,计算,获得最佳匹配速率,并根据最佳匹配速率调整两个服务器之间的传输速率,实现了动态调整两个服务器之间传输速率的技术效果,持续的保证了所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率的技术效果,持续保证了数据传输操作的高效稳定性。
在示例性的实施例中,所述成熟网络模型是以ReLU函数作为激活函数的DeepFM模型,因其同时学习低阶特征组合和高阶特征组合,故,能够兼顾学习基本参数中各不同数据类型的数据,以更好的适应传输速率匹配的应用场景;并且,通过ReLU函数作为激活函数,以制造神经网络的稀疏性,减少了基本参数中各数据之间的相互依赖关系,缓解了过拟合问题的发生。
实施例二:
本实施例为上述实施例一的一种具体应用场景,通过本实施例,能够更加清楚、具体地阐述本申请所提供的方法。在实现本实施例提供的方法时,可以通过周期性地获取源数据库所在服务器及目标数据库所在服务器的基本参数,并通过计算获得的最佳匹配速率,调整所述源数据库与目标数据库之间数据传输的传输速率。
下面,以用于获取源数据库及目标数据库所在服务器的基本参数,并调整所述源数据库与目标数据库之间数据传输的传输速率为例,来对本实施例提供的方法进行具体说明。需要说明的是,本实施例只是示例性的,并不限制本申请实施例所保护的范围。
图2示意性示出了根据本申请实施例二的传输速率控制方法的环境应用示意图。
在示例性的实施例中,源数据库所在服务器4和目标数据库所在服务器5之间建立了通信连接,传输速率控制方法所在的服务器2通过网络3连接源数据库所在服务器4及目标数据库所在服务器5,所述服务器2可以通过一个或多个网络3提供服务,网络3可以包括各种网络设备,例如路由器,交换机,多路复用器,集线器,调制解调器,网桥,中继器,防火墙,代理设备和/或等等。网络3可以包括物理链路,例如同轴电缆链路,双绞线电缆链路,光纤链路,它们的组合和/或类似物。网络3可以包括无线链路,例如蜂窝链路,卫星链路,Wi-Fi链路和/或类似物;
所述服务器2、源数据库所在服务器4、以及目标数据库所在服务器5可以由单个或多 个计算机设备(如,服务器)组成。该单个或多个计算设备可以包括虚拟化计算实例。虚拟化计算实例可以包括虚拟机,诸如计算机***的仿真,操作***,服务器等。计算设备可以基于定义用于仿真的特定软件(例如,操作***,专用应用程序,服务器)的虚拟映像和/或其他数据来加载虚拟机。随着对不同类型的处理服务的需求改变,可以在一个或多个计算设备上加载和/或终止不同的虚拟机。可以实现管理程序以管理同一计算设备上的不同虚拟机的使用。
图3是本申请一个实施例提供的一种传输速率控制的具体方法流程图,该方法具体包括步骤S201至S204。
S201:分别与源数据库所在服务器和目标数据库所在服务器建立连接;其中,通过TCP或UDP与所述服务器建立通信连接。
本步骤中,源数据库所在服务器和目标数据库所在服务器之间通过建立通信连接,实现源数据库将其中的数据传输至目标数据库的技术效果;分别与源数据库所在服务器和目标数据库所在服务器的监控***建立连接,以便于获得源数据库所在服务器和目标数据库所在服务器的基本参数,以及源数据库和目标数据库之间数据传输的传输速率。
需要说明的是,TCP(Transmission Control Protocol传输控制协议)是一种面向连接的、可靠的、基于字节流的传输层通信协议,由IETF的RFC 793定义。在简化的计算机网络OSI模型中,它完成第四层传输层所指定的功能,用户数据报协议(UDP)是同一层内另一个重要的传输协议。在因特网协议族(Internet protocol suite)中,TCP层是位于IP层之上,应用层之下的中间层。不同主机的应用层之间经常需要可靠的、像管道一样的连接,但是IP层不提供这样的流机制,而是提供不可靠的包交换。UDP是User Datagram Protocol的简称,中文名是用户数据报协议,是OSI(Open System Interconnection,开放式***互联)参考模型中一种无连接的传输层协议,提供面向事务的简单不可靠信息传送服务,IETF RFC 768是UDP的正式规范。所述监控***是一种用于实时监测服务器及其数据库的硬件条件、环境状况以及运行状况的计算机***。
S202:周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况。
在应用场景下,源数据库将数据传输至目标数据库的过程,是通过在源数据库所在服务器与目标数据库所在服务器建立通信连接实现的数据传输;为能够动态且全面的了解源数据库所在服务器和目标数据库所在服务器的运行状态,按照预设的提取周期,周期性地获得所述服务器的基本参数。
本步骤中,所述基本参数至少包括:硬件参数、软件参数和业务参数,所述硬件参数至少包括所述服务器的CPU、内存、网络RTT、ACK;所述软件参数至少包括所述服务器中 数据库的连接数、SQL响应指标、负载压力;所述业务参数至少包括所述服务器的业务响应时长;其中,硬件参数用于表征服务器自身的硬件条件,软件参数用于表征服务器当前运行所面临的环境状况,所述业务参数用于表征服务器中数据库的运行状况;因此,通过获取硬件参数、软件参数和业务参数从硬件条件、环境状况以及业务环境三个维度获得当前服务器及其数据库的运行状态。
进一步地,所述基本参数还可包括版本类型,所述版本类型至少包括服务器的磁盘类型,服务器版本,连接数据源的驱动版本。以便于从版本类型维度,并结合所述硬件条件、环境状况以及业务环境的维度对服务器及其数据库的运行状态进行评价,以获得更加准确合理的最佳匹配速度。
需要说明的是,所述基本参数可通过服务器中的监控***获得,因此,基本参数的获得方式在此不做赘述。
S203:通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率。
为从多种维度综合评价当前源数据库和目标数据库之间能够承受的最大传输速率,以避免当前因通过阈值调节导致传输速率过大导致服务器崩溃,或传输速率过小导致服务器传输效率降低的问题,根据当前源数据库和目标数据库所在服务器的状态,获得两个服务器之间能够承受的最大传输速率。
其中,所述成熟网络模型是以ReLU函数作为激活函数的DeepFM模型,因其同时学习低阶特征组合和高阶特征组合,故,能够兼顾学习基本参数中各不同数据类型的数据,以更好的适应传输速率匹配的应用场景;并且,通过ReLU函数作为激活函数,以制造神经网络的稀疏性,减少了基本参数中各数据之间的相互依赖关系,缓解了过拟合问题的发生。
具体地,请参阅图4,所述成熟网络模型通过以下步骤训练获得:
S203-1:获取初始神经网络模型和至少一个训练样本,所述训练样本包括训练参数和训练速率。
其中,所述训练参数是指源数据库所在的服务器和目标数据库所在的服务器在历史上的运行状况;所述训练速率是指历史上基于所述训练参数,源数据库所在服务器与目标数据库所在服务器在保证服务器稳定运行前提下,能够允许的最大传输速率。
所述训练参数至少包括:硬件参数、软件参数和业务参数,所述硬件参数至少包括所述服务器的CPU、内存、网络RTT、ACK,所述软件参数至少包括所述服务器中数据库的连接数、SQL响应指标、负载压力、业务响应时长;所述业务参数至少包括所述服务器的业务响应时长。
进一步的,所述基本参数还可包括版本类型,所述版本类型至少包括服务器的磁盘类型,服务器版本,连接数据源的驱动版本。
本步骤中,所述初始神经网络模型采用的是DeepFM模型,以实现特征交叉,避免手工编码组合特征的问题,以便于更好的适应传输速率匹配的应用场景;采用ReLU函数作为DeepFM模型的激活函数,以制造神经网络的稀疏性,提供了屏蔽不相关因素的机制进而自动关联相关影响因素。
需要说明的是,所述DeepFM模型结合了广度和深度模型的优点,联合训练FM模型和DNN模型同时学习低阶特征组合和高阶特征组合,其中,FM是一个因子分解机,其通过隐向量latent vector做内积来表示组合特征,可以对低阶特征交互和高阶特征交互进行建模,因此实现了特征交叉,而不需要避免手工编码组合特征以进行特征工程的问题。
所述ReLU函数是指线性整流函数(Rectified Linear Unit,ReLU),其为一种人工神经网络中常用的激活函数(activation function),通常指代以斜坡函数及其变种为代表的非线性函数,于本实施例中,使用Relu会使部分神经元为0,这样就造成了网络的稀疏性,并且减少了参数之间的相互依赖关系,缓解了过拟合问题的发生。
于本实施例中,所述基本参数的硬件参数、软件参数和业务参数对于模型来说,是属于低阶特征;而版本类型对于模型来说属于高阶特征;因此,FM模型和DNN模型将同时对作为低阶特征的硬件参数、软件参数和业务参数,以及作为高阶特征的版本类型进行学习,并且FM模型通过隐向量latent vector做内积来表示组合特征,可以对低阶特征交互和高阶特征交互进行建模,因此实现了特征交叉,实现从版本类型维度、硬件条件维度、环境状况维度以及业务环境维度等多维度出发,对服务器及其数据库的运行状态进行评价,以获得更加准确合理的最佳匹配速度的技术效果。
S203-2:将所述训练参数录入所述初始神经网络模型计算获得预测结果,并通过所述预测结果和所述训练速率对所述初始神经网络模型进行迭代。
具体地,将训练参数录入所述初始神经网络模型使其运算获得预测结果,通过损失函数计算所述预测结果与训练速率的差值,通过反向传播算法并根据所述差值更新所述初始神经网络模型中的参数,以使所述初始神经网络模型计算所述训练参数所获得的预测结果,与所述训练速率的差值在预设的训练阈值以内,实现对所述初始神经网络模型的迭代。
S203-3:依次通过各训练样本对所述初始神经网络模型,按照S203-2所述的步骤进行所述迭代形成成熟网络模型。
S204:将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
为避免源数据库和目标数据库所在的服务器因传输速率过大而崩溃,或因传输速率小 而降低传输效率,故,以最佳匹配速率为目标对两个服务器进行数据传输速率的调整,使所述数据传输在保证两个所述服务器稳定性的前提下,能够获得最快的传输速率。
由于是按照预设的提取周期周期性地执行步骤S202-S204,因此,保证了对服务器的基本参数的循环检测,计算,获得最佳匹配速率,并根据最佳匹配速率调整两个服务器之间的传输速率,实现了动态调整两个服务器之间传输速率的技术效果,持续的保证了所述数据传输在保证所述服务器稳定性的前提下,能够获得最快的传输速率的技术效果,持续保证了数据传输操作的高效稳定性。
需要说明的是,所述源数据库与目标数据库之间的传输速率可通过服务器的监控***获得,故在此不做赘述。
在一个优选的实施例中,请参阅图5,将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率的步骤,包括:
S204-1:获取所述源数据库与目标数据库之间数据传输的传输速率并将其设为当前速率。
本步骤中,可通过访问所述源数据库所在服务器的监控***,或目标数据库所在服务器的监控***,获得源数据库与目标数据库之间的通信连接上数据的传输速率。
S204-2:判断所述当前速率是否大于所述最佳匹配速率;若是,则将所述当前速率降至所述最佳匹配速率;若否,则计算所述当前速率和最佳匹配速率的调整差值,通过预设的调整时间计算所述调整差值获得单位调整量并执行所述S204-3。
由于当前速率若大于最佳匹配速率对于服务器来说是具有崩溃死机等风险的,因此,通过将当前速率直接降至最佳匹配速率,将会快速有效的使服务器远离崩溃风险,保证其平稳运行。
本步骤中,通过将当前速率和最佳匹配速率相减获得调整差值,将所述调整差值与预设的调整时间相除获得单位调整量。
S204-3:通过所述单位调整量调整所述当前速率,直至所述源数据库与目标数据库之间数据传输的传输速率到达最佳匹配速率为止。
为避免对当前速率突然进行大幅度调整,造成服务器的CPU和内存占有量骤然提升,导致服务器运行不稳定。本步骤按照所述单位调整量,在单位时间内调整所述当前速率,直至所述当前速率达到最佳匹配速率为止。
于本实施例中,本申请采用TCP协议控制源数据库与目标数据库中间的传输速率。
需要说明的是,TCP发送端发送数据时,并不是直接传输给接收端,而是先将数据复制到发送缓冲区。类似的,TCP接受端从TCP数据段得到数据后,将其放入接受缓冲区。这样可以避免TCP连接滥用内存、CPU、网络带宽,从而阻止其他连接使用这些资源。
实施例三:
请参阅图6,本实施例的一种传输速率控制装置1,所述传输速率控制装置1安装在服务器2中,其包括:
参数提取模块11,用于周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;
速率计算模块12,用于通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;
速率调整模块13,用于将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
可选的,所述传输速率控制装置还包括:
连接建立模块10,用于分别与源数据库所在服务器和目标数据库所在服务器建立连接;其中,通过TCP或UDP与所述服务器建立通信连接。
本技术方案基于大数据领域的数据分析技术,周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;通过构建成熟网络模型(即:神经网络)作为分类模型,计算所述基本参数获得最佳匹配速率,以实现服务器之间的关系网络分析;根据所述最佳匹配速率调整所述源数据库与目标数据库之间数据传输的传输速率,实现更好的资源分配。
实施例四:
为实现上述目的,本申请还提供一种计算机***,该计算机***包括多个计算机设备6,实施例二的传输速率控制装置1的组成部分可分散于不同的计算机设备中,计算机设备可以是执行程序的智能手机、平板电脑、笔记本电脑、台式计算机、机架式服务器、刀片式服务器、塔式服务器或机柜式服务器(包括独立的服务器,或者多个服务器所组成的服务器集群)等。本实施例的计算机设备至少包括但不限于:可通过***总线相互通信连接的存储器61、处理器62,如图7所示。需要指出的是,图7仅示出了具有组件-的计算机设备,但是应理解的是,并不要求实施所有示出的组件,可以替代的实施更多或者更少的组件。
本实施例中,存储器61(即可读存储介质)包括闪存、硬盘、多媒体卡、卡型存储器(例如,SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘等。在一些实施例中,存储器61可以是计算机设备的内部存储单元, 例如该计算机设备的硬盘或内存。在另一些实施例中,存储器61也可以是计算机设备的外部存储设备,例如该计算机设备上配备的插接式硬盘,智能存储卡(Smart Media Card,SMC),安全数字(Secure Digital,SD)卡,闪存卡(Flash Card)等。当然,存储器61还可以既包括计算机设备的内部存储单元也包括其外部存储设备。本实施例中,存储器61通常用于存储安装于计算机设备的操作***和各类应用软件,例如实施例一的传输速率控制装置的程序代码等。此外,存储器61还可以用于暂时地存储已经输出或者将要输出的各类数据。
处理器62在一些实施例中可以是中央处理器(Central Processing Unit,CPU)、控制器、微控制器、微处理器、或其他数据处理芯片。该处理器62通常用于控制计算机设备的总体操作。本实施例中,处理器62用于运行存储器61中存储的程序代码或者处理数据,例如运行传输速率控制装置,以实现实施例一的传输速率控制方法。
实施例五:
为实现上述目的,本申请还提供一种计算机可读存储***,其包括多个存储介质,存储介质可以是非易失性,也可以是易失,如闪存、硬盘、多媒体卡、卡型存储器(例如,SD或DX存储器等)、随机访问存储器(RAM)、静态随机访问存储器(SRAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、可编程只读存储器(PROM)、磁性存储器、磁盘、光盘、服务器、App应用商城等等,其上存储有计算机程序,程序被处理器62执行时实现相应功能。本实施例的计算机可读存储介质用于存储传输速率控制装置,被处理器62执行时实现实施例一的传输速率控制方法。
上述本申请实施例序号仅仅为了描述,不代表实施例的优劣。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。
以上仅为本申请的优选实施例,并非因此限制本申请的专利范围,凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本申请的专利保护范围内。
Claims (20)
- 一种传输速率控制方法,包括:周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
- 根据权利要求1所述的传输速率控制方法,其中,周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数之前,包括:分别与源数据库所在服务器和目标数据库所在服务器建立连接;其中,通过TCP或UDP与所述服务器建立通信连接。
- 根据权利要求1所述的传输速率控制方法,其中,所述基本参数至少包括:硬件参数、软件参数和业务参数,其中,硬件参数用于表征服务器的硬件条件,软件参数用于表征服务器当前运行所面临的环境状况,业务参数用于表征服务器中数据库的运行状况。
- 根据权利要求1所述的传输速率控制方法,其中,所述成熟网络模型是以ReLU函数作为激活函数的DeepFM模型。
- 根据权利要求1所述的传输速率控制方法,其中,所述成熟网络模型通过以下步骤训练获得:获取初始神经网络模型和至少一个训练样本,所述训练样本包括训练参数和训练速率;将所述训练参数录入所述初始神经网络模型计算获得预测结果,并通过所述预测结果和所述训练速率对所述初始神经网络模型进行迭代;依次通过各训练样本对所述初始神经网络模型进行所述迭代形成成熟网络模型。
- 根据权利要求5所述的传输速率控制方法,其中,所述将所述训练参数录入所述初始神经网络模型计算获得预测结果,并通过所述预测结果和所述训练速率对所述初始神经网络模型进行迭代的步骤,包括:将训练参数录入所述初始神经网络模型使其运算获得预测结果,通过损失函数计算所述预测结果与训练速率的差值,通过反向传播算法并根据所述差值更新所述初始神经网络模型中的参数,以使所述初始神经网络模型计算所述训练参数所获得的预测 结果,与所述训练速率的差值在预设的训练阈值以内,实现对所述初始神经网络模型的迭代。
- 根据权利要求1所述的传输速率控制方法,其中,将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率的步骤,包括:获取所述源数据库与目标数据库之间数据传输的传输速率并将其设为当前速率;判断所述当前速率是否大于所述最佳匹配速率;若是,则将所述当前速率降至所述最佳匹配速率;若否,则计算所述当前速率和最佳匹配速率的调整差值,通过预设的调整时间计算所述调整差值获得单位调整量;通过所述单位调整量调整所述当前速率,直至所述源数据库与目标数据库之间数据传输的传输速率到达最佳匹配速率为止。
- 一种传输速率控制装置,包括:参数提取模块,用于周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;速率计算模块,用于通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;速率调整模块,用于将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
- 一种计算机***,其包括多个计算机设备,各计算机设备包括存储器,处理器以及存储在存储器上并可在处理器上运行的计算机程序,所述多个计算机设备的处理器执行所述计算机程序时共同实现如下步骤:周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
- 根据权利要求9所述的计算机***,其中,周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数之前,包括:分别与源数据库所在服务器和目标数据库所在服务器建立连接;其中,通过TCP或UDP与所述服务器建立通信连接。
- 根据权利要求9所述的计算机***,其中,所述基本参数至少包括:硬件参数、软件参数和业务参数,其中,硬件参数用于表征服务器的硬件条件,软件参数用于表征服务器当前运行所面临的环境状况,业务参数用于表征服务器中数据库的运行状况。
- 根据权利要求9所述的计算机***,其中,所述成熟网络模型是以ReLU函数作为激活函数的DeepFM模型。
- 根据权利要求9所述的计算机***,其中,所述成熟网络模型通过以下步骤训练获得:获取初始神经网络模型和至少一个训练样本,所述训练样本包括训练参数和训练速率;将所述训练参数录入所述初始神经网络模型计算获得预测结果,并通过所述预测结果和所述训练速率对所述初始神经网络模型进行迭代;依次通过各训练样本对所述初始神经网络模型进行所述迭代形成成熟网络模型。
- 一种计算机可读存储介质,其包括多个存储介质,各存储介质上存储有计算机程序,所述多个存储介质存储的所述计算机程序被处理器执行时共同实现如下步骤:周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数;其中,所述基本参数表征了服务器及其数据库的硬件条件、环境状况以及运行状况;通过预设的成熟网络模型预测所述基本参数获得最佳匹配速率;其中,所述成熟网络模型用于根据所述基本参数,预测源数据库和目标数据库之间允许的最大传输速率,该传输速率为当前源数据库和目标数据库之间的最佳匹配速率;将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率。
- 根据权利要求14所述的计算机可读存储介质,其中,周期性地分别获取源数据库所在服务器及目标数据库所在服务器的基本参数之前,包括:分别与源数据库所在服务器和目标数据库所在服务器建立连接;其中,通过TCP或UDP与所述服务器建立通信连接。
- 根据权利要求14所述的计算机可读存储介质,其中,所述基本参数至少包括:硬件参数、软件参数和业务参数,其中,硬件参数用于表征服务器的硬件条件,软件参数用于表征服务器当前运行所面临的环境状况,业务参数用于表征服务器中数据库的运行状况。
- 根据权利要求14所述的计算机可读存储介质,其中,所述成熟网络模型是以 ReLU函数作为激活函数的DeepFM模型。
- 根据权利要求14所述的计算机可读存储介质,其中,所述成熟网络模型通过以下步骤训练获得:获取初始神经网络模型和至少一个训练样本,所述训练样本包括训练参数和训练速率;将所述训练参数录入所述初始神经网络模型计算获得预测结果,并通过所述预测结果和所述训练速率对所述初始神经网络模型进行迭代;依次通过各训练样本对所述初始神经网络模型进行所述迭代形成成熟网络模型。
- 根据权利要求18所述的计算机可读存储介质,其中,所述将所述训练参数录入所述初始神经网络模型计算获得预测结果,并通过所述预测结果和所述训练速率对所述初始神经网络模型进行迭代,包括:将训练参数录入所述初始神经网络模型使其运算获得预测结果,通过损失函数计算所述预测结果与训练速率的差值,通过反向传播算法并根据所述差值更新所述初始神经网络模型中的参数,以使所述初始神经网络模型计算所述训练参数所获得的预测结果,与所述训练速率的差值在预设的训练阈值以内,实现对所述初始神经网络模型的迭代。
- 根据权利要求14所述的计算机可读存储介质,其中,将所述源数据库与目标数据库之间数据传输的传输速率调整为所述最佳匹配速率,包括:获取所述源数据库与目标数据库之间数据传输的传输速率并将其设为当前速率;判断所述当前速率是否大于所述最佳匹配速率;若是,则将所述当前速率降至所述最佳匹配速率;若否,则计算所述当前速率和最佳匹配速率的调整差值,通过预设的调整时间计算所述调整差值获得单位调整量;通过所述单位调整量调整所述当前速率,直至所述源数据库与目标数据库之间数据传输的传输速率到达最佳匹配速率为止。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010057739.1 | 2020-01-19 | ||
CN202010057739.1A CN111277511B (zh) | 2020-01-19 | 2020-01-19 | 传输速率控制方法、装置、计算机***及可读存储介质 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2021142971A1 true WO2021142971A1 (zh) | 2021-07-22 |
Family
ID=71000794
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2020/087614 WO2021142971A1 (zh) | 2020-01-19 | 2020-04-28 | 传输速率控制方法、装置、计算机***及可读存储介质 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN111277511B (zh) |
WO (1) | WO2021142971A1 (zh) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113704071A (zh) * | 2021-07-30 | 2021-11-26 | 济南浪潮数据技术有限公司 | 一种服务器健康状况预测的方法、装置、设备及可读介质 |
CN113766564A (zh) * | 2021-10-15 | 2021-12-07 | 武汉联影医疗科技有限公司 | 数据传输的拥塞控制方法、装置、计算机设备和存储介质 |
CN114286356A (zh) * | 2021-12-29 | 2022-04-05 | 河北电信设计咨询有限公司 | 基于数据库构建5g业务模型的***及方法 |
CN114844894A (zh) * | 2022-04-29 | 2022-08-02 | 济南浪潮数据技术有限公司 | 一种服务器节点的数据传输方法、装置以及介质 |
CN115665059A (zh) * | 2022-11-17 | 2023-01-31 | 北京润泽云溪科技有限公司 | 一种数据传输分级储存*** |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112214487A (zh) * | 2020-09-28 | 2021-01-12 | 京东数字科技控股股份有限公司 | 数据写入方法及装置、计算机可读存储介质以及电子设备 |
CN112272074B (zh) * | 2020-10-27 | 2022-11-25 | 国网内蒙古东部电力有限公司电力科学研究院 | 一种基于神经网络的信息传输速率控制方法及*** |
CN117200935A (zh) * | 2022-05-31 | 2023-12-08 | 华为技术有限公司 | 一种信号处理方法及装置 |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101902295A (zh) * | 2009-05-26 | 2010-12-01 | 国际商业机器公司 | 控制应用服务器的负载接收速率的方法及装置 |
CN107181688A (zh) * | 2017-03-31 | 2017-09-19 | 武汉绿色网络信息服务有限责任公司 | 一种sdn网络中实现服务器端跨域数据传输优化的***和方法 |
CN107948004A (zh) * | 2017-12-29 | 2018-04-20 | 北京奇艺世纪科技有限公司 | 一种视频cdn调取优化方法及装置 |
US20180217825A1 (en) * | 2017-01-30 | 2018-08-02 | Dell Products L.P. | Parallel diagnostic/software installation system |
US20180225235A1 (en) * | 2017-02-03 | 2018-08-09 | Futurewei Technologies, Inc. | Systems and methods for utilizing ddr4-dram chips in hybrid ddr5-dimms and for cascading ddr5-dimms |
CN108540506A (zh) * | 2017-03-01 | 2018-09-14 | ***通信有限公司研究院 | 一种信息传输方法、第一服务器以及终端设备 |
CN109417514A (zh) * | 2018-03-06 | 2019-03-01 | 华为技术有限公司 | 一种报文发送的方法、装置和存储设备 |
CN110022378A (zh) * | 2019-04-22 | 2019-07-16 | 苏州排忧网络技术有限责任公司 | 基于状态的vpn服务器智能分配方法 |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109474444A (zh) * | 2017-09-07 | 2019-03-15 | ***通信有限公司研究院 | 一种传输控制协议tcp窗口的调整方法及调整设备 |
US10616121B2 (en) * | 2017-11-16 | 2020-04-07 | Distech Controls Inc. | Inference server and environment control device for inferring an optimal wireless data transfer rate |
US11510097B2 (en) * | 2017-11-16 | 2022-11-22 | Distech Controls Inc. | Environment control device and method for inferring an optimal wireless data transfer rate using a neural network |
CN110475295B (zh) * | 2018-05-10 | 2023-12-01 | 中国信息通信研究院 | 一种超密集组网的信息传输速率控制方法及装置 |
CN108848566B (zh) * | 2018-06-12 | 2021-02-02 | Oppo广东移动通信有限公司 | 无线通信传输方法、装置、移动终端及计算机可读取存储介质 |
-
2020
- 2020-01-19 CN CN202010057739.1A patent/CN111277511B/zh active Active
- 2020-04-28 WO PCT/CN2020/087614 patent/WO2021142971A1/zh active Application Filing
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101902295A (zh) * | 2009-05-26 | 2010-12-01 | 国际商业机器公司 | 控制应用服务器的负载接收速率的方法及装置 |
US20180217825A1 (en) * | 2017-01-30 | 2018-08-02 | Dell Products L.P. | Parallel diagnostic/software installation system |
US20180225235A1 (en) * | 2017-02-03 | 2018-08-09 | Futurewei Technologies, Inc. | Systems and methods for utilizing ddr4-dram chips in hybrid ddr5-dimms and for cascading ddr5-dimms |
CN108540506A (zh) * | 2017-03-01 | 2018-09-14 | ***通信有限公司研究院 | 一种信息传输方法、第一服务器以及终端设备 |
CN107181688A (zh) * | 2017-03-31 | 2017-09-19 | 武汉绿色网络信息服务有限责任公司 | 一种sdn网络中实现服务器端跨域数据传输优化的***和方法 |
CN107948004A (zh) * | 2017-12-29 | 2018-04-20 | 北京奇艺世纪科技有限公司 | 一种视频cdn调取优化方法及装置 |
CN109417514A (zh) * | 2018-03-06 | 2019-03-01 | 华为技术有限公司 | 一种报文发送的方法、装置和存储设备 |
CN110022378A (zh) * | 2019-04-22 | 2019-07-16 | 苏州排忧网络技术有限责任公司 | 基于状态的vpn服务器智能分配方法 |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113704071A (zh) * | 2021-07-30 | 2021-11-26 | 济南浪潮数据技术有限公司 | 一种服务器健康状况预测的方法、装置、设备及可读介质 |
CN113766564A (zh) * | 2021-10-15 | 2021-12-07 | 武汉联影医疗科技有限公司 | 数据传输的拥塞控制方法、装置、计算机设备和存储介质 |
CN113766564B (zh) * | 2021-10-15 | 2023-05-30 | 武汉联影医疗科技有限公司 | 数据传输的拥塞控制方法、装置、计算机设备和存储介质 |
CN114286356A (zh) * | 2021-12-29 | 2022-04-05 | 河北电信设计咨询有限公司 | 基于数据库构建5g业务模型的***及方法 |
CN114844894A (zh) * | 2022-04-29 | 2022-08-02 | 济南浪潮数据技术有限公司 | 一种服务器节点的数据传输方法、装置以及介质 |
CN115665059A (zh) * | 2022-11-17 | 2023-01-31 | 北京润泽云溪科技有限公司 | 一种数据传输分级储存*** |
CN115665059B (zh) * | 2022-11-17 | 2023-02-28 | 北京润泽云溪科技有限公司 | 一种数据传输分级储存*** |
Also Published As
Publication number | Publication date |
---|---|
CN111277511A (zh) | 2020-06-12 |
CN111277511B (zh) | 2022-03-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2021142971A1 (zh) | 传输速率控制方法、装置、计算机***及可读存储介质 | |
CN107590001B (zh) | 负载均衡方法及装置、存储介质、电子设备 | |
US20180039897A1 (en) | Forcasting interest in an object over a future period of time using a three-stage time-series analysis process | |
US11206706B2 (en) | Method and apparatus for web browsing on multihomed mobile devices | |
CN103414579A (zh) | 一种适用于云计算的跨平台监控***及其监控方法 | |
CN103986786A (zh) | 一种远程云桌面操作*** | |
US20200112499A1 (en) | Multiple quorum witness | |
CN103428198B (zh) | 网络设备、***和方法 | |
US10387214B1 (en) | Managing data processing in a distributed computing environment | |
US10540238B2 (en) | Adjusting replication operations in a distributed environment | |
US9684538B1 (en) | Enhanced power method on an electronic device | |
JP2022188275A (ja) | 量子チャンネルクラシック容量の推定方法及び装置、電子機器と媒体 | |
WO2023284387A1 (zh) | 基于联邦学习的模型训练方法、装置、***、设备和介质 | |
US9448827B1 (en) | Stub domain for request servicing | |
WO2018130161A1 (zh) | 基于云计算服务的高效传输方法和装置 | |
WO2023040203A1 (zh) | 一种人工智能平台的数据获取方法、装置、设备、介质 | |
CN111712799B (zh) | 供在非边缘装置和边缘装置上执行的模型的自动分布 | |
CN117196014B (zh) | 基于联邦学习的模型训练方法、装置、计算机设备及介质 | |
US20090132582A1 (en) | Processor-server hybrid system for processing data | |
US9021109B1 (en) | Controlling requests through message headers | |
US8682958B2 (en) | Decoupling cluster data from cloud deployment | |
US9806956B1 (en) | Managing computing resources by predicting resource usage | |
US20220244976A1 (en) | Containers on demand | |
US20210303516A1 (en) | Format agnostic database migration | |
CN114756312A (zh) | 用于对本地服务进行远程协助优化的***和方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20913999 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 10/11/2022) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20913999 Country of ref document: EP Kind code of ref document: A1 |