124 results about "Server load" patented technology

The invention discloses a Web end video monitoring distributed decoding method and system, and belongs to the technical field of network video playing. The method comprises the following steps of: actively initiating WebSocket connection by a browser end through a local loopback address port (127.0. 0.1: 19022), and detecting whether a distributed decoding program is installed in a user terminal or not; calling a local FFmpeg library by a distributed decoding program, decoding a video stream in real time by utilizing a terminal CPU (Central Processing Unit), and converting decoded video key frames into an image format originally supported by a browser frame by frame through OpenCV; the distributed decoding program packages binary data of each frame of image and an address identifier of the binary data into a WebSocket message, the WebSocket message is actively pushed to the browser, and the browser refreshes and displays the WebSocket message in a specified area of a page in real time according to the address identifier; according to the method, zero-plug-in, low-delay, cross-platform, safe and controllable Web end video decoding is realized, the server load and the network delay are remarkably reduced, and the user experience and the system compatibility are improved.

The invention relates to the technical field of edge computing resource management, in particular to an edge server load evaluation method based on analytic hierarchy process and performance loss. According to the method, firstly, edge servers are divided into a common type, a computing type, a memory type and an I / O type according to heterogeneous types; aiming at each type of edge server, constructing a judgment matrix by adopting an analytic hierarchy process and solving a feature vector so as to calculate load weight coefficients of CPU (Central Processing Unit), memory and disk I / O (Input / Output) resources; calculating a real-time load based on the weight coefficient, wherein the real-time load is the weighted sum of the resource utilization rates; the performance loss is further calculated according to the nonlinear relation between the real-time load and the total load ratio; and finally, restraining the real-time load through the load threshold value, and evaluating the performance state of the server in combination with the performance loss. According to the method, the accuracy and reliability of load evaluation of the heterogeneous edge server are effectively improved, the resource utilization rate is optimized, overload of the server is avoided, and the system stability is enhanced.

The invention discloses a double-layer collaborative energy management and control method and system for a cold plate type liquid cooling data center, and belongs to the technical field of data center energy management. The method comprises the steps that upper-layer day-ahead economic dispatching is based on time-of-use electricity price and outdoor environment temperature prediction, and a server load scheme and a cooling system reference flow track are generated with the minimum total operation cost as the target; in lower-layer intraday dynamic regulation and control, model prediction control is adopted to track the trajectory, the load rate of a server and the rotating speed of a secondary pump are adjusted in real time, and it is ensured that the temperature of a chip is within the safety range of 45-70 DEG C; and the lower layer feeds back the actual chip temperature and energy consumption to the upper layer for online correction of dynamic thermodynamic model parameters to form closed-loop optimization. The system comprises a data acquisition module, a dynamic thermodynamic modeling module, an upper and lower layer optimization control module and an information interaction and feedback module. According to the invention, the overall operation cost and the PUE value of the data center are reduced on the premise of ensuring the temperature safety of the chip.

The invention belongs to the field of deep learning prediction, and particularly relates to a deep learning prediction method for patient medical experience. A deep learning prediction method for patient medical experience comprises the following steps: acquiring sample data containing patient medical images and doctor-patient dialogue audio, converting the sample data into scale data in combination with a stored scale, and unifying data types of the scale data; screening the scale data through a preset missing rate threshold value; performing data amplification according to the score of each label in the scale data; a prediction model is constructed based on a multi-task deep learning model, and the bottom layer of the prediction model is composed of a linear layer and a convolution auto-encoder. According to the scheme, through multi-modal data processing, a multi-task model and edge calculation, the recognition accuracy is improved, the load of the server is reduced, and the problem that the economic burden of a hospital is increased due to the high-performance requirement of the server is solved; meanwhile, resource consumption is balanced, and data security is guaranteed.

The invention relates to a method for accelerating file acquisition by using RDMA (Remote Direct Memory Access), and belongs to the technical field of AI (Artificial Intelligence) training, reasoning and file transmission in a large-scale CPU (Central Processing Unit) calculation-intensive scene. Aiming at the problems of IO bottleneck, client memory limitation and high special hardware deployment cost existing in traditional file reading, a client-server architecture is adopted, a control plane and a data plane are separated, a server loads a remote shared file through a large memory and registers as an RDMA memory area, and the remote shared file is stored in the RDMA memory area. And establishing a quick index of the file and the memory area by means of Hash mapping. The client directly accesses the memory of the server based on RDMAreadverbs, data does not need to be forwarded and copied by a kernel, oversized files can be read in blocks, and the system is compatible with a common NFS storage system and an ROCE network card. The method effectively breaks through the memory limitation of the client, reduces the deployment and hardware cost, guarantees the data transmission accuracy, is suitable for a multi-client concurrent access scene, and provides a low-delay and high-bandwidth efficient solution for high-frequency data set access.

The invention discloses a precise liquid feeding structure control method applied to a single-phase immersion liquid cooling data center box body, which comprises the following steps: step 1, data acquisition and modeling: acquiring parameters by using a sensor, and simulating a cooling liquid flow path and temperature field distribution through thermal analysis; 2, simulation analysis and problem positioning are carried out, the flow velocity, flow distribution and temperature field of cooling liquid under different loads are simulated, and an unsmooth flow area, a hot spot area and a high-energy-consumption node are positioned; 3, hardware deployment and debugging: adjusting the installation position of a mainboard water pump to optimize the initial axial direction of a cooling liquid, simulating different server loads, and comparing test data with a simulation result to perform fine tuning optimization; 4, dynamically controlling the flow, executing multi-mode control logic based on the temperature parameters, and dynamically adjusting the rotating speed of a main board water pump and the rotating speed of a circulating pump; cooling liquid is forced to flow through the mainboard water pump, flow distribution is optimized in combination with the heat-flow coupling model, and forced flowing of the cooling liquid is achieved.

The invention relates to the technical field of mobile edge computing, and discloses a method for service migration, and the method comprises the steps: predicting the traffic flow of a set road section in a preset time period; planning a moving path of at least one target user terminal according to the traffic flow of the set road section; at least one target user terminal belongs to the same interest alliance; determining a target server of a to-be-migrated service corresponding to the interest alliance according to the moving path of each target user terminal; and migrating the to-be-migrated service to the corresponding target server. According to the method, the moving path of the user terminal is planned through the predicted traffic flow in the future time period, a congested road section can be avoided for the user terminal, and the to-be-migrated service is shunted in advance, so that the probability that the load of the target server is too large can be reduced, the probability of service interruption is reduced, and the user experience is improved. The invention further discloses a service migration device, electronic equipment and a storage medium.

The invention discloses a data center dynamic energy-saving scheduling method and system based on AI prediction, electronic equipment and a storage medium, and the method comprises the steps: collecting load data of a data center server and external service flow data in real time to construct an original data set, and processing the original data set through employing a long short-term memory network to obtain a load change trend; according to the load change trend, determining distribution characteristics of the thermal load trend in a future time window, extracting idle computing resource data from the server group based on the distribution characteristics, and computing a computing resource distribution proportion needing to be adjusted; carrying out iterative optimization on the calculation resource distribution proportion by adopting a genetic algorithm to obtain an optimized server load distribution scheme; through a server load distribution scheme, obtaining associated refrigeration system parameter data, determining an adjustment vector of a refrigeration strategy, fusing real-time energy consumption data of a power supply module, and judging a correction demand of power supply scheduling; and obtaining the operation configuration of the overall energy efficiency optimization according to the correction demand.

The invention provides a game server state monitoring method and system based on cloud computing, and the method comprises the steps: obtaining a player interaction event record and a server resource consumption index of a game server cluster in a preset monitoring period, constructing a multi-dimensional data association structure, carrying out the behavior mode analysis of the player interaction event record, and carrying out the analysis of the behavior mode. Generating a virtual economic flow map and scene interaction thermodynamic distribution, and performing feature extraction on server resource consumption indexes in the virtual economic flow map and the scene interaction thermodynamic distribution to obtain fragmentation features of a memory access sequence and response delay features of a network session; inputting the virtual economic flow atlas, the scene interaction thermodynamic distribution and fragmentation features and the response delay features into a pre-constructed cross-dimension correlation model to perform server load state evaluation, generating a load state evaluation result, generating a resource scheduling instruction according to resource bottleneck positioning information in the load state evaluation result, and performing resource scheduling on the server according to the resource scheduling instruction. And adjusting a computing resource allocation strategy of the game server cluster. According to the method and the device, the accuracy and the effectiveness of server resource scheduling are effectively improved.

This disclosure discloses a terminal device control method, apparatus, electronic device, and computer-readable medium. One specific implementation of the method includes: generating a set of associated user groups based on an acquired set of user attribute information; performing time-series information fusion processing on the associated user group set based on acquired user virtual item circulation information to obtain a heterogeneous graph of virtual item circulation among associated nodes; extracting node features from the heterogeneous graph of virtual item circulation among associated nodes using an associated node feature extraction model to obtain a set of associated node feature vectors; reconstructing nodes in the associated node feature vector set to obtain a reconstructed feature vector set of associated nodes; generating an abnormal associated user group set based on the reconstructed feature vector set and the associated node feature vector set; and performing circulation interception control on the user terminal device set corresponding to the abnormal associated user group set. This implementation can reduce server load and minimize system resource waste.

The invention discloses an integrated air cooling and liquid cooling data center heat dissipation device and a heat dissipation method thereof, and belongs to the technical field of data center heat dissipation, and the integrated air cooling and liquid cooling data center heat dissipation device comprises an air cooling unit which comprises a fan and cools a data center; the liquid cooling unit comprises a heat exchanger module, a pump set module and a liquid supplementing module; the pump set module and the liquid supplementing module are connected with the heat exchanger module. And the control unit is connected with the air cooling unit and the liquid cooling unit and controls the air cooling unit to cool the data center and the liquid cooling unit to perform liquid cooling heat dissipation on the data center. Through cooperative work of air cooling and liquid cooling heat dissipation, the fan output flow and the pump output flow can be dynamically adjusted according to the load change of the server, the heat dissipation efficiency is optimized, and the overall energy efficiency control of the data center is improved.

A method of noise reduction, including: collecting configuration data with agents from configuration items, transmitting the collected configuration data to a server for processing, blocking receipt of selected configuration data by a server load protection module to protect the server from overload, storing the configuration data at the server in a storage, processing the configuration data with a batch noise detection module that identifies noise in the configuration data; the batch noise detection module deduces rules based on the identified noise and stores the rules in a data knowledge repository, receiving a user query from a user using a front-end module, analyzing the configuration data with an analytical module to transform the configuration data to be aligned with the user query, responsive to the analysis applying a runtime noise detection module to identify and suppress transformed configuration data to remove irrelevant transformed configuration data.

Systems, methods, and computer readable medium are disclosed for performance of network connection operations. Performance of network connection operations includes accessing a cloud service via a client device, the cloud service containing real-time network information including at least two of a real-time server load for a plurality of servers, real-time server capacity, a customer preference, or geolocation IP; receiving from the cloud service, identities of a plurality of candidate servers from the servers, wherein, based on the real time information, the candidate servers meet criteria for potential establishment of a connection with the client device; transmitting probe packets from the client device to at least some of the candidate servers; receiving responses to the probe packets; determining from the responses, communication characteristics for at least some of the probe packets; selecting a server based on the communication characteristics; and establishing a connection between the client device and the selected server.

The application relates to the technical field of artificial intelligence and password security, and discloses a large language model privacy reasoning method and system capable of verifying no-log behavior and output integrity; the method first sends an encrypted question to the server by the client; the server loads a pre-compiled no-log reasoning program in a trusted execution environment, decrypts and reasons; the reasoning is completed, and zero is cleared; a joint hash value is obtained, a custom data field of a remote proof report is embedded, and a program integrity measurement value is obtained; finally, the response and the proof report are returned to the user. The user client compares the local joint hash value with the custom data in the proof, and verifies the program measurement value, so that it can be confirmed that the log function is not enabled in this reasoning, and the received model response is generated by the specified model for the question, and is not tampered with. The application first realizes double-verified guarantee of privacy and output credibility in large model services, and significantly improves the trust degree of the user for the AI system.

The application relates to the technical field of servers, in particular to a server load balancing method and system based on edge computing, which comprises the following steps: monitoring state data of heterogeneous server nodes based on an edge computing scenario, identifying parameter fluctuation abnormal nodes, analyzing task scheduling and resource change trends, screening and grouping nodes, evaluating node continuous online and communication active performance, classifying nodes according to grading standards, and obtaining an availability distribution structure. The application realizes hierarchical dynamic collection of node availability by real-time mining of the coupling rules of load fluctuation, resource cooperation and bandwidth flow dynamic change among nodes, the resource grouping result is continuously optimized according to the node task bearing capacity and multi-time window communication activity, the node scheduling and distribution fully adapt to the network state and heterogeneous computing power scene, the task flow and node distribution tend to be balanced, the resource utilization and grouping structure of the cluster as a whole are more flexible, and the problems of high-frequency change and resource unevenness under the edge deployment environment can be adapted.

The invention provides a self-adaptive distribution system and method for GPU server load prediction, and the method comprises the steps: obtaining a monitoring data stream of a GPU server in real time, and constructing a monitoring sequence matrix according to the monitoring data stream; performing load mutation extraction on each sensitive mutation index in the monitoring sequence matrix to obtain a load pulse factor of each sensitive mutation index, constructing a multi-scale feature set, and generating a context representation vector sequence according to the multi-scale feature set; carrying out load prediction by using the context representation vector sequence to obtain a GPU load prediction sequence, and mapping the GPU load prediction sequence into a resource demand profile; and generating a resource migration cost matrix according to the current resource state information and the resource demand profile, and generating a resource allocation optimization strategy based on the resource migration cost matrix. According to the technical scheme provided by the invention, the abrupt change mode of the load can be accurately captured and represented, and the allocation decision is optimized by coupling the resource migration cost, so that the overall stability of self-adaptive resource allocation is realized.

The present disclosure provides a data processing method and device and electronic equipment, and relates to the technical field of data processing. The method comprises: receiving a data transmission request, and performing encryption processing on the data to be processed to obtain encrypted data; obtaining service quality parameters of a plurality of second servers, the service quality parameters comprising interface response time, channel encryption strength and server load rate; obtaining weight values of the plurality of second servers based on a trained weight model according to the service quality parameters; selecting a target server according to the weight values of the plurality of second servers; and sending the encrypted data to the target server to enable the target server to process the encrypted data. The first server of the present disclosure increases the security of data transmission by encrypting the data to be processed, and determines the second server capable of improving the data processing speed through the service quality parameters and the weight model, thereby improving the data processing speed under the premise of ensuring the security of data transmission.

Systems and methods for providing a seamless automatic repeat request (ARQ) stream are provided. The system can include a plurality of streaming servers, a load balancer, and an ARQ streaming service.

The application discloses a kind of server power consumption adjustment method, electronic equipment and storage medium, especially in the technical field of server operation, comprising: in the case where determining that the server to be processed satisfies power consumption adjustment condition, based on the reverse mapping relationship of the power consumption load association model of the server to be processed, determine the reverse mapping function corresponding to power consumption load association model;Reverse load analysis is carried out on the target power consumption data of the server to be processed using the reverse mapping function, and the target load threshold value matched with the target power consumption data is obtained;Based on target load threshold value, generate the power consumption adjustment strategy of the server to be processed, to adjust the power consumption of the server to be processed according to power consumption adjustment strategy.It can effectively adapt to the dynamic change of server load, accurately capture the complex association between load and power consumption, realize the flexible and accurate adjustment of server power consumption, and guarantee the stable operation of server.

The application discloses a kind of federal social recommendation method and device based on adaptive diffusion denoising, belong to personalized recommendation technical field in electronic commerce and social platform. For the problem of serious noise interference, server load is too high and slow response in existing social federal recommendation, the application constructs interaction graph and social graph by collecting user interaction and social data, initializes user and project embedding representation, carries out adaptive diffusion denoising process in client local, wherein noise scheduling is dynamically adjusted according to user social structure characteristics, and through federal training framework, client local model updating and server global parameter aggregation are realized, finally, denoised social embedding and user interaction embedding are fused to carry out recommendation score prediction and list generation. The application effectively suppresses social noise interference, improves recommendation accuracy and robustness, reduces server load and communication overhead, and is suitable for personalized recommendation service of electronic commerce and social platform.

The application provides a webpage data acquisition method and device, an electronic device and a storage medium. The method comprises the following steps: in response to an operation instruction of a user on a webpage interface, determining whether response data corresponding to the operation instruction is stored in a local cache; when the local cache has the response data, acquiring and displaying the response data; when the local cache does not have the response data, constructing an AJAX asynchronous request message to acquire the response data from a server and display the response data. In this way, the server load can be reduced, the waiting response time can be avoided, the webpage loading speed can be improved, and the user experience can be improved.

The application discloses a content filtering method and system based on role portraits and a storage medium. The method comprises the following steps: a client establishes a first role portrait library for a user role, and sends the first role portrait library to a server; the server learns each received first role portrait library, and generates a second role portrait library for each role; when the client receives conversation information sent by the user role, it is judged whether the user role matches the first role portrait library; if yes, the first role portrait library is used to filter the content of the conversation information; otherwise, the client sends the conversation information to the server, the server matches a corresponding second role portrait library according to the user role, and uses the second role portrait library to filter the content of the conversation information. The application solves the problems of the prior art, such as rigid filtering rules, incapability of combining with context, and uneven load of the client and the server, and realizes intelligent, accurate and adaptive content filtering.

The invention provides a server resource configuration method which can be applied to the technical field of cloud computing. The method comprises the following steps: acquiring a server load caused by a transaction request, wherein the transaction request comprises a conventional transaction request and an operation transaction request; in response to the condition that the server load meets a first preset condition, expanding the capacity of the standby cluster to a first scale, distributing the conventional transaction request to the common cluster, and distributing the operation transaction request to the standby cluster, the server resource quantity of the common cluster being the first scale, and the standby cluster being preset independent of the common cluster; or, in response to the fact that the server load meets a second preset condition, the standby cluster is expanded to a second scale, and the transaction request is distributed to the common cluster and the standby cluster based on a preset proportion, the second scale is smaller than or equal to the first scale, and the second scale is a positive number larger than the first scale. The invention further provides a server resource configuration device and equipment, a storage medium and a program product.

Described herein are methods and systems to save and resume data inputs from users. A server receives interactions with an electronic application; instructs a second server to display prompts on a webpage; retrieves, from the computing device, responses to the prompts, to generate a profile for the user. When the server received an indication of abandonment representing a lack of interaction with the electronic application, the server accesses a database to identify a temporary profile that matches the generated profile for the user to extract a communication identifier from the temporary profile and transmit a notification using the communication identifier. In response to the server receiving a response to the notification, the server loads, for display on the computing device, a second page having at least one prompt of the electronic application prefilled using data extracted from the generated profile.

Systems and methods are provided for controlling load on a server hosting block storage volumes by dynamically controlling latency for input / output (I / O) operations to the block storage volumes. The system can selectively inject synthetic latency into the I / O operations according to how the server is loaded, thus enabling control of server load. For example, when the server is overloaded, additional synthetic latency can be injected into I / O operations to counteract overloading. This synthetic latency can then be removed when the server is no longer overloaded. Modifications of synthetic latency can be targeted to individual clients, individual types of I / O operations, or both, facilitating targeted performance shaping for servers hosting block storage volumes.

The invention relates to a DNS encrypted transmission and flow dynamic allocation method and system based on TLS, and the method comprises the steps: executing a TLS protocol handshake to generate an encrypted transmission channel through building a TCP connection, and limiting an 853 port to only receive a TLS handshake message; the encrypted DNS query is sent and processed through the encrypted transmission channel, and the transmission efficiency is optimized by using pipeline scheduling and a disordered response mechanism; by monitoring a network state and a server load, an encrypted transmission mode or an unencrypted transmission mode is dynamically switched based on a preset threshold value. By decoupling the DoT module and the recursion function, flexible configuration and dynamic loading are supported, full-process encryption communication between the recursion server and the authoritative server is achieved, and the problem that in the prior art, a safety link is incomplete is solved. A dynamic allocation mechanism is introduced, encrypted and unencrypted paths are intelligently switched according to the network state and the load, and the safety and the availability are balanced.

The invention belongs to the technical field of load prediction, and particularly relates to a computing power center server load prediction method, device and equipment and a medium, and the method comprises the following specific steps: collecting load data of a computing power center server; inputting the load data into a pre-trained server load rate prediction model based on GRU, and outputting and obtaining a server load rate at a future moment; and based on the server load rate at the future moment, calculating through a preset nonlinear regression model to obtain server load power consumption at the future moment. The invention provides a computing power center server load prediction method, which comprises the following steps: firstly, constructing a server load prediction model, solving by adopting a regression model aiming at the problem that the load rate in the model is complex and dynamic change is difficult to estimate, constructing a server load rate prediction model based on GRU, and realizing accurate prediction of the server load rate.

A cloud game service task coordination system based on edge computing comprises a task scheduling module, an edge computing node, a task monitoring module, a self-adaptive load balancing module and a safe and reliable transmission module, and the task scheduling module is used for scheduling tasks of the cloud game service; the edge computing node is used for deploying a task execution environment; the task monitoring module monitors the execution condition of the task and the load condition of the server, and high availability and stability of the service are ensured; the self-adaptive load balancing module is used for realizing a self-adaptive load balancing strategy and dynamically adjusting the weight of task allocation; the safe and reliable transmission module is used for ensuring confidentiality and integrity of task data, and a safe and reliable data transmission protocol is adopted, so that the safety and reliability of service are improved. Through the technical means based on edge computing, the performance and efficiency of the cloud game service are improved, and development and application of the cloud game service are further promoted.

The embodiment of the invention provides a multi-page communication method, electronic equipment, a storage medium and a program product. The method comprises the steps of receiving a processing request of a first task sent by a first page, wherein the processing request is used for requesting to start the first task; sending a starting signal corresponding to the first task to the first page; when a first task processing request sent by a second page is received, the second page is added to a first list corresponding to the first task, and the first list is used for recording pages subscribing to the first task; and receiving an execution result of the first task sent by the first page, and sending the execution result to each second page in the first list. According to the method, only the first page which initiates the request for the first time and the server send the request (namely the first task is executed), other pages (namely the second page) do not need to repeatedly send the request to the server, the problem that the load is too high due to the fact that the server processes a large number of redundant requests at the same time is solved, and the load of the server is reduced.

The front-end page dynamic recommendation method and device, the storage medium and the computer device provided by the application can input the real-time burying point data of a target user into the latest front-end recommendation model after obtaining the real-time burying point data of the target user through a front-end burying point technology. The model generates a recommendation list corresponding to the user behavior by using the real-time burying point data, and dynamically adjusts the display content of the current page according to the recommendation list. In the process, the system processes the user behavior data in real time through the front end, realizes millisecond-level recommendation adjustment, and even in an offline state, the front-end recommendation model can provide basic recommendation functions, thereby responding to the change of user interest in real time, significantly improving the recommendation relevance, and increasing the click rate by 20-30%. Moreover, the system migrates 50-70% of the recommendation calculation to the client, significantly reduces the server load, and reduces the total cost of the recommendation system by 30%. At the same time, the application adjusts the recommendation content and the display mode in real time, and increases the product conversion rate by 15-30%.