A cloud service architecture quality evaluation method includes displaying a cloud architecture diagram, where the cloud architecture diagram represents a cloud service architecture, and where the cloud architecture diagram includes a first cloud graphical element, a second cloud graphical element, and an integration relationship between the first cloud service and the second cloud service in the cloud service architecture. The first cloud graphical element corresponds to a first cloud service, and the second cloud graphical element corresponds to a second cloud service.
G06F 3/0484 - Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range
H04L 67/10 - Protocols in which an application is distributed across nodes in the network
H04L 67/75 - Indicating network or usage conditions on the user display
2.
Model Training Method and Apparatus, and Compute Device
A model training method includes: receiving values of a configuration parameter of a first model that are input by a user, where the configuration parameter of the first model includes a training parameter and/or a model parameter of the first model; performing prediction based on a second model to obtain pieces of training indicator data of the first model that respectively correspond to the values, where the training indicator data of the first model includes training process data of the first model and/or hardware indicator data consumed by a server for training the first model; determining target training indicator data from the pieces of training indicator data; determining a target value that is in the plurality of values; and training the first model based on the target value.
This application discloses a cluster and a communication method, to present various resources in a pooled manner and decouple resource pools, so as to improve resource utilization. The cluster in this application includes: a data processing unit pool and a plurality of compute resource pools of different types. The data processing unit pool includes a plurality of data processing units, and any one of the plurality of compute resource pools includes a plurality of compute resources. The data processing unit pool is connected to the plurality of compute resource pools through a high-speed interconnect bus. The data processing unit pool is configured to manage communication between the plurality of compute resource pools and the exterior of the cluster, communication between the plurality of compute resource pools, and communication inside any one of the compute resource pools.
An application running method includes a first container that is in a first container group loads bytecode in a first JAVA archive to obtain a second JAVA archive of an application, where the second JAVA archive includes first bytecode of a to-be-enhanced class in the application, the first container runs a first enhancement component, and the first bytecode is obtained by modifying, by the first enhancement component, bytecode of the to-be-enhanced class in the first JAVA archive in a process in which the first container loads the bytecode in the first JAVA archive; a second container in the first container group loads bytecode in the second JAVA archive to obtain a first loading result; and the second container runs the application based on the first loading result.
An auto scaling method including: when the tenant needs to complete auto scaling for a scaling group of the tenant, the cloud management platform may receive, through a network interface, an auto scaling formula that is input by the tenant for the scaling group. Then, the cloud management platform may obtain a current load indicator of the scaling group, and substitute the current load indicator of the scaling group into the variable that indicates the current load of the scaling group and that is in the auto scaling formula, to obtain a target capacity of the scaling group. Finally, the cloud management platform may adjust a quantity of cloud instances included in the scaling group, until a capacity of an adjusted scaling group matches the target capacity of the scaling group.
There is provided a method and device for hosting multiple player games on a cloud gaming platform. Each player of a gaming session may communicate with a single game engine instance, which may host one or more players on the same game session. The present disclosure improves resource utilization on the game engine by avoiding duplicating the use of resources such as computation and storage for different players. This is achieved by isolating global entities, which are relevant to every player, and local entities, which are relevant to only one player, from each other. The game server may produce audio and video streams for each player based only on the relevant entities for a given player. Gaming logic may also be isolated by declaring either a local or a global scope for game events, and restricting the scope of ensuing sub-events. This isolation mechanism also facilitates multiplayer game development.
The method includes determining service information corresponding to the target tenant, where the service information indicates information about a service instance of the target service in the managed cluster; creating a gateway corresponding to the target tenant, where the gateway includes a traffic monitoring rule; determining a traffic processing rule corresponding to the target service, where the traffic processing rule is associated with a target monitoring rule in the traffic monitoring rule; and sending the service information, the target monitoring rule, and the traffic processing rule to a traffic proxy in a virtual private cloud of the target tenant, to cause the traffic proxy to perform service governance on the target service. A service governance rule of a tenant is sent to a proxy in a virtual private cloud of the tenant, to implement data isolation, and improve data security.
An iterative image-to-3D generation method (100) utilizing a pre-trained diffusion model. The method (100) comprising generating multiple views of an object using an image-to-multiview method based on a sampled view, selecting one of the multiple views of an object, using the selected view as a prompt to the pre-trained diffusion model and optimizing based on a loss function employing weights based on homoscedastic uncertainty, wherein homoscedastic uncertainty is defined as task-dependent uncertainty that stays constant for any input.
A cloud service providing system is described. The cloud service providing system includes a first host that runs an active virtual instance, where the active virtual instance is used for providing a cloud service for a user. The cloud service providing system also include a second host that runs a first standby virtual instance. The first host stores, in a first shared memory during running of the active virtual instance, memory data needed for running the active virtual instance, and the second host runs, when the first host fails, the first standby virtual instance based on the memory data obtained from the first shared memory, to cause the first standby virtual instance to provide a cloud service for the user. This application improves timeliness and efficiency of hot standby of memory data, provides a real-time redundancy backup capability, and helps ensure performance of services provided based on backed-up memory data.
G06F 11/20 - Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
G06F 9/455 - EmulationInterpretationSoftware simulation, e.g. virtualisation or emulation of application or operating system execution engines
In the method, a data exchange platform obtains status data that is of a first object and that is sent by a first device. The first object is an environmental object in a real scene or an item included in the real scene. The data exchange platform sends the status data of the first object to a second device. The second device obtains the status data of the first object, and adjusts, based on the status data of the first object, a second object that is included in a virtual scene and that corresponds to the first object.
This application discloses a method for providing a container service, a container system, and a related apparatus, relating to the field of container technologies. The method includes: The container system provides base images and a public image for users, creates a container by using the base image, and mounts the public image. The public image does not need to be downloaded, thereby reducing occupation on storage space of a node on which the container is located. In addition, the shared public image exists independently, thereby facilitating management and quick update or upgrade for the public image. Because the public image is mounted, simply updating the public image can quickly update the container created for the users. There is no need to download the public image, and the mount-based approach also makes container creation more efficiently.
A method for training a semantic vector-based information retrieval model in the field of artificial intelligence includes: obtaining a first training data set and a second training data set, the first training data includes a first query text, a first retrieval text corresponding to the first query text and having a positive label, and a second retrieval text corresponding to the first query text and having a negative label; updating a first information retrieval model based on the first training data set, to obtain a second information retrieval model; and updating the second information retrieval model based on the second training data set, to obtain a third information retrieval model.
A check rule processing method includes: obtaining to-be-checked code and/or related information about the to-be-checked code, where the related information about the to-be-checked code includes at least one of the following: a programming language of the to-be-checked code, a development framework of the to-be-checked code, a deployment form of the to-be-checked code, a service scope of the to-be-checked code, a third-party library used in the to-be-checked code, a coding style of the to-be-checked code, or description information of a target project to which the to-be-checked code belongs; determining a target rule set from a static analysis rule library based on the to-be-checked code and/or the related information about the to-be-checked code, where the target rule set includes one or more candidate check rules, and the one or more candidate check rules are used to perform static analysis on the to-be-checked code; and performing control to display the target rule set.
G06F 21/57 - Certifying or maintaining trusted computer platforms, e.g. secure boots or power-downs, version controls, system software checks, secure updates or assessing vulnerabilities
G06F 21/56 - Computer malware detection or handling, e.g. anti-virus arrangements
14.
Object Storage System, Data Processing Method, and Related Device
This disclosure provides an object storage system including a service layer, a persistence layer, and a metadata layer. The service layer receives a batch upload request sent by a client, where the batch upload request indicates a plurality of to-be-uploaded first objects; sends data of the plurality of first objects to the persistence layer; and generates metadata of the plurality of first objects, and sends the metadata of the plurality of first objects to the metadata layer. The persistence layer stores the data of the plurality of first objects. The metadata layer stores the metadata of the plurality of first objects into the persistence layer; and determines that metadata of at least one first object fails to be stored into the persistence layer, and sends first indication information to the service layer, where the first indication information indicates to perform garbage collection on the plurality of first objects.
A method, wherein automatic execution of a series of complex operations including at least one AI operation and at least one data processing operation may be triggered by using a table operation instruction. A user only needs to input a simple table operation instruction, and does not need to input a series of complex operation instructions.
A network model training method includes providing a parameter configuration user interface; obtaining a parameter count and a type of a target network model from the parameter configuration user interface; determining training configuration information of the target network model based on the parameter count and the type of the target network model, where a training process of the target network model is divided into a plurality of training phases, different training phases correspond to different model parameter counts, the model parameter counts of the plurality of training phases increase sequentially with the order of training, and training the target network model based on the configuration information of the plurality of training phases.
A cache replacement method includes a compute node in a distributed system that sends a first read operation request to a memory node. Then, the compute node separately calculates popularities of the plurality of data objects based on the access information of the plurality of data objects using a first cache replacement algorithm in a plurality of cache replacement algorithms. Finally, the compute node sends a cache replacement request to the memory node based on the popularities of the plurality of data objects, to indicate to replace cold data in the plurality of data objects with a data object to be inserted into the memory node. In this way, the compute node executes the cache replacement algorithm based on global access information.
The cloud management system is configured to manage an infrastructure on which a first service of a first tenant and a second service of a second tenant are configured. The cloud management system obtains an authorization request of the first tenant which includes a tenant identifier of the second tenant, an authorization name, and a permission for the second service to invoke the first service. The cloud management system creates the permission based on the authorization request; obtains a permission obtaining request that is inputted by the second tenant and that includes the authorization name and a tenant identifier of the first tenant, configures, based on the permission obtaining request, the second tenant to obtain the permission; and allows, based on the permission when the second tenant configures the second service to invoke the first service, the second service to invoke the first service.
The present disclosure relates to methods and devices for processing multimodal data using a Large Multimedia Model. The method includes an encoding process to receive multimodal data, generate multimodal tokens, and store them in a cache; a prefilling process to combine multimodal and text tokens, generate attention states and an initial output token, and store them in memory; and a decoding process to retrieve stored data, generate successive output tokens in an autoregressive manner, and produce an output response. A communication process facilitates independent operation of these processes by mapping memory blocks across graphics processing units, and thereby enables inter-process data transfer. The encoding, prefilling, and decoding processes operate asynchronously, independently, and in parallel.
Methods and devices for processing multimodal data using a Large Multimedia Model. The method includes an encoding process to receive multimodal data, generate multimodal tokens, and store them in a cache; a prefill process to combine multimodal and text tokens, generate attention states and an initial output token, and store them in memory; and a decoding process to retrieve stored data, generate successive output tokens in an autoregressive manner, and produce an output response. A communication process facilitates independent operation of these processes by mapping memory blocks across graphics processing units, and thereby enables inter-process data transfer. The encoding, prefill, and decoding processes operate asynchronously, independently, and in parallel.
This application provides a mesh model generation method and apparatus, and a device. The method includes: obtaining visual data that is of an object and that is uploaded by a user, where the visual data includes at least one of an image, a point cloud, and a voxel model; performing geometric feature extraction on the visual data of the object, to obtain a first geometric feature of the object; obtaining, from a prior knowledge base, prior knowledge corresponding to an object category to which the object belongs, where the prior knowledge is used to guide establishment of a geometric generation network; establishing a first geometric generation network based on the first geometric feature and the prior knowledge corresponding to the object category to which the object belongs; and generating a first mesh model of the object based on the first geometric generation network.
G06T 19/20 - Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
G06V 10/764 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
G06V 10/82 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
22.
METHOD AND SYSTEM FOR INTERACTIVE SEMI-AUTOMATIC LABELING
Methods and systems for semi-automatic video object segmentation combining a plurality of modules, and in particular a coarse mask generation module (CMG) and a mask refinement module (MR). The CMG module uses a vision foundation model (VFM) to generate coarse masks for each frame. The MR module takes user input as clicks on the coarse masks and refines them using a foreground-background segmentation model (FB). The merged mask from the MR module is then passed back to the CMG module for further processing.
A software module running method includes obtaining a running request for a first software module, where the first software module is associated with a first user, the running request for the first software module is to run the first software module that has been deployed in a first computing instance set, or deploy the first software module in a provisioned first computing instance set for running, or provision a first computing instance set for the first software module and run the first software module by using the first computing instance set, and the first computing instance set includes at least one computing instance; and obtaining a first cloud resource authorization specification corresponding to the first software module.
A permission check method is disclosed. A node receives an operation request, the operation request includes a target identifier. The node determines an operation permission of the operation request based on a file ACL of the target identifier and a file user identifier associated with the operation request, and determines a permission check result based on the operation permission of the operation request and a target operation type included in the operation request.
A model switching method, comprising: receiving a first query request, and extracting a first feature on the basis of the first query request; selecting, from among a plurality of AI models on the basis of the first feature, a first model matching the first feature, each model among the plurality of AI models having one or more matched features; using the first model among the plurality of AI models to perform inference on the first query request so as to obtain a first model generation result, and providing the first model generation result on a user interface; receiving a second query request, and extracting a second feature on the basis of the second query request; selecting, from among the plurality of AI models on the basis of the second feature, a second model matching the second feature; and using the second model to perform inference on the second query request so as to obtain a second model generation result, and providing the second model generation result on the user interface. In this way, adaptive model switching is achieved, the same model is not fixedly invoked, the advantages of the AI models are fully utilized, the capability of solving tasks from simple tasks to complex tasks is improved, and service requirements are satisfied.
Disclosed in the present application are a request processing method and apparatus, and a device cluster, a program product and a storage medium, which are applied to a service device, wherein the service device provides a request processing service by means of a broker cluster, and a plurality of broker nodes in the broker cluster share a physical storage unit. The method comprises: acquiring a first request of a logical storage unit; and processing the first request by means of a first broker node, wherein the first broker node is a broker node meeting a load balancing condition in a broker cluster. The logical storage unit is used for managing, on the basis of a mapping relationship, message data stored in the physical storage unit, wherein the mapping relationship indicates a relationship between at least one of the position and publishing time of the message data in the physical storage unit and the position of the message data in the logical storage unit, and the order of positions of the message data in the logical storage unit is consistent with a publishing order of the message data. The global ordering of the message data is ensured while achieving relatively balanced loads among the broker nodes.
Provided in the present application are a testing method and apparatus for an artificial-intelligence cluster, and a computing device cluster. The method is applied to a management platform, which is used for managing infrastructures, wherein the infrastructures comprise an artificial-intelligence cluster, which comprises a plurality of computing units, the plurality of computing units being used for operating an artificial-intelligence model. The method comprises: acquiring architecture information of an artificial-intelligence model, wherein the architecture information is used for describing the deployment architecture of the artificial-intelligence model in a plurality of computing units; on the basis of the architecture information, determining first fault testing information, which is used for indicating a condition for a target computing unit from among the plurality of computing units to perform fault testing when operating the artificial-intelligence model; and sending the first fault testing information to the target computing unit from among the plurality of computing units. Online fault testing is implemented during the normal operation of an artificial-intelligence model, without consuming additional time, thereby reducing the time for model training and inference, and increasing the resource utilization rate.
Provided in the present application are a data retrieval method, an apparatus and a computing device. The method comprises: on the basis of a question input by a user, determining a plurality of pieces of first data to be retrieved; retrieving from a data storage medium a plurality of candidate source tables related to said plurality of pieces of first data, each candidate source table comprising part of target data among said plurality of pieces of first data; on the basis of coverage of said plurality of pieces of first data by each candidate source table and an association relationship between the plurality of candidate source tables, determining a target source table combination from the plurality of candidate source tables; and, on the basis of the target source table combination, generating a target wide table, and retrieving values of said plurality of pieces of first data from the target wide table. The solution can help users find data-related content that cannot be directly searched from external knowledge bases, thereby improving the accuracy of data retrieval.
Disclosed are a server system based on public cloud technology, an instance creation method and a cloud management platform, belonging to the technical field of cloud services. The server system comprises a first server and a second server, between which connection channels are established. A first service module and a first security module are provided in a first instance of the first server; a second service module and a second security module are provided in a second instance of the second server; the first service module is configured to send an access request for the second service module, the access request being used for requesting the second service module to provide a second tenant service; the first security module is configured to acquire the access request, encrypt the access request, and send the encrypted access request to the connection channel; and the second security module is configured to acquire the encrypted access request from the connection channel, decrypt the encrypted access request, and send the decrypted access request to the second service module. The present application implements end-to-end encryption protection of access requests.
A communication method includes a first dynamic host configuration protocol (DHCP) server that allocates a plurality of internet protocol (IP) addresses to an elastic network interface of a first elastic cloud server (ECS), and is configured based on first configuration information of a tenant.
This application discloses a point cloud data processing method. The method includes: obtaining an original point cloud data set of a scene, where the scene includes a target object; augmenting point cloud data of a target object, to generate an augmented point cloud data set; training a first model by using the augmented point cloud data set, to obtain a trained first model; training a second model by using the at least one original point cloud data set and the trained first model; and obtaining a trained second model. An original point cloud is augmented to obtain a dense point cloud, the dense point cloud and the original point cloud are separately used for training to obtain a model, and inter-model training is performed, allowing a target model to have a capability of extracting high-quality information from sparse point clouds and executing point cloud processing tasks.
A computer-implemented method is disclosed. The method includes: receiving a request to process using a pre-trained machine learning model, the request including an input prompt, a first budget, and a task identifier of a first task; pruning the machine learning model, without storing a separate pruned model in memory, wherein the pruning comprises determining, for each layer of the machine learning model, a layer gate value indicating whether to use the layer for inference based on the first budget and the first task; and utilizing the pruned machine learning model in accordance with the layer gate values to generate an output corresponding to the request.
Provided in the present application is a fault diagnosis method. The method is applied to a fault diagnosis service for a CI/CD pipeline, and comprises: acquiring a category of a fault of the CI/CD pipeline, a first problem description of the fault, and a plurality of SOP nodes of the fault diagnosis service, wherein the SOP nodes are used to obtain a diagnosis result on the basis of a problem description; on the basis of the first problem description and at least one first SOP node among the plurality of SOP nodes that corresponds to the category of the fault, determining at least one first diagnosis result; when the at least one first diagnosis result can be used as a solution to the fault, outputting the first diagnosis result; or, when the first diagnosis result cannot be used as a solution to the fault, on the basis of the at least one first diagnosis result and an RAG agent program and/or an LLM agent program of the fault diagnosis service, obtaining a second diagnosis result, wherein the second diagnosis result is used as a solution to the fault; and outputting the second diagnosis result.
The present application provides a hardware device configuration method and a related device applied to the technical field of cloud computing. In the method, a format conversion module in a virtual machine converts a channel creation instruction in a first format into a channel creation instruction in a second format, and sends the channel creation instruction in the second format to a virtio front-end driver. The first format is a format capable of being processed by a hardware device. The second format is a format capable of being processed by a virtio device, and is an instruction format specified in a virtio protocol. The virtio front-end driver sends the channel creation instruction in the second format to a hardware device driver. The hardware device driver converts the channel creation instruction from the second format to the first format, calls the hardware device on the basis of the converted channel creation instruction, and creates a control channel. By means of conversion between the first format and the second format, the standardized virtio front-end driver is used to drive the hardware device, and there is no need to additionally install a driver for the hardware device in a Guest OS, thereby increasing the speed of configuring the hardware device.
A code generation method, a model training method, and a related product, applied to the fields of computer technology and AI. In the present application, at least one model is deployed in a code generation apparatus or the code generation apparatus has the capability to invoke at least one model. The code generation apparatus can acquire code to be converted that is written in a first programming language, perform semantic mapping on said code to obtain an intermediate representation, and use the at least one model and the intermediate representation to obtain target code in a second programming language. In the present application, at least one model is used to convert code from one programming language into another programming language, and lexical‑level semantic mapping can be achieved during conversion, thereby ensuring semantic consistency between the generated code and the original language, allowing for efficient and accurate programming language conversion of code. The present application also accordingly provides a model training method and other related products.
Methods and systems are disclosed. The method comprises receiving the input text sequence; generating, a first likelihood score for each of a plurality of candidate output tokens using an expert model on the input text sequence, the expert model being a LoRA adapter trained to generate text based on a task; generating a respective second likelihood score for each of the plurality of candidate output tokens using an amateur model on the input text sequence, the amateur model having been previously used to train the expert model; generating a contrastive score for each of the plurality of candidate output tokens by maximizing a positive difference in their respective first likelihood score and second likelihood score; selecting a given candidate output text token as the output token based on the respective contrastive score.
This application discloses a method and an apparatus for hybrid deployment across runtime instances based on a cloud computing technology, and belongs to the field of cloud technologies. In this method, a resource scheduling system manages resources of a plurality of hybrid nodes. Because the hybrid node supports deployment of a plurality of types of runtime instances, the resource scheduling system can deploy different types of runtime instances on the hybrid node. Specifications of different types of runtime instances differ greatly, and fragmented resources of the hybrid node can be allocated to a runtime instance with a small specification, so that the fragmented resources of the hybrid node can be fully utilized.
The computer graphics system 600 includes a controller 602. The controller is configured to obtain a view- dependent cut of a graph of nodes representing a hierarchical, cluster-based micro-polygon Level of Detail, LOD system. The controller is configured to compute a selection error, an aliasing error and rasterization cost metric that considers caching benefits by shared node rasterization, for each node. The controller is configured to receive a plurality of LOD selections from multiple viewers, start with an LOD selection having the finest LOD and continuing with LOD selections of coarser LODs optimizing the graph cut towards higher performance with bounded quality reduction, determine whether the node can be moved to a higher detailed node for remaining nodes, determine a current memory load, and reduce the selection of nodes to be merged or moved to fit inside a predefined memory budget.
Disclosed are a permission verification method based on a cloud management platform, and a cloud management platform, which relate to the technical field of cloud computing. In the method, infrastructure managed by a cloud management platform is used for providing object storage services, the infrastructure stores permission control information, which is used for indicating the correspondence between data identifier information and an access control label, and the data identifier information comprises an identifier of a bucket and a prefix of an identifier of data. Since the granularity of a prefix of an identifier of data is larger than that of the identifier of the data, the number of permission entries in the permission control information is much smaller than the number of pieces of data stored in the infrastructure. On this basis, when the cloud management platform has received an operation request for target data and then traverses the permission control information to determine an access control label corresponding to data identifier information of the target data, the number of permission entries that need to be traversed is relatively small, so that the speed of traversing the permission control information can be significantly increased, thereby significantly improving the performance of permission verification.
Disclosed are an application fault injection method and apparatus, and a computing device cluster, relating to the technical field of computers. Upon receiving a fault injection request of a user, an endogenous fault module of a target application directly injects a fault into a service module of the target application. The fault injection scope is limited to the service module of the target application, and the fault impact scope is limited to a service of the target application. The fault injection scope is controllable, thereby ensuring that the fault impact scope is controllable, and further providing a guarantee for normal processing of the service. In addition, the endogenous fault module of the target application directly injects the fault, and fault injection can be implemented without the need to install a fault injection agent on a node on which the target application is deployed, thereby improving the convenience of use, reducing the requirements for the node on which the target application is deployed, and expanding the use scenarios.
Provided in the present application is a data processing method for improving the efficiency of solving for an eigenvalue and an eigenvector of a matrix. The data processing method comprises: acquiring a target matrix; on the basis of a plurality of characteristic parameters of the target matrix, searching a first database to determine at least one target reference eigenvector, wherein the first database stores a plurality of groups of first reference data, each group of first reference data corresponds to one first reference matrix, each group of first reference data comprises a plurality of characteristic parameters and eigenvectors of the first reference matrix, a plurality of target reference eigenvectors are eigenvectors of a target first reference matrix, and a plurality of characteristic parameters of the target first reference matrix match the plurality of characteristic parameters of the target matrix; and on the basis of the at least one target reference eigenvector, determining an initial reference vector, wherein the initial reference vector is used for solving for eigenvalues and eigenvectors of the target matrix. In addition, further provided in the present application are a corresponding apparatus, a computing device cluster, a computer-readable storage medium and a computer program product.
Disclosed in the present application are a permission detection method based on a cloud management platform, and a cloud management platform, which facilitate the resolution of potential data security risks of cloud resources of tenants. The method in the present application comprises: when a first tenant needs to determine whether a second tenant has an operation permission for cloud resources of the first tenant, the first tenant sending to a cloud management platform a permission detection request for the cloud resources of the first tenant; on the basis of the request, the cloud management platform generating a first operation of the second tenant for accessing the cloud resources of the first tenant; if a first authentication strategy bound to the second tenant indicates that the execution of the first operation is not allowed, the cloud management platform further generating a second operation of the second tenant for replacing the first authentication strategy with a second authentication strategy, wherein the second authentication strategy indicates that the execution of the first operation is allowed; and then, the cloud management platform returning a permission detection result to the first tenant, wherein the result is used for indicating the first operation and the second operation of the second tenant, and the second operation precedes the first operation in execution order.
A method and a system for double-objective learned indexing in a key-value store is disclosed. The key-value store is a log-structured merge-tree based store. A sorted sting table is divided into data blocks, where a sum of values of the keys in each data block is equal to or less than the maximum block size value. A piecewise linear approximation function is generated for the plurality of data blocks, where the maximum interpolation error of the piecewise linear approximation function is equal to or less than the model error value. The maximum block size value and the model error value are provided by the reinforcement learning agent. Following receipt of the lookup query request to find the target key, the system initiates a discovery operation to discover a data block among the data blocks and generates a search range for the discovered data block.
A cloud computing technology-based snapshot management method, comprising: acquiring a target instance identifier input by a user, the target instance identifier being configured for indicating a plurality of target cloud instances in a cloud instance set; acquiring an association attribute input by the user, the association attribute being configured for indicating an association relationship among the plurality of target cloud instances; and creating a first snapshot according to the target instance identifier and the association attribute. The method in the present application involves receiving a user selection of a plurality of cloud instances and an association relationship among the plurality of cloud instances to construct a snapshot for executing a target task. The snapshot created by the method enables rapid creation and restoration of resources in specified cloud instances, and establishment of an association relationship among the cloud instances, thereby ensuring that the resources in the specified cloud instances can execute a target service. The method does not perform resource orchestration for a corresponding cloud instance when a target service is being executed, allowing the resources of the corresponding cloud instance to execute other services after completing execution of the target service.
Provided in the present application is a transaction processing method, which is applied to the processing of different transactions in a distributed storage system. A transaction management apparatus may be used to perform logical clock synchronization on index modules on storage nodes involved in transactions, so as to improve the accuracy of ordering among different transactions, thereby reducing inaccuracies in transaction processing results caused by conflicts among different transactions in the distributed system.
Embodiments of the present application disclose an industrial model generation method. The method comprises: acquiring a design intent inputted by a user, the design intent comprising relevant information for generating a target industrial model; retrieving a first industrial model matching the design intent from an industrial model database, the first industrial model being a reference model; and on the basis of the design intent, performing on the first industrial model: basic shape fitting, independent appearance subcomponent fitting, and surface relief detail fitting, thereby obtaining a second industrial model that serves as the target industrial model. By understanding a multi-modal intent inputted by a user, the method matches from a model database an industrial model satisfying the multi-modal intent as a reference model, and also continuously optimizes a basic model on the basis of deep understanding of the multi-modal intent inputted by the user, enabling quick and accurate generation of an industrial model conforming to the design intent of the user.
G06F 30/12 - Geometric CAD characterised by design entry means specially adapted for CAD, e.g. graphical user interfaces [GUI] specially adapted for CAD
A code management method is provided that includes receiving an input task description, decomposing the task description into a plurality of subtask descriptions, and generating code for a plurality of subtasks based on the plurality of subtask descriptions, where the code for the plurality of subtasks corresponds one-to-one to the plurality of subtask descriptions. A task is decomposed into a plurality of general-purpose tasks or atomic tasks (i.e., tasks that do not support further decomposition) by introducing decomposition of a task description to improve accuracy of code generation in a complex multi-step task, achieve a favorable code generation effect, and meet a service requirement.
A method of operating a serverless query engine is provided. The method involves receiving a user query, generating an initial execution plan, and then optimizing that plan by fusing multiple tasks into a single cloud function. This fusion reduces overhead and improves efficiency by minimizing function invocations and data transfers. Upon receiving an invocation request from a scheduler, the optimized plan is executed using the fused function. This approach enhances the performance of serverless query engines, particularly for complex queries involving multiple operations. The method enables faster query processing, better resource utilization, and improved scalability in serverless data processing environments.
A digital image processing method of performing real-time ray tracing of hierarchical micro-poly geometry is provided. The method includes receiving a plurality of Axis Aligned Bounding Boxes, AABBs, corresponding to a respective plurality of levels of detail corresponding to a plurality of clusters, and adding the plurality of AABBs to a first acceleration structure. Each cluster includes a plurality of triangles. The method includes invoking a ray tracing intersection shader associated with the first acceleration structure, when each of the plurality of AABBs is intersected by a ray. The method includes determining whether a geometric error of a cluster associated with the ray which has intersected a respective AABB, is within a specific range of error for a plurality of predetermined view parameters by the ray tracing intersection shader. The method reports the result to the ray tracing pipeline.
A method of creating a confidential cloud computing platform 502 in a disaggregated resource architecture 514, 800. The method includes (i) verifying a fabric manager 508, 708A-C which runs inside of a Trusted Execution Environment, TEE module 506, 822, (ii) submitting a security policy from a user at an attestation and key management system 510, (iii) performing a secure boot at a device to generate a report including details of the device's hardware and software components, (iv) signing the report at the device, using the device's hardware root of trust, HWRoT, of the device's TEE module, to create a quote, (v) sending the signed report to the fabric manager by the device, which generates a trustworthy certificate, and (vi) receiving the trustworthy certificate from the fabric manager by the device, that allows the device to establish trust with other devices.
H04L 9/32 - Arrangements for secret or secure communicationsNetwork security protocols including means for verifying the identity or authority of a user of the system
51.
STREAM DATA TRANSMISSION METHOD AND RELATED DEVICE
A stream data transmission method, which is applied to a data transmission system, wherein the system comprises a routing scheduling component and a first bus node. The method comprises: a first bus node receiving a data transmission request sent by a client, and forwarding the data transmission request to a routing scheduling component; on the basis of service information carried by the data transmission request, the routing scheduling component scheduling the data transmission request to a service cluster corresponding to the service information; and the first bus node receiving stream data returned by the service cluster, wherein the stream data is generated by means of the service cluster executing the data transmission request, and returning the stream data to the client. In the method, stream data transmission is performed by means of a bus node, thereby supporting decoupling of user stream data transmission from control logic in a serverless system. When request traffic increases, capacity expansion can be performed only on the bus node or a stream data bus component in the bus node, without the need for performing capacity expansion on an event routing component and a function scheduling component. Thus, the resource utilization rate can be increased, and costs can be reduced.
The present application provides a page generation method, which is applied to a migration process for a historical page. The method comprises extracting a control image from a page image of a historical page input by a user, and then matching a corresponding control, thereby enabling batch generation of new pages with minimal manual intervention, and improving page generation efficiency while reducing page generation costs.
A method includes: A cloud platform provides a network configuration interface for a user, where the network configuration interface prompts the user to enter network feature information, and the network feature information indicates a cloud service network needed for implementing a target cloud service. The cloud platform provides a network association interface for the user, where the network association interface prompts the user to enter association information, the association information indicates a plurality of service nodes associated with the cloud service network, the service node is used to deploy an application of the user, the application is used to implement the target cloud service, and the plurality of service nodes are deployed based on a resource in one or more resource deployment regions. The cloud platform establishes an association relationship between the plurality of service nodes based on the network feature information and the association information.
A method and an apparatus for pre-aggregating time series data are provided. The method includes: obtaining at least one piece of time series data; determining a pre-aggregation method based on the identifier of the at least one piece of time series data, where a preset binding relationship exists between the pre-aggregation method and the identifier of the at least one piece of time series data; processing the at least one piece of time series data using the pre-aggregation method, to generate pre-aggregated data; and when a usage of a memory satisfies a trigger condition, writing the pre-aggregated data and the at least one piece of time series data into a disk. The time series data and the pre-aggregated data can be evenly distributed, and storage space utilization can be improved.
A method for performing profiling of a software application (202) to measure parameters related to the running of the software application is provided. The method includes instrumenting a portion of the software application to generate a plurality of metric measurements related to the portion of the software application. A context is then attached to the generated plurality of metric measurements generated by instrumenting the portion of the software application. The context is an object of interest when measuring one or more parameters, and the context is related to the portion of the software application that has been sampled. The method uses the attached context to build a flow, related to one or more parameters related to the running of the software application.
A model deployment method and a management platform. The method comprises: acquiring a plurality of model blocks, wherein the plurality of model blocks are obtained by means of splitting a model to be deployed, and different model blocks from among the plurality of model blocks have different resource requirements; creating a computing instance corresponding to each of the plurality of model blocks by using resources of at least one of a plurality of servers, wherein the computing instance corresponding to the model block is assigned, from the resources of the at least one server, to resources that match the resource requirements of the model block; and deploying each model block in the computing instance corresponding to the model block, such that each model block runs in the computing instance corresponding to the model block. The method can avoid resource mismatch, thereby improving resource utilization and the running efficiency of models.
The present application provides a data processing method and apparatus, a device, and a storage medium. During data processing, a first data block is processed in a corresponding processing mode on the basis of the value of data comprised in the first data block. When the first data block is a full target value data block, the first data block is directly stored in a storage space of a thin volume. When the first data block is a non-full target value data block, the first data block is encrypted to obtain a first ciphertext block, and the first ciphertext block is stored in the storage space of the thin volume. While ensuring data security, confusion between a non-full target value data block and a full target value data block caused by conversion between the full target value data block and the non-full target value data block is avoided, and while maintaining the characteristics of efficient utilization of storage resources of a thin volume, encrypted storage of service data is realized, and the utilization efficiency of a storage space is improved.
Provided are a communication method, distributed database middleware, and a computing device cluster. The method comprises: acquiring an access request from a client, wherein the access request is used for instructing a processing operation on batch data of a data table, the data table belonging to a database system, and the database system comprising a plurality of database nodes; determining at least one target node among the plurality of database nodes that is related to the batch data, and determining data corresponding to each target node in the batch data; and on the basis of a parsing-binding-execution (PBE) protocol, sending to each target node an operation request corresponding to each target node, wherein the operation request corresponding to the target node is used for instructing a processing operation on data corresponding to the target node.
A service deployment method includes providing a configuration interface, obtaining service configuration information from the configuration interface, determining a resource deployment policy of a target service based on the service configuration information and a number of service copies, and deploying the target service on a plurality of service units according to the resource deployment policy, where the resource deployment policy indicates that one service unit bears at least one user group of a tenant, every Rf service units bear service copies of a same user group, and Rf is the number of service copies. For any user in any user group, even if a service unit that bears the user is faulty, the user can alternatively obtain a corresponding service from another service unit.
A voice cloning model generation method includes: obtaining results of scoring a plurality of pieces of reference audio; performing training based on the plurality of pieces of reference audio and the results of scoring the plurality of pieces of reference audio by the user, to obtain an acoustic feedback unit; obtaining a first voice data set that is input by the user via the terminal device; and training a voice cloning model based on the first voice data set and the acoustic feedback unit, and obtaining a voice cloning model. In a process of training the voice cloning model, in consideration of preference of the user for different pieces of audio, the results of scoring the plurality of pieces of reference audio by the user are added to the process of training the voice cloning model.
A data compression method and a related device, used for reducing I/O overhead. The data compression method comprises: acquiring attribute information of a logic block, wherein the logic block comprises a plurality of data pages, and the attribute information indicates data sizes of data contained in the plurality of data pages across a plurality of time periods; if the attribute information indicates that the total data size of the data contained in the plurality of data pages during a target time period is greater than or equal to a first threshold, compressing the data contained in the logic block, and storing the result obtained by compression, wherein the target time period includes at least time period among the plurality of time periods, and the time interval between the target time period and the current moment is greater than or equal to a time threshold; and if the attribute information indicates that the total data size of the data contained in the plurality of data pages during the target time period is less than the first threshold, determining not to compress the data contained in the logic block.
An execution plan generating method, comprising: acquiring state information of an infrastructure, and on the basis of the content of the state information of the infrastructure, obtaining a target state of the infrastructure; acquiring the current state of the infrastructure, and comparing the current state of the infrastructure with the target state of the infrastructure to obtain first-level execution topology information; and parsing an EL expression on the basis of each operation in a change operation to obtain second-level execution topology information. According to the present application, in the method, after the first-level execution topology information of an execution plan is obtained, EL is used to perform code-based description on a change trend of a resource attribute, and the EL expression is parsed on the basis of the change operation for each differential resource to obtain fine-grained internal operations of the change operation for each differential resource, so that the content of the execution plan is enriched, the transparency of a resource change execution process is improved, users can intuitively understand a detailed resource change process, and the risks and costs related to manual change verification are significantly reduced.
The present application relates to the technical field of cloud services, and discloses a public cloud technology-based memory management method and system. The method comprises: acquiring a real-time usage status of memory of a first server, the first server being one of a plurality of servers; obtaining a memory scheduling policy of the first server on the basis of the real-time usage status of the memory of the first server, wherein the memory scheduling policy is used for indicating one or more of the following: allocating memory of one or more second servers to the first server for use, canceling the allocation of memory of one or more second servers to the first server for use, and canceling the allocation of the memory of the first server to one or more second servers for use, and the second server is one of the plurality of servers other than the first server; and on the basis of the memory scheduling policy, scheduling memory for the first server. The present application improves the memory utilization of servers.
The present application provides a retrieval method and an apparatus. The method comprises: receiving a query instruction; performing retrieval in an index base on the basis of the query instruction, so as to obtain a plurality of candidate chunks related to the query instruction; evaluating the plurality of candidate chunks to obtain evaluation results of the plurality of candidate chunks, the evaluation results of the plurality of candidate chunks being used for indicating correlations between the plurality of candidate chunks and the query instruction; and determining one or more target chunks from among the plurality of candidate chunks on the basis of the evaluation results of the plurality of candidate chunks, the one or more target chunks being used for constructing a prompt word, and the prompt word being used for instructing a large language model (LLM) to generate a response to the query instruction on the basis of the one or more target chunks. The solution of the embodiments of the present application facilitates improvement of the retrieval quality, thereby facilitating improvement of the effect of applying RAG to a code research and development scenario.
Provided in the present application are a meeting participation method and apparatus, and a terminal device and a computer program product. The method comprises: a first terminal device acquiring a first instruction, and then on the basis of the first instruction, sending a second instruction to a second terminal device, wherein a first user enters a first meeting on the basis of a meeting application on the first terminal device, and the meeting application comprises a plurality of modules having different functions; the first instruction instructs the second terminal device to replace a first module among the plurality of modules to process first data, and the first data comprises part or all of meeting data, which is related to the function of the first module, in the first meeting; and the second instruction instructs the second terminal device to run a second module in the second terminal device to process the first data, the second module has the same function as the first module, and the first data processed by the second terminal device is sent to the second terminal device by the first terminal device/a meeting server, or is collected by the second terminal device. The solution in the present application can improve the meeting participation experience of a user.
A method for automated feedback generation in an adaptable large language model (TTM) agent (304, 404) is provided. The method includes continuously monitoring and analyzing the outputs generated by the TTM (306, 406) in response to user inputs. The method includes detecting a sequence of similar outputs generated by the TTM in response to repetitive user prompts aimed at improving the output. The method includes determining that the final output in the sequence is a preferred output when the topic, subject, or intent of the user input changes. The method includes using the characteristics of the preferred output as reference parameters for future output generation.
A test code generation method, comprising: receiving a test code generation request; on the basis of the test code generation request, determining a domain tag of a source domain and a domain mapping relationship, the domain tag of the source domain being used for indicating a domain to which source code of a test object belongs, and the domain mapping relationship comprising a mapping relationship between the source domain and a target domain; performing assembling on the basis of the domain tag of the source domain and the test code generation request to obtain a code generation prompt; inputting the code generation prompt into a code generation model to obtain test code of the test object in the source domain; and then, on the basis of the domain mapping relationship, mapping the test code of the test object in the source domain to a feature space of the target domain, so as to obtain test code of the test object in the target domain. The method determines the source domain and the target domain by means of adaptive domain identification, generates the test code of the source domain by understanding and analyzing test requirements of different domains, and performs code mapping on the basis of the domain mapping relationship, thereby implementing efficient generation of cross-domain test code.
The present application belongs to the technical field of storage. Provided are a data processing method and apparatus in a distributed file system, and a storage cluster. In the method, on the basis that a distributed file system stores a log of a file in a cache space, persistent storage is performed on the log on the basis of whether the data volume of the log in the cache space meets a condition. During this process, when the data volume of a certain log does not meet the condition, the log is redirected to release an initial position of the log in the cache space, such that the distributed file system can continue to process the next log of the log. Therefore, the possibility of aggregating discrete data over a relatively long time span is provided. The method is applicable to any type of file. Moreover, since the persistent storage process is performed on the basis of the data volume of the log, the log which meets the condition is persistently stored, thereby avoiding the fragmentation of a persistent space, and improving the access performance of the system.
Disclosed in the embodiments of the present application is a solution process control method. In the method, a cloud management platform may determine a first physics discipline type of a first physical equation to intelligently determine, on the basis of the first physics discipline type, one or more of a first auxiliary equation, a first coupling strategy and a first numerical solution strategy which correspond to the first physical equation, so as to configure mathematical equation sets included in first solution tasks of a current physical-field service and/or a solution mode of each first solution task, thereby solving the first physical equation. It can be seen that in the solution, the cloud management platform may intelligently determine one or more of the first auxiliary equation, the first coupling strategy and the first numerical solution strategy which correspond to the first physical equation, so as to conveniently and efficiently determine, for various physical scenarios, mathematical equation sets included in first solution tasks to be solved and/or a solution mode of each first solution task, thereby solving the first physical equation by means of an appropriate solution process.
The present application provides a video generation method and apparatus, and a related device, with the aim of generating a flexible virtual character image according to user requirements. The video generation method comprises: acquiring an input image, source speech, and affective information, the input image comprising a facial image of a target object, and the affective information being used for indicating an affect of a target video to be generated; performing affective synthesis on the source speech on the basis of the affective information so as to obtain target speech, the target speech expressing content of the source speech with the affect indicated by the affective information; generating a target image set on the basis of the input image, the affective information, and the target speech, the target image set comprising a plurality of target images, and the plurality of target images being configured for displaying an image of the target object producing the target speech with the affect; and combining the target image set with the target speech to obtain the target video.
The present application provides a network communication method and a device. The method comprises: receiving service data and an address of a destination service node for the service data from a source service node; acquiring an address of a destination forwarding node on the basis of an association table and the address of the destination service node, wherein the association table records an association relationship between the address of the destination service node and the address of the destination forwarding node; and sending a data packet, wherein the data packet comprises the service data, the address of the destination service node, and the address of the destination forwarding node, and the address of the destination service node and the address of the destination forwarding node are recorded in a packet header of the data packet. The address of the destination forwarding node is used for instructing at least one forwarding node to forward the data packet from the source forwarding node to the destination forwarding node, and the address of the destination service node is used for instructing the destination forwarding node to forward the data packet to the destination service node. The method can reduce the complexity of a network structure and simplify operation and maintenance, while ensuring tenant isolation.
Provided in the present application are a data processing method and apparatus, and a computing device. The method is applied to a first DPU, and comprises: receiving a notification message from a first computing node, wherein the notification message is used for instructing a first DPU to perform a data shuffle operation on a first intermediate result, and the first intermediate result is obtained by means of the first computing node executing a mapping task; on the basis of the notification message, acquiring the first intermediate result from the first computing node, and performing the data shuffle operation on the first intermediate result, so as to obtain a second intermediate result; and performing a serialization operation on the second intermediate result, and sending a serialized second intermediate result to a second DPU, wherein the second DPU is connected to a second computing node, and the second computing node is used for executing a reducing task corresponding to the mapping task. The solution can completely offload a data shuffle stage from a server to a DPU for execution, thereby increasing the resource utilization rate of the server.
This disclosure provides an object download method. The method includes: A first client sends a first seed obtaining request to an object storage service on a first server; the first server sends a seed file of a target object to the first client based on the first seed obtaining request; the first client sends a first multi-source download provider obtaining request to a second server based on the seed file, where the first multi-source download provider obtaining request is used to request to provide a list of at least one client that downloads the target object; the second server sends a first multi-source download provider list to the first client based on the first multi-source download provider obtaining request; and the first client receives a first part of the target object from the first server, and receives a second part of the target object from the second client.
H04L 67/06 - Protocols specially adapted for file transfer, e.g. file transfer protocol [FTP]
H04L 9/32 - Arrangements for secret or secure communicationsNetwork security protocols including means for verifying the identity or authority of a user of the system
74.
Code Processing Method and Apparatus, Device, and Medium
Code inputted by a user is received through a user interface. A feature of context of to-be-completed code is determined based on the code inputted by the user. Next, at least one candidate item of the to-be-completed code is determined from a context database based on the feature of context of the to-be-completed code. The context database stores sample code and a feature of context of the sample code. Then, the at least one candidate item is presented to the user through the user interface.
A video processing apparatus and a cloud application are disposed in a data center. A communication connection is established between the data center and a terminal device. The method includes: performing rendering to generate raw video data, where the raw video data is obtained by rendering raw model data corresponding to the raw video data, the raw model data is generated by the cloud application according to an application operation instruction sent by the terminal device through the communication connection, and the raw video data includes M rendering layers; and encoding the M rendering layers in the raw video data into i encoding layers, to generate a target video bitstream, where a first part of rendering layers in the M rendering layers is encoded into a first encoding layer, . . . , and an ith part of rendering layers in the M rendering layers is encoded into an ith encoding layer; M≥i≥1.
H04N 19/33 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using hierarchical techniques, e.g. scalability in the spatial domain
H04N 19/154 - Measured or subjectively estimated visual quality after decoding, e.g. measurement of distortion
76.
METHODS AND DEVICES FOR IMPROVED RETRIEVAL AUGMENTED GENERATION
Methods and systems for improving information retrieval system, particularly in the context of a retrieval augmented generation (RAG) system. A knowledge base is embedded in vector space as knowledge base vectors using a pre-trained transformer model. Using a set of embedded query vectors and a multi-agent reinforcement learning algorithm, the knowledge base vectors are fine-tuned through actions that shift their positions based on relevance to a nearby query vector. Once equilibrium is reached, the modified vector space may be used for information retrieval to augment queries in a RAG system.
A computer-implemented method for extending a first JavaScript runtime is disclosed. A WebAssembly (Wasm) interpreter is embedded within the first JavaScript runtime based on integrating, in the first JavaScript runtime a runtime component and a Wasm interpreter manager. The runtime component implements, at least, a memory allocator configured to manage a memory space that is mutually accessible by Wasm and JavaScript and a WebAssembly System Interface (WASI) -compliant Wasm runtime. The Wasm interpreter manager is configured to track compiled Wasm modules and instantiated Wasm modules.
A data processing method and related apparatus are provided in embodiments of the present disclosure. The method includes: obtaining, based on a first number of instances of a first cloud-based microservice, a plurality of unified abstract syntax trees (ASTs) corresponding to a plurality of fragments of a target code; and loading, based on a second number of instances of a second cloud-based microservice, the plurality of unified ASTs into a graph database as a code property graph (CPG) of the target code; where the first number is configurable based on a workload requirement corresponding to the first cloud-based microservice, and/or, the second number is configurable based on a workload requirement corresponding to the second cloud-based microservice. The separation of the first cloud-based microservice and the second cloud-based microservice makes it possible to separately configure their respective instances, that is, the first number and the second number are both configurable, separate configuration for the instances of the two cloud-based microservices may be beneficial for improving scalability, resource utilization and latency.
This application discloses a resource provisioning method and apparatus, and a computing device cluster, relating to the field of cloud computing technologies. The method includes: obtaining a first resource request, where the first resource request includes a first resource type of a first virtual resource that is requested to be provisioned; determining, in a plurality of physical resources, at least one physical resource on which a virtual resource of the first resource type is deployed; obtaining a provisioning priority of the at least one physical resource for the virtual resource of the first resource type, where the provisioning priority indicates a provisioning sequence of the at least one physical resource for the virtual resource of the first resource type; and controlling, based on the provisioning priority, the at least one physical resource to provision the first virtual resource.
This disclosure discloses a garbage collection method and apparatus. In the method, a client of a distributed storage system may send a garbage collection request to a server of the distributed storage system. The server may determine, based on the garbage collection request, that a target unit data block is a valid data block, and send a move command to a storage device storing the target unit data block, to cause the storage device to move the target unit data block from a first storage position to a second storage position. As a server moves data between storage positions of storage devices accessed by the server, instead of a client moving data between servers, network bandwidth between the client and the server is reduced.
A data access control method includes configuring authentication information for data access control, where the authentication information includes one or more of an identity clearance level, a resource security level, or an environment security level, the identity clearance level indicates access permission of a user, the resource security level indicates a security level of a cloud resource, and the environment security level indicates a security level of a cloud environment; receiving a data access request from the user; and processing the data access request according to authentication logic, where the authentication logic is for controlling data access permission of the user in the cloud resource or the cloud environment, or is for controlling a flow direction of the data between cloud resources or cloud environments, and the flow direction is from a low security level to a high security level.
Provided in the present application are a monitoring method, a device, a product, a medium and a system. The method comprises: receiving a panoramic image sent by a vehicle, the panoramic image being composited on the basis of images acquired by a plurality of image acquisition devices of the vehicle; displaying a projection image of the panoramic image in a first region of a terminal interface, the projection image comprising a plurality of portions; in response to a first viewing operation input by a user in a first portion, sending a first viewing request to the vehicle, the first portion belonging to the plurality of portions; receiving a first viewing response returned by the vehicle on the basis of the first viewing request, the first viewing response comprising an image acquired by a first image acquisition device corresponding to the first portion, and the first image acquisition device belonging to the plurality of image acquisition devices; and displaying on the terminal interface the image acquired by the first image acquisition device.
Methods and system to dynamically tune an application operating on a cloud computing infrastructure. The dynamic tuning generates a dependency graph based on the applications configurable parameters while operating on cloud computing infrastructure. One or more subgraphs are selected from the dependency graph which include a subset of the parameters of the dependency graph, and corresponding subgraph variations are created. The subgraphs and subgraph variants undergo a simulation used to identify a subset of the subgraphs with viable operation. This subset of subgraphs undergoes probabilistic modelling to generate a ranked set of configurations, which are then implemented in test replicas of the application and evaluated against the production replica. Once the test replicas are determined to outperform the production replica, they are implemented within the production replica.
Disclosed are a data processing method and apparatus, which relate to the technical field of cloud computing. The method is applied to a data backup system. The data backup system pre-stores a data processing policy for target data, wherein the data processing policy is used for indicating a target performance for processing the target data, and the target performance is the performance promised to a user by the data backup system. On this basis, after detecting a target processing task that indicates the processing of the target data, the data backup system executes the processing task by means of M execution units; during the execution of the processing task, the performance of the data backup system for processing the target data is acquired; and when the performance of the data backup system for processing the target data does not reach the target performance promised to the user, the number of execution units for executing the processing task in parallel is increased, so as to improve the performance of the data backup system for processing the target data, such that the data backup system provides for the target data the target performance promised to the user, and the data backup system thus provides for the user the performance that can be promised.
G06F 11/14 - Error detection or correction of the data by redundancy in operation, e.g. by using different operation sequences leading to the same result
A task processing method, applied to an operation and maintenance platform. The operation and maintenance platform comprises an event queue and a task executor. The method comprises: the event queue receives at least one operation and maintenance task of an application, and adds task events corresponding to the at least one operation and maintenance task to the event queue; the task executor fetches a first task event from the event queue, calls a first execution unit in an execution unit resource pool, executes a first subtask, delivers an operation of the first subtask on a resource to the resource, and releases the first execution unit; the event queue receives a first resource event sent by the resource; and on the basis of the state of the resource described by the first resource event, the task executor determines a next task or a next subtask to be processed, and executes the next task or the next subtask. The method is event-centric, and an operation and maintenance task, a subtask, and the state of a resource are processed as an event, so that a task is processed from a smaller granularity. Moreover, an execution unit is not blocked by a waiting state of an operation and maintenance task, so that the resource utilization is improved.
The present application provides a screen generation method and apparatus. The method comprises: acquiring a resource release time interval; on the basis of the resource release time interval, periodically releasing resources in a resource pool, the resources being used for generating a screen; and when a screen generation request is received, using target resources released from the resource pool to generate a target screen requested to be generated by the screen generation request. According to the method, screen stuttering can be prevented or reduced while preventing screen tearing.
Disclosed in the present application are an RDMA-based data transmission method and a related device, which can reduce, to a certain extent, the time cost and performance overhead required in the overall data transmission process. The method in the present application comprises: a first processor providing a memory block to a first RDMA network interface card, wherein the memory block comprises a plurality of sub-memories, including a first sub-memory and a second sub-memory; after a second RDMA network interface card sends a packet to the first RDMA network interface card, which packet carries first data generated by a second processor, the first RDMA network interface card generating a first address for the first data; then, the first RDMA network interface card storing the first address in the first sub-memory and storing the first data in the second sub-memory, wherein the first address is used for indicating the location of the first data in the second sub-memory; and subsequently, the first RDMA network interface card notifying the first processor to read the first address from the first sub-memory, so as to read the first data from the second sub-memory on the basis of the first address, thereby realizing data transmission.
A parallel-strategy search method and apparatus, which are used for improving the flexibility of parallel-strategy search. The method comprises: a parallel-strategy search system acquiring a plurality of parallel strategies, each of the plurality of parallel strategies being used for indicating a parallel manner for training an AI model; providing an input interface, the input interface being used for acquiring input content, the input content being used for indicating at least one strategy screening rule, and the at least one strategy screening rule being used for screening the plurality of parallel strategies; on the basis of at least one strategy screening rule, screening the plurality of parallel strategies, so as to determine a plurality of candidate parallel strategies among the plurality of parallel strategies; and evaluating the plurality of candidate parallel strategies to determine a target parallel strategy, the target parallel strategy being a parallel strategy to be used for training the AI model.
G06F 18/214 - Generating training patternsBootstrap methods, e.g. bagging or boosting
G06F 30/27 - Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
89.
ELASTIC IP ADDRESS CONFIGURATION METHOD BASED ON HYBRID CLOUD SCENARIO, AND PUBLIC CLOUD SYSTEM
The present application discloses an elastic IP address configuration method based on a hybrid cloud scenario, and a public cloud system, which can improve the public network access experience of tenants to a certain extent. The method of the present application comprises: a tenant can provide, to a cloud management platform, access point information of an elastic IP address used by an off-cloud resource, and cloud gateway creation information, wherein the access point information is used for indicating a target cloud data center where an access point of the elastic IP address is located, and the cloud gateway creation information is used for indicating a cloud gateway which is to be created and is bound to the elastic IP address; the cloud management platform can create said cloud gateway in the target data center, and establish a network connection channel between said cloud gateway and the off-cloud resource; and when the remaining tenants in the public network send, to the access point, a first packet of which a destination address is the elastic IP address, the access point can notify said cloud gateway to send the first packet to the off-cloud resource of the tenant by means of the network connection channel, so that the off-cloud resource of the tenant processes the first packet.
The present application provides a cross-cluster application access method, a gateway apparatus, and a related product, capable of realizing communication between cross-cluster applications and reducing the resource occupation of a network proxy component in a cluster. The method is applied to a multi-cluster system, the multi-cluster system comprises an application layer gateway, a first K8s cluster and a second K8s cluster, the first K8s cluster comprises a network proxy component and a first application, and the second K8s cluster comprises a second application. Specifically, the network proxy component intercepts an access request sent by the first application, and forwards the access request on the basis of a routing rule matching the request. When the routing rule instructs to perform application layer processing on the access request, the access request is sent to the application layer gateway, so that the application layer gateway performs application layer processing on the access request, and forwards the processed access request to the second application. When the routing rule instructs to perform transport layer processing on the access request, the access request is sent to the second application.
A method comprises: obtaining, by a source terminal, first media data of a real person and comprising a first video or a first audio of the real person; extracting, by the source terminal and from the first media data, feature data for describing the real person and comprising a facial feature, a morphological feature, or a voice feature of the real person, wherein the facial feature and the morphological feature are from the first video, and wherein the voice feature is from the first audio; sending, by the source terminal and to a cloud server, the feature data; generating, by the cloud server and based on the feature data and a multimodal AI-driven model, second media data of a digital human and comprising a second video or a second audio of the digital human; and sending, by the cloud server and to a destination terminal, the second media data.
A method includes: receiving downsampling policies at a plurality of levels that are defined by a user, where a downsampling policy at each level indicates to downsample data in the plurality of time shards based on a defined sampling time interval after raw data is stored in a plurality of time shards and a defined time interval has elapsed; determining a first downsampling policy in the downsampling policies at the plurality of levels based on a first time interval that has elapsed after first raw data is stored in a first time shard, where the longer first time interval indicates a longer sampling time interval of the first downsampling policy; obtaining first data currently stored in the first time shard; and downsampling the first data according to the first downsampling policy, to obtain second data.
This application relates to the field of cloud computing technologies and provides a cloud disk service change recommendation method and apparatus. An example method is applied to a cloud management platform configured to manage infrastructure that provides a cloud service. The infrastructure includes a storage pool including storage space of a plurality of cloud servers. The storage pool is virtually divided into a plurality of cloud disks. The method includes obtaining running data of a first cloud disk in the plurality of cloud disks. The running data of the first cloud disk includes a flow control parameter and capacity usage. The method further includes determining a change policy for the first cloud disk based on the running data of the first cloud disk and providing a recommendation interface for a tenant of the first cloud disk. The recommendation interface includes the change policy.
Disclosed are a media data processing method and system, and a cluster, a storage medium and a program product, which relate to the technical field of cloud computing. During media data processing, a cloud management platform creates a media processing template on the basis of an AI model and transcoding parameters comprised in an algorithm configuration request; and when receiving a first request sent by a client, the cloud platform processes first media data indicated by the first request on the basis of the media processing template, so as to obtain second media data. A user configures an AI model in a media task processing process by themselves, such that a media processing template has higher flexibility and applicability. In addition, AI models available to a user comprise AI models preset in a cloud platform and AI models customized by the user, such that by increasing the data volume of available AI models, the degree of adaptation between the media processing template and a media processing scenario can be improved, thereby ensuring the reliability of a media processing result.
H04L 65/752 - Media network packet handling adapting media to network capabilities
H04N 21/2662 - Controlling the complexity of the video stream, e.g. by scaling the resolution or bitrate of the video stream based on the client capabilities
H04N 21/637 - Control signals issued by the client directed to the server or network components
95.
RESOURCE ISOLATION METHOD AND APPARATUS, AND COMPUTING DEVICE CLUSTER
A resource isolation method, comprising: acquiring the usage amount of computing resources of each tenant among a plurality of tenants; on the basis of the usage amount of computing resources of each tenant, obtaining the total usage amount of computing resources of the plurality of tenants; and when the total usage amount of computing resources of the plurality of tenants is less than a first lower resource threshold value, increasing the number of threads for the plurality of tenants, or keeping the number of threads for the plurality of tenants unchanged. In the present application, on the basis of the method described in the present application, the usage amounts of computing resources of different tenants can be continuously acquired, and the number of threads invoked by different tenants can be dynamically adjusted on the basis of the usage amounts of computing resources of the tenants, so as to adjust the usage amounts of computing resources of different tenants. Thus, a plurality of tenants can fully use computing resources in a system, thereby improving the utilization rate of computing resources of a processor.
G06F 21/53 - Monitoring users, programs or devices to maintain the integrity of platforms, e.g. of processors, firmware or operating systems during program execution, e.g. stack integrity, buffer overflow or preventing unwanted data erasure by executing in a restricted environment, e.g. sandbox or secure virtual machine
96.
LOG DETECTION METHOD AND APPARATUS, COMPUTING DEVICE CLUSTER, MEDIUM AND PROGRAM PRODUCT
The present application provides a log detection method, a log detection apparatus, a computing device cluster, a computer-readable storage medium, and a computer program product. The method comprises: identifying anomaly sample logs from among sample logs; on the basis of the anomaly sample logs, generating an anomaly log template, the anomaly log template containing common features of the multiple anomaly sample logs; converting the anomaly log template into a regular expression; and, by means of the regular expression, determining whether a log under test is an anomaly log. The present application can comprehensively detect anomaly logs, and consume fewer computing resources.
Disclosed in the embodiments of the present application is a method for determining an initial solution. In the method, preferable preset solutions of a preset mathematical programming model in one or more historical service scenarios can be pre-collected by means of, for example, an uncertainty algorithm, so as to provide deterministic preferable initial solutions corresponding to the mathematical programming model in different historical service scenarios, thereby preventing the uncertainty of solving results caused by the uncertainty algorithm, etc., from being exposed to a user. In this way, in an actual application process, the preset solutions corresponding to the preset mathematical programming model, the similarity between which and a first mathematical programming model meets a specified condition, can be efficiently and quickly obtained and can be used as one or more target initial solutions of the first mathematical programming model, such that the user can be provided with a reproducible and relatively appropriate initial solution required by the first mathematical programming model corresponding to a current service, thereby effectively improving the solving efficiency of the mathematical programming model and meeting the reproducibility requirements of the user.
A cloud service-based data forwarding method applying to a cloud platform, the cloud platform includes a first cloud system, the first cloud system includes a cloud computing device and a forwarding device, and the first cloud system and an extranet system are connected to each other through an extranet gateway. The method includes: The forwarding device obtains a target packet sent by the cloud computing device to the extranet system, and obtains a forwarding attribute of the cloud computing device. When the forwarding attribute of the cloud computing device is a first attribute, the forwarding device forwards the target packet by using the extranet gateway as a next hop of the target packet. When the forwarding attribute of the cloud computing device is a second attribute, the forwarding device forwards the target packet to the extranet gateway based on a routing relationship of the forwarding device.
An SQL defense method includes an SQL defense rule configuration unit configuring corresponding SQL defense rules for a plurality of SQL engines through a unified entrance, where each SQL defense rule includes an effective SQL engine and a corresponding execution action, and the SQL defense rules corresponding to the plurality of SQL engines have a same format; and the SQL defense rule configuration unit sending each configured SQL defense rule to a corresponding SQL engine based on the effective SQL engine in each SQL defense rule, so that the SQL engine performs defense on a received SQL statement according to the obtained SQL defense rule.
Provided in the present invention are a data forwarding method and apparatus based on application network technology. The method is applied to a first acceleration node in an application network deployed in a physical network, and comprises: acquiring a service attribute table sent by a controller, which service attribute table comprises tenant access identifiers and forwarding planes corresponding to the tenant access identifiers, wherein a forwarding plane is determined on the basis of a quality of service requirement configured by a tenant corresponding to each tenant access identifier, and comprises at least a portion of acceleration nodes in the application network; acquiring a data request from a terminal, wherein the data request comprises a target tenant access identifier; on the basis of the target tenant access identifier and the service attribute table, determining a target forwarding plane; and sending the data request to a second acceleration node in the target forwarding plane. In summary, by means of dividing a physical network into virtual networks of different quality of service requirements, network resources of the physical network can be fully utilized, thereby improving the utilization rate of the network resources; and adaptation to diverse quality of service requirements is realized, thereby improving the user experience.
H04L 67/61 - Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources taking into account QoS or priority requirements
