52 results about "Resource specification" patented technology

The invention discloses an inference service resource allocation method, electronic equipment and a readable storage medium, and relates to the technical field of computers, the inference service resource allocation method comprises the following steps: obtaining a model file of an inference service, and analyzing the model file to determine a minimum storage resource demand required for running the inference service; determining a resource specification meeting the minimum storage resource requirement, wherein the resource specification is used for representing nodes required for running the reasoning service and storage resource configuration corresponding to the nodes; and according to the storage resource state of the node in the reasoning cluster, determining the resource configuration strategy matched with the resource specification, and configuring the storage resource in the reasoning cluster according to the resource configuration strategy to operate the reasoning service, thereby solving the technical problems of low automation degree and poor utilization rate of the storage resource when resource configuration is performed on the reasoning service, and improving the resource configuration efficiency. The error risk and cost of manual operation are reduced, and the utilization rate of storage resources is also improved.

The invention is applicable to the technical field of cloud computing resource management and scheduling, and provides a resource scheduling method, device, equipment and medium based on instance specification awareness, and the method comprises the following steps: collecting multi-dimensional resource information and task load information of a plurality of computing nodes in a cloud computing platform in real time, and according to the multi-dimensional resource information and the task load information, scheduling the computing nodes according to the multi-dimensional resource information and the task load information; a target scheduling action meeting the instance specification requirement at the current moment is obtained through a pre-trained multi-dimensional resource specification perception scheduling model, the multi-dimensional resource specification perception scheduling model is obtained through population-based reinforcement learning training mechanism training, and task resource allocation is executed according to the target scheduling action. And the task execution result is monitored, and the multi-dimensional resource specification perception scheduling model is adaptively updated based on the feedback information obtained by monitoring, so that the cloud environment resource utilization efficiency, the task matching rate and the dynamic environment response speed are improved.

The present invention relates to a resource object (26) operated in a modular processing system (10) for execution of at least one task (28) in support of modular processing system operation. Every task (28) is defined by a requested capability specification representing an expected functionality for task execution and / or by a requested resource specification representing at least one processing resource expected for task execution. As tasks describe capabilities and / or resources to be provided for related task execution the resource object (26) can execute autonomous process resource configuration. The present invention also relates to a processing resource pool providing processing resources for use through the resource object (26) during autonomous process resource configuration.

Embodiments of the present application provide a resource reservation method and a resource allocation method, a device, a storage medium and a program product. In the embodiments of the present application, cloud resources reserved for multiple clients are unified into a reserved resource pool, and the reserved resource pool is shared by the multiple clients, so that the binding relationship between the cloud resources and the clients can be eliminated, the reserved cloud resources in the reserved resource pool can be reused by the multiple clients, the utilization rate of the cloud resources is improved, and the resource selling rate of a cloud vendor is improved. In the resource reservation, the overall use condition information of the current cloud resource reservation is reasonably predicted by learning the historical information related to the cloud resource reservation of the multiple clients, and the overall resource reservation quantity of each resource specification is determined in combination with the resource specification and the resource quantity of the cloud resources reserved by the multiple clients, so that the reserved resource quantity can be saved, and the resource reservation cost is reduced.

This application relates to the field of artificial intelligence technology, and in particular to a computing power resource scheduling method and apparatus. This method can overcome the bottleneck of computing power resources of a single physical machine. In response to a first user instruction to create a function, the method displays first resource specification information; the first resource specification information describes the first resource specification corresponding to the user-created main function; in response to a second user instruction to enable a slave function, the method displays second resource specification information; the second resource specification information describes the second resource specification corresponding to at least one slave function attached to the main function; the total computing power obtained based on the first and second resource specifications is greater than the computing power limit corresponding to a single physical machine; the first and second resource specification information are uploaded to a server; the server is used to obtain a first container image corresponding to the main function and at least one second container image corresponding to at least one slave function; a cloud service program is deployed in the first container image.

The technology disclosed describes a Quantum Utility Exchange (QUE) framework for managing and trading Quantum Utility Exchange Traded Resources (QU-ETRs). Providers define quantum resource packages (with time windows, usage limits, and resource specifications) and list them in a Recorded Database. Entities granted usage rights can trade these packages on the QUE through buy, sell, swap, or bid transactions, with all ownership transfers recorded and expired / consumed packages delisted. A Pending Database supports offers to buy or sell, searchable with AI / ML tools. The exchange uses secure blockchain-based communications, encrypted private tunnels, and verification processes to ensure integrity. Interfaces for providers and consumers enable package submission, searching, bidding, and downloading through a networked computer system. QU-ETRs can be traded under flexible contract types (“buy now-use now / later” or swaps), with transaction, validation, and alert mechanisms built into the system. Overall, the method establishes a secure, structured marketplace for allocating and exchanging quantum computing resources.

The invention provides a service resource determination method and device, equipment and a storage medium, and the method comprises the steps: obtaining a target resource type and a to-be-estimated energy measurement parameter, determining a target energy measurement reference value matched with the to-be-estimated energy measurement parameter in a preset reference scheme list, calculating a proportional relation between the to-be-estimated energy measurement parameter and the target energy measurement reference value, and determining a service resource according to the proportional relation. Determining resource specification configuration information corresponding to the target energy reference value, and determining target resource specification configuration information in a preset configuration sorting list corresponding to the target resource type based on the proportional relation and the resource specification configuration information corresponding to the target energy reference value; according to the method, the corresponding resource specification configuration information is dynamically calculated by matching the energy quantity parameter with the reference value in the benchmark scheme list in combination with the proportional relation and the configuration sorting list, so that inaccurate resource allocation caused by empirical judgment is avoided, the accuracy of the resource specification configuration information is effectively improved, and the user experience is improved. And the phenomenon of insufficient resource support or redundant resource waste caused by inaccurate resource allocation is reduced.

The application relates to a resource intelligent allocation method and device of a virtual desktop based on multi-dimensional constraints, which comprises the following steps: identifying the CPU performance type of a blade server, dynamically adjusting the super-division coefficient based on the CPU performance type of the blade server; calculating the total core number of a virtual CPU, calculating the theoretical allocation number of a virtual machine based on the total core number of the virtual CPU, combining the virtual desktop specification demand of a user, the physical resource of the blade and the number of the allocated virtual machines of the blade; calculating the actual allocation number of the virtual machine based on the total core number of the virtual CPU and the upper limit of the physical resource specification of the blade; judging whether the allocation result meets the multi-dimensional constraints based on the theoretical allocation number of the virtual machine and the actual allocation number of the virtual machine; when the allocation result meets the multi-dimensional constraints, establishing an allocation priority rule, selecting an optimal target blade based on the allocation priority rule, executing the virtual machine resource allocation by using the optimal target blade, and updating the virtual machine resource state.

The embodiment of the invention provides a server screening method and device based on multiple screening conditions, a storage medium and a program product. In a server screening method, a plurality of parallel screening conditions are divided into an environmental dependency condition and at least one resource specification condition. Wherein the cloud servers in the first cloud server set can be sorted according to the priority attribute of the environment dependence condition, and a second cloud server set is obtained. And then, according to the sequence of the cloud servers in the second cloud server set, screening out a target cloud server matched with the at least one resource specification condition from the second cloud server set. The cloud servers are sequenced based on the priority attributes of the environment dependence conditions, so that the cloud servers meeting the environment dependence conditions can be preferentially subjected to specification matching, the target cloud server can be screened out by traversing the second server set once, the traversing times during cloud server screening are greatly reduced, and the screening efficiency is improved. And the screening efficiency of the cloud server is improved.

The invention provides a data processing method. The method comprises the steps of obtaining a basic load of a target application; splitting the basic load into at least one minimum execution unit; according to the resource specification required by the at least one minimum execution unit, determining a target resource instance corresponding to the at least one minimum execution unit from a target resource pool, and executing the at least one minimum execution unit by the target resource instance, the target resource pool comprising a plurality of pooled resource instances, and the resource specification required by the at least one minimum execution unit; the plurality of pooled resource instances are obtained by virtualizing physical resources, and the physical resources are arranged in at least one cloud data center managed by the cloud management platform. According to the scheme, waste of computing resources can be avoided, and the execution speed of the application is increased.

The method comprises: generating a new firewall instance virtual machine according to new system resource specification information; obtaining an original firewall instance virtual machine to be updated; calling a configuration lock to add the configuration lock to data of the original firewall instance virtual machine; obtaining the data of the original firewall instance virtual machine, importing the data into the new firewall instance virtual machine, and leading traffic to the new firewall instance virtual machine through a virtual gateway; releasing the configuration lock, and closing and recycling the original firewall instance virtual machine. In the method, the data of the original firewall instance virtual machine is obtained, and the data is imported into the new firewall instance virtual machine, so that the update of the firewall instance capacity can be implemented in the security protection of the firewall without closing the original firewall instance virtual machine and affecting existing production services.

The invention discloses a cloud computing resource intelligent configuration and automatic opening method and system based on a large language model. The method comprises the following steps: receiving and preprocessing an original demand described by a user in a natural language, and generating standardized information; inputting the data into a large language model for deep understanding and intention recognition, and automatically generating an optimal configuration scheme comprising specific resource specifications, deployment positions and load balancing strategies; according to a load balancing strategy dynamically decided by a large language model in the scheme, the resources are optimally distributed in combination with real-time resource load data; the scheme is analyzed and converted into a specific API calling sequence for different cloud service providers through a multi-cloud hybrid management system, and resource opening and deployment are automatically completed. According to the method, end-to-end intelligent processing from natural language requirements to multi-cloud resource automatic configuration is realized, and the configuration efficiency, the resource utilization rate and the agility of cross-cloud management are remarkably improved.

Various embodiments of the present technology generally relate to systems and methods for providing a service model engine. In an aspect a method includes identifying, by a service model engine, a swagger file from a network vendor and parsing the swagger file for API endpoints. The service model engine may generate resource specifications based on the API endpoints and may generate delivery actions for each API endpoint. Each delivery action may include delivery parameters for each API endpoint, The service model engine may also map each of the delivery parameters to corresponding characteristics in the resource specifications to generate parameter mappings. The service model engine may generate a service delivery model including the resource specifications, the delivery actions, and the parameter mappings, where the service delivery model is used by a service orchestration system to deliver respective services.

The invention provides a resource allocation method of a virtual cipher machine and a cloud server, and relates to the technical field of information security. The method comprises the following steps: acquiring a current resource specification grade of a target virtual cipher machine, and a processing resource quota and an encryption card occupation unit corresponding to the current resource specification grade; according to the processing resource quota, distributing a corresponding target processing resource for the target virtual cipher machine from the processing core of the physical cipher machine; according to the encryption card occupation unit, allocating a corresponding target memory resource to the target virtual cipher machine from the memory resources of the physical cipher machine; and according to the encryption card occupation unit, distributing a corresponding target password card and a target encryption channel on the target password card for the target virtual cipher machine from a plurality of password cards inserted into the physical cipher machine. The problem of low resource utilization rate in resource allocation of the virtual cipher machine is solved, and the resource use efficiency of the cipher service in the cloud environment is improved.

The application provides a model production method and device, electronic equipment and a storage medium. The method comprises the following steps: in response to a data access request, accessing training data from a data source and processing the training data to obtain a target data set; in response to a model access request, selecting a target model matched with the access request from a model warehouse, wherein the model access request carries model type information and model accuracy information; in response to a resource scheduling request, determining a target resource through resource specification information, wherein the resource specification information is determined based on the target data set and the target model; and in response to a model training request, executing a training task based on the target resource, wherein the training task is created based on the target data set and the target model. The application can decouple each stage of model production, each stage can complete its own function, and the scalability, robustness, reusability and portability of model production are improved.

The invention relates to the technical field of GPU resource allocation, in particular to a GPU resource dynamic allocation and cross-node migration method and system based on a containerization environment, and the method comprises the steps: monitoring a container state corresponding to a container in real time, and collecting dimension data corresponding to each dimension in a corresponding data collection dimension set; evaluating the GPU resource use condition corresponding to the container at the current moment, and generating corresponding GPU resource evaluation result data; judging whether the container needs to perform data migration at the current moment or not; if yes, identifying and capturing an incremental data layer generated when the container runs; pre-deploying the same basic mirror image layer on the target node, and configuring a required GPU resource specification; quickly migrating the captured incremental data layer to a target node; and starting a new container on the target node based on the basic mirror image layer and the migrated incremental data layer.

This invention relates to the field of GPU resource allocation technology, specifically to a method and system for dynamic allocation and cross-node migration of GPU resources in a containerized environment. The method includes: real-time monitoring of the container status corresponding to the container; collecting dimensional data corresponding to each dimension in a data collection dimensional set; evaluating the GPU resource usage of the container at the current moment and generating corresponding GPU resource evaluation result data; determining whether the container needs data migration at the current moment; if so, identifying and capturing the incremental data layer generated during container runtime; pre-deploying the same base image layer on the target node and configuring the required GPU resource specifications; quickly migrating the captured incremental data layer to the target node; and starting a new container on the target node based on the base image layer and the migrated incremental data layer.

The invention relates to a cloud host hybrid deployment method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring the total resource of each physical node and the resource specification of a cloud host; the cloud host comprises a first virtual machine and a second virtual machine; determining each initial combination scheme of each physical node based on the total resource of each physical node, the resource specification of the first virtual machine and the resource specification of the second virtual machine; for each initial combination scheme of each physical node, calculating a comprehensive distribution rate corresponding to the initial combination scheme, and determining a target combination scheme of each physical node based on each comprehensive distribution rate; and deploying the first virtual machine and the second virtual machine at the corresponding physical nodes according to the target combination scheme. By adopting the method, the resource utilization rate of the physical nodes can be improved.

The invention discloses a resource programmable elastic acceleration method based on YARN, and relates to the field of big data storage computing, the method comprises a user, a resource management master node, a resource management slave node and an application program coordinator, the method provides a dynamic resource application and life cycle management method, and resource idleness and waste are reduced. Through a resource validity period matching mechanism, task resource recovery failure is avoided; a request ReSource interface is provided, a client is allowed to dynamically apply for resources through a YARN API, and resource specifications and use time windows can be flexibly configured; the applied dynamic resources are bound to a specified queue, only tasks of the queue can be used, and the scheduler is preferentially scheduled to a dynamic resource node; after the resources expire, calling a release interface to release the resources, and updating the queue quota; according to the method, the problem of rigidity of YARN static resource allocation is effectively solved, elastic scheduling and dynamic management of resources are realized, and the method is suitable for high-concurrency and high-fluctuation complex service scenes.

The current document is directed to methods and systems that generate recommendations for resource specifications used in virtual-machine-hosting requests. When distributed applications are submitted to distributed-computer-system-based hosting platforms for hosting, the hosting requestor generally specifies the computational resources that will need to be provisioned for each virtual machine included in a set of virtual machines that correspond to the distributed application, such as the processor bandwidth, memory size, local and remote networking bandwidths, and data-storage capacity needed for supporting execution of each virtual machine. In many cases, the hosting platform reserves the specified computational resources and accordingly charges for them. However, in many cases, the specified computational resources significantly exceed the computational resources actually needed for hosting the distributed application. The currently disclosed methods and systems employ machine learning to provide accurate estimates of the computational resources for the VMs of a distributed application.

The application discloses a reasoning service resource configuration method, an electronic device and a readable storage medium, relates to the technical field of computers, and comprises the following steps: obtaining a model file of a reasoning service, analyzing the model file to determine minimum storage resource requirements required for running the reasoning service; determining a resource specification that meets the minimum storage resource requirements, wherein the resource specification is used to represent nodes required for running the reasoning service and storage resource configurations corresponding to the nodes; determining a resource configuration strategy matched with the resource specification according to storage resource states of the nodes in a reasoning cluster, and configuring storage resources in the reasoning cluster according to the resource configuration strategy to run the reasoning service, thereby solving the technical problems of low automation and poor utilization of storage resources when resources are configured for the reasoning service, reducing the error risk and cost of manual operation, and improving the utilization of storage resources.

The invention relates to the technical field of computing power scheduling, in particular to a computing power resource scheduling method based on GPU virtualization, electronic equipment and a storage medium, and the method comprises the following steps: an application end initiates a computing power application through a computing power portal, and the application at least comprises a computing power scheduling task type and a required resource specification; analyzing the application, obtaining a task type and a resource specification, and setting a priority category for the to-be-scheduled instance; according to an analysis result, scheduling the task to a data center physical resource node according to a priority category through a container cloud platform scheduler, and creating a corresponding instance to execute the task; virtual GPU resources are provided for the instance, and the virtual GPU resources are obtained by virtually segmenting physical GPU resources. According to the invention, three core technologies of GPU virtualization, container and virtual machine mixed arrangement and priority scheduling are fused, an efficient, flexible and intelligent computing power resource scheduling system is constructed, and the problems of low GPU resource utilization rate, single scheduling mode, insufficient task priority guarantee and the like in the prior art are solved.

The present application provides a method, system and storage medium for rust detection based on edge computing, which relates to the field of computer vision technology. The method includes: using an enhanced shuffle network to perform lightweight improvements on the backbone network of the target detection model, using a bolt data set to train the lightweight target detection model, and then obtaining a lightweight model that adapts to the edge computing hardware platform through format conversion and quantization; using a number of cores that matches the hardware resource specifications in the edge computing hardware platform, and performing collaborative reasoning on the lightweight model based on real-time acquired bolt images to quickly obtain real-time rust detection results. By performing targeted optimization on the architecture and deployment method of the rust detection model, the speed and performance of the model deployed in the edge computing device can be effectively improved, and real-time robust bolt rust detection can be achieved based on edge computing.

The embodiments of the present disclosure provide a resource scheduling method, apparatus, device and medium. The method includes: obtaining parameter information of the resource object according to a resource application request for applying to create one or more resource objects, the parameter information including the resource specifications of the resource object and the application quantity of the resource object; determining a first resource sequence according to the resource specifications and the cluster monitoring database, the first resource sequence including the correspondence between at least one first cluster and the amount of resource objects that can be created, the cluster monitoring database being used to record the host information and resource object information of multiple clusters; determining the target cluster and the resource object allocation amount corresponding to the target according to the application quantity and the first resource sequence; sending a request to create a resource object to the target cluster according to the resource object allocation amount of the target cluster, so as to create the resource object on the host machine in the target cluster. The method of the embodiment of the present disclosure improves the resource scheduling effect between multiple clusters.

This invention discloses a stress testing method and system for a large-scale task scheduling system based on a Kubernetes cluster, belonging to the field of container orchestration technology. The method first creates a simulation environment and deploys the large-scale task scheduling system on it; using the simulation node tool Kubemark, simulation nodes with the same computing resource specifications as real computing nodes are created in the simulated Kubernetes cluster, and the simulation node size is set; a task load management tool is used to create and terminate the stress testing task load of the task scheduling system; and a task data acquisition tool is used to collect task data in the real environment and collect task scheduling information in the simulation environment to achieve scheduling performance analysis. This method tests the scheduling function and performance of the large-scale task scheduling system, providing reference and verification for functional improvement and performance optimization of the scheduling system, reducing the testing and verification costs in large-scale computing scenarios, and ensuring the stability of the scheduling system.

The embodiment of the application provides a large-specification priority scheduling method and device for GPU resources, dynamically adjusts the queue order based on the large task requirement and the resource specification demand satisfaction, performs the priority matching according to the large resource quantity, sequentially from large to small, performs the matching again from large to small after the matching is successful, preferentially allows the large task to completely occupy the large node, and concentrates the small task in the small node matched with the resource, so that the method can avoid the problems of the small task occupying the large resource and the large task running across nodes, prevents resource fragmentation, reduces the task cross-node communication cost and resource management overhead, and improves the resource utilization.

The embodiment of the invention provides a resource scheduling method and a cloud computing platform, and the resource scheduling method relates to the technical field of cloud computing, and comprises the following steps: receiving a resource request for a target computing instance deployed on a target physical machine, the target physical machine being deployed with computing instances of various resource use types, different resource use types correspond to different resource mapping relationships, the resource mapping relationships represent mapping relationships between virtual resources and physical resources, and the resource request carries a virtual resource specification and a target resource use type; based on a target resource mapping relationship corresponding to the target resource use type, mapping to obtain a physical resource specification corresponding to the virtual resource specification; and dividing the physical resources of the target physical machine to obtain target physical resources conforming to the physical resource specification, and deploying a target calculation instance based on the target physical resources. The resource strong-mixing deployment is realized, and the flexibility and the availability are improved.

The invention provides a cloud platform capacity water level injection method and device based on multi-dimensional resource assessment, and belongs to the technical field of computers.The method comprises the steps that multi-dimensional resource data of at least one resource pool in a target test environment is obtained; according to a preset target water level proportion, the multi-dimensional resource data and the resource specification of a single cloud host, the maximum number of cloud hosts which can be supported by each resource pool under different dimensions is calculated, and the minimum value is taken as the number of target cloud hosts created by each resource pool; and creating the cloud hosts of the corresponding target cloud host number in the corresponding resource pool so as to realize the injection of the capacity water level. According to the method, by adopting the short board principle, it is ensured that the finally injected load cannot fail due to exhaustion of any resource, the test result can truly reflect the comprehensive capacity bottleneck of the cloud platform, and misjudgment caused by evaluation based on a single resource is avoided.

The invention discloses a game asset intelligent management and script analysis method and system based on a multi-modal large language model, and relates to the technical field of artificial intelligence and game development tools. The method comprises the following steps: constructing an AI Agent workflow framework based on a state machine, and supporting a ReAct reasoning mode to realize a thinking-action-observation cycle decision; a three-layer asset naming check and renaming algorithm is realized; an intelligent asset arrangement function is realized; a resource perception map generation function supporting multi-modal input is realized; a visual script intelligent analysis and generation function is realized; a resource specification checking function is realized; and a three-layer data storage and synchronization mechanism is realized. According to the method, the efficiency and accuracy of asset management, map generation and script analysis in the game development process are improved by utilizing the semantic understanding ability of the large language model and the multi-modal understanding ability of the visual language model.

The invention discloses a resource sharing method, device and equipment, a storage medium and a computer program product, the method is applied to a computing cluster, the computing cluster comprises a management node and at least one computing node, and a physical GPU on the computing node is divided into a plurality of virtual GPU instances. The method comprises the following steps: acquiring resource information of a virtual GPU instance on each computing node, and constructing a global resource view on a management node; receiving a creation request of the target task, wherein the creation request comprises a resource specification required by the target task; based on the global resource view, a virtual GPU instance meeting the resource specification is searched for; and in response to the existence of the distributed task and the candidate virtual GPU instance of which the current residual resource quantity meets the resource specification, determining the candidate virtual GPU instance as a target virtual GPU instance, and scheduling the target task to a computing node where the target virtual GPU instance is located. According to the method, the problems of inflexible resource scheduling and resource waste caused by fixed division granularity in the existing hardware isolation type virtual GPU technology can be effectively solved.