114 results about "Asynchronous operation" patented technology

A cloud deployment automation system consisting of: a deployment automation device housed in a rack-mountable enclosure, the device comprising: a multi-core orchestration processor configured to execute deployment logic as compiled execution graphs; a storage module operatively coupled to the orchestration processor, the storage storing a set of deployment templates, real-time execution states, telemetry logs, and policy configurations; a secure credential management processing unit embedded in the device, configured to generate, store, and rotate cloud access tokens, API keys, and user-specific credentials, and to provide encrypted access to those credentials during deployment execution; an in-memory workflow execution engine executed by the orchestration processor, configured to analyze a user-defined deployment configuration that includes a declarative specification of infrastructure resources and compile that configuration into a directed acyclic graph (DAG) that represents the resource deployment order, dependency mapping, and rollback relationships, a cloud provider interface subsystem communicatively connected to multiple heterogeneous cloud platforms via appropriate API adapters, the subsystem enabling the orchestration processor to send provisioning requests and receive status events from the platforms; a customizable workflow compiler unit configured to convert graphical workflow definitions or domain-specific language (DSL) scripts into execution sequences that can be used by the workflow execution engine, where the workflow compiler unit supports conditional branching, asynchronous operations, and runtime variable resolution; and A policy enforcement control unit integrated into the deployment automation device, with the policy engine configured to apply organization-specific compliance rules, tagging conventions, security group configurations, and runtime resource limits to all deployment actions in a context-aware manner prior to execution.

One embodiment provides a graphics processor comprising a base die including a plurality of chiplet sockets and a plurality of chiplets coupled with the plurality of chiplet sockets. At least one of the plurality of chiplets including a plurality of processing elements, a distributed shared local memory coupled with the plurality of processing elements, a plurality of matrix engines coupled with the distributed shared local memory, and an asynchronous matrix multiply accumulate (MMA) controller configured to perform an asynchronous MMA operation via the plurality of matrix engines.

The embodiment of the invention provides a distributed training system and a distributed training method. The distributed training system and the distributed training method are used for improving the training efficiency of model distributed training. Training nodes in the distributed training system are used for storing check points in video memories of GPUs of the training nodes and sending the check points to memories of CPUs of the training nodes; the training node is also used for sending the check points in the memory of the CPU to the storage system through asynchronous operation; the training node is also used for sending abnormal information to the management node under the condition that the node state information of the training node is abnormal information; the management node is also used for determining whether the training node is a fault node based on the abnormal information, and isolating the fault node under the condition that the training node is the fault node; and the management node is also used for loading a target check point corresponding to the fault node from the storage system to other training nodes under the condition that the training node is the fault node, so that other training nodes complete a model training task of the fault node based on the target check point.

The invention relates to the technical field of communication, in particular to an RoCE protocol-oriented reliable multicast transmission method, which comprises the following steps of: S1, dividing a multicast source and a multicast group in a hierarchical manner; s2, a sending end sends control type information through a UDP protocol, a receiving end monitors a control message, connection is established through an RDMA CM API, and QP ready state conversion is completed; s3, the sending end and the receiving end adopt an RC mode to carry out reliable data transmission, a forwarding node carries out concurrent forwarding in a Reactor multi-thread mode, and a main thread polls and receives a CQ event, determines a forwarding target node, copies message content to a sending buffer area and initiates asynchronous operation; and S4, carrying out API (Application Program Interface) development, and realizing SEND-RECEIVE bilateral operation and READ-WRITE unilateral operation in a multicast scene. The method has the advantages that a reliable multicast transmission function is realized, and a many-to-many concurrent forwarding scene is supported.

The invention discloses a cloud mobile phone interaction method and device, a storage medium and electronic equipment, and relates to the technical field of cloud computing, and the method comprises the steps: responding to an interaction request initiated by at least one cloud mobile phone client, carrying out the conflict verification of the interaction request, and carrying out the merging processing of the interaction request under the condition of determining that the interaction request conflict exists, generating a unified operation instruction; controlling the cloud mobile phone instance to execute the unified operation instruction to operate the execution instruction, and generating interaction data; and distributing the interaction data to a cloud mobile phone client connected with the cloud mobile phone instance. According to the method, multiple terminals can be supported to initiate interaction requests to the same cloud mobile phone instance at the same time, effectiveness and consistency of operation instructions are ensured through conflict verification and merging processing, real-time performance and accuracy of multi-terminal cooperative control are achieved, and the problems of asynchronous operation, rhythm breakage and the like in a multi-user cooperative scene are effectively solved.

The embodiment of the invention provides a transaction execution flow obtaining method, electronic equipment, a storage medium and a program product, and the method comprises the steps: responding to a received transaction processing request of a target transaction, and obtaining request metadata of the transaction processing request; in the process of processing the target transaction based on the request metadata, acquiring a system thread event and respectively acquiring an application program thread event corresponding to at least one preset event acquisition point, each application program thread event comprising a transaction identifier; determining a target application program thread event from the application program thread events according to the transaction identifier; and generating a transaction execution flow of the target transaction according to the system thread event and the target application program thread event. In the embodiment of the invention, in a complex transaction processing scene facing asynchronous operation between threads or concurrent execution of multiple transactions, the transaction execution flow of each transaction can be accurately tracked, so that the accuracy and the reliability of system performance analysis are guaranteed.

The invention discloses a corresponding management method for IPV4 and IPV6 addresses of the same user of network dual stacks, and relates to the technical field of network address management, network system nodes are divided into a main control node, a replica synchronization node and an edge cache node, a layered management architecture is constructed, hotspot users with frequent write-in are identified, a flow partitioning and path arbitration strategy is adopted, and the IPV4 and IPV6 addresses of the same user of the network dual stacks are correspondingly managed. The method comprises the following steps of: dynamically allocating an auxiliary book group, constructing a state deviation evaluation function, judging the consistency degree according to a version and hash information, selecting a corresponding synchronization strategy, fusing various mechanisms to form an execution layer consistency maintenance closed loop, and performing IPV4 and IPV6 address mapping relation management in a distributed environment. According to the management method, the stability, the accuracy and the reliability of the dual-stack user address mapping management system in a multi-source asynchronous operation environment are effectively enhanced by constructing a layered execution architecture which takes a main control node as a core, a copy synchronization node as a distribution guarantee and an edge cache node as a response terminal.

The invention provides a system and a method for optimizing the cold start time of a. NET Serverless function. The system comprises a dependency analysis module, an AOT compiling optimization module, a dynamic code optimization module, a program set loading optimization module, a type system optimization module, a memory management optimization module, a state storage and recovery module and an adaptive optimization system. Dependence analysis is carried out through static analysis and dynamic tracking, partial AOT compiling and self-adaptive JIT compiling are achieved, program set loading and a type system are optimized, a special memory allocation strategy and lightweight state serialization are achieved, and the strategy is continuously monitored and dynamically adjusted and optimized. The method further comprises the steps of asynchronous operation optimization and security assurance. The method has the advantages that the cold start time of the. NET Serverless function is remarkably shortened, the function execution efficiency is improved, meanwhile, cross-platform compatibility and safety are guaranteed, and a more efficient solution is provided for Serverless calculation.

The invention discloses a system login and operation log asynchronous recording method and system based on a message queue, belongs to the technical field of data security, and aims to solve the technical problem of how to realize security monitoring of login operation behaviors across platforms and among different databases. Comprising the steps that each to-be-monitored platform or system serves as a source end, an asynchronous operation thread is introduced for each source end, login and service operation of the source ends are recorded through the asynchronous operation threads, log records are formed, and the content of the log records comprises login information, operation behaviors and data changes before and after operation; a producer reads a log record of a source end, converts the content of the log record into a uniform format, and sends the content to a message queue; consuming the log record in the message queue through a consumer, and sending the log record to a monitoring platform; and checking and analyzing the log record of each source end through the monitoring platform, and providing an alarm service based on an analysis result.

The embodiment of the invention provides a task scheduling and notification method and device, a storage medium and a program product. In the embodiment of the invention, at least part of asynchronous operation instructions with the incidence relation in the target task are divided into the same group, and during the execution period of the asynchronous operation instructions in the same group, a task scheduler schedules and executes other tasks. Under the condition that all the asynchronous operation instructions in the same group are executed, the task scheduler controls the target task to enter a scheduling waiting state so as to wait to be scheduled again, and then under the condition that the target task is scheduled, task context switching is carried out so that the target task can be switched back to continue to be executed; compared with a scheme that task context switching is triggered once after each asynchronous operation instruction is completed, the task context switching frequency can be reduced by performing task context switching under the condition that the asynchronous operation instructions in the same group are completely executed by taking the group as a unit, and the task execution performance is improved on the whole.

The invention discloses an automatic test method and device for streaming asynchronous jobs, electronic equipment and a storage medium, and the method comprises the steps: generating to-be-tested message data through a message generator, and sending a test message to an Flink job operation environment, the Flink job operation environment comprising a plurality of Flink streaming asynchronous jobs; monitoring the processing logic of the Flink streaming asynchronous operation through an asynchronous monitor, and obtaining a processing result in the Flink streaming asynchronous operation; and through a result verifier, generating a test report based on an expected result message and a verification result of a processing result in the Flink streaming asynchronous job. According to the invention, on one hand, the testing efficiency is improved, and on the other hand, the automatic testing of the Flink streaming asynchronous operation is realized.

The invention provides a non-intrusive asynchronous processing method based on transaction event monitoring. An asynchronous event is triggered through a transaction event monitor after a database transaction is submitted, it is ensured that asynchronous operation strictly follows a transaction life cycle, and the risk of data inconsistency caused by dislocation of a thread pool task and a transaction time sequence in a traditional scheme is fundamentally avoided; by intercepting a business method carrying asynchronous execution annotations and automatically extracting parameters to generate an asynchronous request event, physical isolation of business logic and an asynchronous mechanism is realized, and architecture pollution is eliminated; and further dynamically routing to an instant execution or persistent execution module according to the policy type annotated and declared, so as to provide elastic execution capability for different service scenes. Therefore, on the premise of guaranteeing transaction atomicity, a zero-intrusive and extensible asynchronous processing framework is constructed, and the reliability and maintainability of the system are remarkably improved.

The invention relates to the technical field of large model processing, in particular to a fault tolerance method and device based on asynchronous storage check points in a large model training scenario, and the method comprises the steps: obtaining the state data of the asynchronous storage check points; converting the state data into serialized data by using a serialization tool of a deep learning framework; the serialized data is written into persistent storage through asynchronous I / O operation, the serialized data of the persistent storage is used for recovering the training task in the large model training scene, and the serialized data of the persistent storage allows data writing operation to be executed in the background. Therefore, the problems that in the related technology, in the large model training data storage process, the storage overhead is large due to frequent storage operation, the overall training progress is affected if hardware breaks down, and the training efficiency is low are solved.

The invention discloses a reaction cup arranging device and a cup arranging control method. The cup arranging device comprises a material groove, a conveying belt, a cup fishing mechanism, a cup hitting mechanism, a vertical cache channel, a cup placing mechanism, a rotary cache mechanism and a control module. According to the cup arranging control method, a first part state machine and a second part state machine are included, the first part state machine is used for converting reaction cups in an unordered state into reaction cups in an ordered state to be output, and the second part state machine is used for outputting the reaction cups in the ordered state to other devices. According to the cup arranging device and the cup arranging control method, the reaction cups can be continuously, timely and efficiently supplemented, the operation performance and the automation level of in-vitro diagnosis equipment can be remarkably improved, and the problems that in an existing cup arranging system, the reaction cups are not supplemented in time, operation is not synchronous, and fault processing is complex are solved; and the operation efficiency of the equipment and the detection continuity are improved.

The invention discloses a multi-mode fusion control method for a man-machine interaction system. Firstly, a reinforcement learning algorithm is developed to enable the robot to learn an optimal modal priority strategy by taking human control of the mobile robot as an example; in addition, a multi-modal fusion technology is innovated based on a null-space behavior control (NSBC) method, and after a reinforcement learning algorithm obtains a specific modal priority, the method can seamlessly integrate potential conflict control instructions from multiple modals. According to the NSBC method, the expansion from synchronous operation to asynchronous operation is realized, so that the NSBC method is suitable for a multi-mode HRI system. In order to verify the effectiveness of the framework method, a movement control experiment of the mobile robot in the obstacle environment is carried out. Results show that the method is superior to a traditional single-mode HRI control method and a fixed weighted multi-mode method in the aspects of average task completion time, collision rate, artificial participation degree, task error rate and the like. The questionnaire survey at the last stage of the experiment further verifies the acceptance of the user to the HRI framework provided by the user.

The embodiments of the disclosure relate to a method, apparatus, device and storage medium for application creation. The method proposed herein includes: providing, in an application editing interface, a first configuration control associated with a target processing component; setting the target processing component to an asynchronous operating mode via the first configuration control; and creating a target application including the target processing component, the target application being configured to, during an operation of the target processing component based on a trigger by a first message and in response to receiving a second message, process the second message without waiting for completion of the operation of the target processing component.

The embodiments of the present disclosure relate to a method and apparatus for creating an application, and a device and a storage medium. The method provided herein comprises: in an application editing interface, providing a first configuration control associated with a target processing component; by means of the first configuration control, setting the target processing component to an asynchronous operation mode; and creating a target application comprising the target processing component, wherein the target application is configured such that during a period in which the target processing component is triggered on the basis of a first message to operate, in response to receiving a second message, the target application processes the second message without waiting for the target processing component to complete operation. In this way, the embodiments of the present disclosure can support configuring specific processing components in an application to be executed asynchronously, thereby improving the efficiency of application execution.

The application discloses a memory management method and device, relates to the technical field of heterogeneous computing, and comprises the following steps: obtaining a memory allocation request of a device, and searching for a page-locked memory block corresponding to the memory allocation request to allocate the memory block to the corresponding device; receiving a request for releasing the memory block, identifying a device computing flow associated with the memory block, and recording a synchronization event for each computing flow; storing the recorded synchronization event and metadata of the memory block in a to-be-processed event queue; querying whether an event in the to-be-processed event queue is completed; and when it is queried that a specific event is completed, recycling the corresponding memory block to an idle memory pool, so that the technical problem that in the related art, high-cost memory management function calls are too frequent, performance is low, and it is difficult to cooperate with device asynchronous operation, and data competition and errors are easily caused is solved, and the technical effect that the efficiency of memory allocation and the overall throughput of a system are greatly improved under the premise of ensuring memory safety is achieved.

The invention discloses a method, a device and equipment for converting log data into storage, and the method comprises the following steps: generating a dynamic log file in a memory file system at regular time by utilizing a system log service tool of a Linux operating system; using the system log service tool to read log data to a latest dynamic log file; calling a log merging script of a user mode of the Linux operating system, when it is determined that log file to storage conditions are met by executing the log merging script, reserving at least one dynamic log file according to the sequence of generation time from back to front, and merging the rest dynamic log files according to the sequence of time from front to back; and transferring and storing the log file obtained after merging into a disk file system. According to the method, the problem that log data may be lost due to the fact that log is converted into storage through logate and Rsyslod is used for collecting files into new log files in an asynchronous mode in the prior art is solved.

This invention relates to the field of computers and discloses a client-AutoCAD task processing method and system based on named pipes. The method includes: when AutoCAD loads a DLL plugin, initializing a communication service through the DLL plugin, and establishing a named pipe between the DLL plugin and the client to achieve a communication connection; the DLL plugin obtains the client's request data through the established named pipe, parses the client's request data to generate a parsing result, and executes asynchronous or synchronous task processing based on the parsing result. The advantages of this invention are: it reduces the coupling between the Electron client and AutoCAD, improves communication reliability, supports simultaneous synchronous and asynchronous operations, and avoids blocking the main thread.

The invention provides a large data volume import and export processing method, system and device based on Excel and a medium, and belongs to the technical field of batch data process.The method comprises the steps that statistics is conducted on user query information of a service system, and a unified query model is constructed; creating an asynchronous data processing component, packaging the asynchronous data processing component based on the unified query model, and adding the packaged asynchronous data processing component to a service system; the asynchronous data processing component responds to an Excel-based data import or export request, analyzes user query information by using a unified query model, determines a target operation and a data flow direction, and generates a target task; and the scheduling service receives the target task, starts an asynchronous operation, divides the target task according to the target task and the server resource condition, and feeds back to the front-end page of the business system until the target task is completed. By constructing the unified query model and the asynchronous data processing component, efficient data processing and reasonable resource scheduling are realized, the cost is reduced, the data security is guaranteed, and the user experience is improved.

During execution of software a processing unit issues asynchronous operations such that there are multiple asynchronous operations, such as memory operations, in flight (that is, pending execution completion) from a single set of instructions, such as a wavefront or warp. In some cases, the processing unit executes other operations while the multiple asynchronous operations are pending. Performance monitor circuitry records information, such as asynchronous operation count information, register file scoreboard information, and the like, that allows a software engineer to identify which of a plurality of asynchronous operations caused a stall.

The invention discloses a graphical programming multi-language interactive thread asynchronous calling method and terminal, which receive a first script generated by a graphical programming result; determine whether there is an interface method for calling an external interface in the first script, and if so, generate a callback method corresponding to the interface method and with a unique identifier in the first script according to preset conditions to obtain a second script; execute the second script, send the unique identifier as a parameter when executing the interface method, obtain the return value of the external interface called by the interface method through the unique identifier, and execute the callback method corresponding to the unique identifier; the invention introduces a callback method, corresponds the callback method to the interface method, can suspend the interface method, wait until a corresponding interval of time, and then obtain the return value of the external interface through the callback method for processing, and can execute other parts of the code when the interface method is suspended, thereby realizing asynchronous operation.

The invention discloses a programming method and a related device, a first operation building block is provided, the first operation building block comprises a first operation slot position and a first callback function slot position, the first operation slot position is used for receiving operation information of first operation, and the first callback function slot position is used for receiving function information of a first callback function. In response to the fact that the first operation building block is added to the to-be-executed event, the first programming code can be generated according to the operation information of the first operation, the function information of the first callback function and the first code template corresponding to the first operation building block, and then the event programming code corresponding to the to-be-executed event is generated according to the first programming code. The first programming code comprises a first operation code corresponding to the first operation and a callback code corresponding to the first callback function. The first operation building block is an asynchronous operation building block, the first callback code is executed after the first operation code is executed, other operations can be executed in parallel with the first operation, program blocking is reduced, and operation execution efficiency is improved.

The application discloses a communication flow control method, device and equipment and a storage medium, and relates to the technical field of communication flow control, and specifically discloses the following technical scheme: a step sequence corresponding to a to-be-executed communication flow and a conditional transfer rule are acquired; a target primitive instance matched with a current flow step is acquired from a preset primitive pool, and the target primitive instance is activated; the target primitive instance is driven to execute an asynchronous communication operation, and a step execution result and a condition identifier fed back by the target primitive instance are received; a target step is determined from the step sequence according to the step execution result, the condition identifier and the conditional transfer rule, and the current flow step is updated to the target step. According to the application, the step sequence corresponding to the to-be-executed communication flow and the conditional transfer rule are acquired, and the asynchronous operation and the active feedback execution result are encapsulated by using the primitive instance. Compared with the prior art, the application reduces the state management complexity, eliminates the logic fragmentation, and thus improves the development and maintenance efficiency of the embedded communication flow.

The present invention provides a method, device, electronic device and storage medium for processing concurrent asynchronous operations, including: continuously receiving external access request information; grouping the external access requests according to preset conditions to obtain a request information queue group; for each request information queue group, storing the target URLs therein in sequence according to the receiving timing, and obtaining access data of all the target URLs; and outputting the access data according to the receiving timing. The method, device, electronic device and storage medium provided by the present invention solve the problem of serious coupling between various modules in concurrent processing, and realize efficient sequential response to high-concurrency access requests.

The device may queue the threads for execution by the thread pool, including configuring the environmental context with the cancel token. Based on the call of the thread to the synchronization method, the device may determine that the thread uses a green thread model and identify a cancel token from the environmental context. Based on the thread using the green thread model, the device may invoke an asynchronous method that performs tasks of the synchronization method using asynchronous operations, including passing a cancel token to the asynchronous method. Based on the state change identifying the cancel token, the device may terminate the asynchronous operation.

The invention provides an application software performance optimization method based on a cache technology, and relates to the technical field of computer software. The application software performance optimization method based on the cache technology specifically comprises the following steps of S1, cache introduction, S2, read-write performance improvement, S3, cache Miss, S4, short-term massive failure avoiding, S5, cache consistency, S6, advanced processing, S7, delayed processing, S8, parallel asynchronous operation and S9, optimization verification. According to the application software performance optimization method based on the cache technology, by introducing the local cache, the pressure of the distributed cache can be relieved, the network and serialization overhead can be reduced, the reading performance can be optimized, and when the request hits the cache, the request can be directly returned, so that the 10 reading is omitted, the reading cost is reduced, the writing is optimized, the writing operation is merged in the cache, and the performance of the application software is optimized. The IO equipment can perform batch processing, and the writing cost is reduced, so that software can run quickly.

The invention provides a local analysis method for a large-scale GLTF model, and the method specifically comprises the steps: configuring a virtual browser environment in a server-side environment, and solving a problem of compatibility between a server-side development platform and a Web side; the file structure of the GLTF is analyzed, and the GLTF file is analyzed through a Web end 3D display module; an asynchronous calling mechanism is adopted for function calling, and analysis and rendering are conducted through synchronous calling after asynchronous operation is completed; the method comprises the following steps of: after reading a cache file, analyzing data into a cascade data structure; analyzing the picture from the GLTF binary file and storing the picture as an independent picture file, and obtaining and mapping material textures; and determining a data storage format, naming the cache file, and analyzing the file to the local in combination with a model import processing mechanism. The invention provides a technical scheme which is used for improving the display efficiency and can finish the loading of models and animations in a local display engine, mainly aiming at the problem that the display efficiency of the GLTF file is low when the GLTF file is displayed in a webpage.

The invention provides an asynchronous event management system, method and device for multiple department logo platforms and a storage medium, and relates to the technical field of vehicle data transmission, and the method comprises the steps that a platform obtaining module obtains a first department logo platform in a communication preparation state; the callback function configuration module configures a connection state callback function, a receiving state callback function and a sending state callback function; the first asynchronous operation module asynchronously operates the three callback functions on the first ministry label platform to obtain first write-in data; an Nth asynchronous operation module asynchronously operates the three callback functions on an Nth logo platform to obtain Nth write-in data; and the data fusion storage module takes union set data of the first write-in data to the Nth write-in data to obtain fusion write-in data and store the fusion write-in data. The technical problem of low data efficiency during multi-platform interaction caused by data transmission depending on a synchronous uploading mechanism in the prior art is solved, and the efficiency, accuracy and reliability of vehicle data transmission are improved.