44 results about "Polling" patented technology

The invention discloses a message issuing speed control method and system based on an AI large model. The method comprises the steps of collecting multi-source data in real time and constructing a multi-dimensional fusion feature vector; based on the BERT-CNN distillation architecture, detecting abnormity and executing hierarchical fusing; based on the fusion feature vector, using a reinforcement learning model to predict message issuing rates of a plurality of time windows in the future and dynamically adjusting priorities; in the offline stage, a mixed integer programming model is adopted to generate minimum cloud resource cost, and in the online stage, a deep reinforcement learning model is adopted to carry out real-time resource fine tuning; based on the node health degree, calculating a traffic allocation weight through a smooth weighted polling algorithm so as to carry out load balancing; different business scenes are adapted through domain adversarial training and multi-task learning, and different languages are adapted through multi-language BERT coding. The message rate is accurately controlled to improve the system stability and the resource utilization rate; low-cost scheduling fine tuning of cloud resources is realized; various business scenes and languages can be quickly adapted, and the development cost and time are reduced.

A system and method for using a semi-polling model to monitor a Non-Volatile Memory Express (NVMe) completion queue (CQ). The method, implemented at an interrupt thread, includes receiving an input / output (I / O) request from an application and submitting a submission queue (SQ) entry into an NVMe SQ in response to the I / O request. The method further includes registering for notifications from a polling thread after submitting the SQ entry and receiving, from the polling thread, a notification of the presence of a CQ entry within an NVMe CQ. Additionally, the method involves removing the CQ entry from the NVMe CQ and notifying the application of completion of the I / O request.

Systems and methods for determining and reporting actual utilization of a core of a central processing unit (CPU) of a host. Prior to implementation of aspects of the present disclosure, running a poll querying endpoints of a process for work appears to the host's operating system as busy work (e.g., taking full use of the core for the poll duration). However, only a percentage of the duration of the poll is used to process a task of the process, where the remaining duration of the poll is spent querying the endpoints (idle time) and the core is not performing a task. Accordingly, a core utilization reporting system and method automatically detects the processing time of the tasks of a process, determines actual CPU utilization of the core based on a percentage of the time the core is busy polling (doing effectively no work) versus doing actual work (processing a task).

The invention discloses a system certainty enhancement method, which comprises the following steps of: S1, defining a core judgment reference: defining a key CPU (Central Processing Unit) and a key real-time task, detecting the ready states of the key CPU and the key real-time task, and if the ready states exist, triggering a periodic forced scheduling mechanism; s2, deterministic transformation of a scheduler: cutting unnecessary paths, simplifying a user mode preemptive path, and providing an SMP forced scheduling mechanism and a silent background thread; step S3, real-time interrupt processing: key interrupt is set as a high priority, a key CPU is isolated, the bottom half part is limited, and a periodic forced scheduling mechanism is activated; step S4, protocol stack determinacy transformation: optimization is carried out from four aspects of NAPI polling, congestion fragmentation, Qdisk delay and filtering rules; step S5, priority signal processing; and step S6, deterministic transformation of the container: releasing CGroup limitation, binding a key CPU, forbidding NAT and port mapping, using a native IO and closing a non-core daemon process.

The present application relates to the technical field of data pushing, in particular to a method for pushing and real-time previewing and updating of stream file operation based on SSE, which comprises the following steps: collecting a cursor associated sequence, performing comparison, extracting density characteristics, matching and determining a threshold interval, determining length, establishing a pushing load, merging a ratio quota, generating a cursor latching instruction, monitoring an instruction state, grabbing a character bottom layer code, comparing a preset escape sequence, obtaining a truncation identifier, extracting a pre-node end code, combining a full-zero state span value, performing numerical operation, and generating a view rendering signal.In the present application, the transmission slice length is dynamically controlled by frequency comparison combined with interval threshold to construct the load, the quota is compared by table lookup to prevent resource exhaustion and blockage, the bottom layer storage bit XOR span analysis is used to accurately locate the truncation boundary to trigger cross-section rendering, and fixed polling redundant communication overhead and server connection waste are eliminated, thereby ensuring low-latency delivery of batch stream file change data and smooth redrawing and rendering of the front-end page.

The invention discloses a CAN bus hybrid scheduling and fault-tolerant communication method and system for multi-module equipment, and the method comprises the steps: receiving and confirming an instruction of an upper computer in real time through a hardware interruption mechanism, and endowing the instruction with the highest priority through an event-driven task; serial polling communication is executed on a plurality of lower computer modules based on the state machine model, and different fault-tolerant strategy parameters including the maximum number of retry times and timeout are called according to instruction types; in a response receiving stage of the lower computer module, establishing an application layer session verification mechanism based on a two-factor session verification mechanism of an identifier and a data field; and under the condition that the lower computer module has no response, fault isolation is realized by adopting a limited number of retry and fault node skipping strategy, and system-level communication blockage is prevented. Compared with the prior art, the real-time performance, the certainty and the robustness of multi-device CAN bus communication are remarkably improved, and the method is suitable for the fields of power supply control, embedded systems and the like.

The application discloses a kind of interrupt fault tolerance and acceleration response method under heterogeneous platform, including, in interrupt service program, response accelerator message signal interrupt, read the first interrupt state indicating data transmission completion and the second interrupt state of user logic interrupt request, if both are no interrupt, be identified as user logic interrupt, increment kernel state interrupt count value, if second state has interrupt request, also increment count value, if the first state shows data transmission completion, then handle as data transmission interrupt;In application layer thread, obtain kernel state interrupt count value, compare with application layer processing count value, if kernel state count value is greater than application layer count value, send data to accelerator and increment application layer count value, if equal, thread hibernates first preset duration and retry, the application cooperates with application layer, solves interrupt loss problem, uses IOCTL polling acceleration response, and ensures system stability by intelligent hibernation.

The application discloses a kind of based on the write optimization scheduling method of key-value storage system of coroutine, task scheduler and key-value storage system, by encapsulating write request as coroutine task, avoid operating system kernel context switching storm;Second, by introducing the queue state perception scheduling strategy for three queues, efficient overlap of processor execution phase and disk execution phase is realized, the resource utilization of processor and disk is significantly improved, and the scheduling strategy is according to the order of I / O task completion queue priority highest, post task queue second, write request queue lowest, so that the key processing path after disk persistence completion can be executed preferentially, and the next batch of write request is awakened in time, so as to reduce the overall waiting time of write request and system write delay;Third, the completion state is detected by independent I / O polling task, instead of using interrupt or callback mode, avoid the kernel state and user state switching overhead caused by interrupt processing.

The invention discloses a method for realizing polling. The method comprises the following steps: receiving triggering of polling trigger configuration; aPI request configuration information and polling rule configuration information are received. The API request configuration information is used for configuring an API request needing to be sent during data polling, and the polling rule is used for configuring attributes related to polled data; based on the API request configuration information and the polling rule configuration information, a polling trigger is generated, and the polling trigger is used for achieving a data polling function corresponding to the API request configuration information and the polling rule configuration information. By means of the scheme, if a user hopes to achieve the polling function for the data, software codes do not need to be written manually, the user only needs to trigger the polling trigger configuration operation and input the API request configuration information and the polling rule configuration information, and the efficiency of achieving the polling function is improved.

The invention discloses a process-level power consumption analysis method and device, electronic equipment and a storage medium, and relates to the technical field of computers.According to the process-level power consumption analysis method and device, due to the fact that an event-driven monitoring mechanism deployed in an operating system kernel is adopted to achieve accurate capture and time metering of process execution events, the process-level power consumption analysis efficiency is improved; therefore, delay and sampling omission caused by a polling mechanism are avoided, extra overhead caused by frequent system calling is remarkably reduced, and therefore the technical problems that an existing process-level power consumption monitoring method is poor in real-time performance and large in data updating delay, and short-life-cycle processes are omitted in a sampling interval can be solved. The technical effects of high real-time performance and high-precision monitoring of process energy consumption and meeting of the requirements of edge computing and other scenes for refined management of system energy efficiency are achieved.

The invention relates to the technical field of API (Application Program Interface) data processing and machine translation, and discloses a precise translation and asynchronous blocking processing method for API response body content, which comprises the following steps that: a tree path encoder identifies a field to be translated by utilizing a reflection mechanism; and generating task metadata containing a path vector of the to-be-translated field in the JSON nested structure according to a path coding formula. The identifier generator and the serializer generate a global request serial number and task snapshot data and distribute asynchronous tasks. A regular cleaner generates a cleaned text, and a semantic compensation extraction algorithm executes logic truncation and splices a semantic compensation window to generate a to-be-translated input sequence. And the thread pool concurrently schedules a third-party engine to obtain translation blocks, and the structure reconstruction engine executes in-situ backfilling to generate a translated response body after eliminating redundant translation contents. And the exponential backoff retry device processes the exception, and the callback polling module adopts a double-track delivery strategy to deliver the result. The problem of context deficiency is solved, and translation continuity is guaranteed.

This application discloses an electric cylinder control method, system, and storage medium, belonging to the field of electric control technology. The electric cylinder control method includes establishing a three-level hierarchical architecture of host computer-controller-driver, which completes a closed loop from instruction issuance and logic processing to power execution; dividing the CPU's runtime of the host computer into several fixed time slices and configuring the time slices; using a timer to enable the controller to generate a reference clock and configuring the reference clock; based on the three-level hierarchical architecture, performing time slice polling scheduling according to the reference clock and the time slices to achieve control of the electric cylinder by the driver; by adopting time slice polling scheduling, without relying on an expensive real-time operating system, and due to the matching of the reference clock and the time slices, it effectively ensures the stable switching of the driver's cyclic start, run, and stop state machines, as well as the real-time performance of multi-task concurrent processing in actual use.

The invention discloses a polling arbitration method and device supporting access of multiple master control peripheral devices, and the method comprises the steps: monitoring an access request of the master control devices to a target resource, judging whether the target resource can be arbitrated or not according to an arbiter enable signal when the access requests of the multiple master control devices to the target resource are detected at the same time, and if yes, judging whether the target resource can be arbitrated or not; if the target resource can be arbitrated, determining an arbitration mode according to the mode selection signal, selecting a corresponding arbiter to arbitrate the plurality of main control devices according to the determined arbitration mode so as to enable the main control devices to access the target resource according to an arbitration authorization sequence, ending and exiting; otherwise, not authorizing access to all the main control devices. The arbitration mode can be determined through the mode selection signal, fusion of multiple arbitration modes can be achieved, the effect of the multiple arbitration modes is achieved, the arbitration modes and the arbiter can be expanded according to needs, and the advantages of being good in expansibility and high in flexibility are achieved.

The application relates to a request arbitration method and an arbitrator, and belongs to the technical field of data processing. The request arbitration method comprises the following steps: acquiring a first array representing input requests of each input channel at a current time, the number of elements in the first array being consistent with the number of input channels; and arbitrating an input request with the highest current priority according to the first array and a preset two-dimensional array representing priority relationships of each input channel, wherein the priority relationships of each input channel in the two-dimensional array are configurable, the number of elements in each row or each column of the two-dimensional array is consistent with the number of elements in the first array, and each row or each column of the two-dimensional array represents a priority comparison result of a specified input channel and each input channel. The application no longer adopts a polling mode, but performs arbitration based on the two-dimensional array with the configurable priority relationships of the input channels, so that the defect that the polling method cannot be dynamically changed during polling after the priority is determined is improved.

The invention relates to the technical field of resource management, in particular to a software resource dynamic management method and system. According to the method, a multi-dimensional physical performance evaluation system is established by fusing the frequency of the central processing unit, the bandwidth of the disk and the state of the network buffer, so that the integrating degree of a resource allocation decision and an actual load of underlying hardware is greatly improved; a fluctuation residual set is constructed by using response delay jitter, verification integrity and instruction period deviation, and an information entropy quantization algorithm is introduced, so that the disorder degree and the uncertainty risk of the operation behavior of the storage node are accurately captured; a dynamic security reference line is determined by referring to a cluster whole deviation degree mean value, so that an outlier sub-health server can be effectively identified, state identification overturning is executed, and a penalty factor and a deviation multiplying power reciprocal are combined to generate a weight adjustment factor to endow a storage cluster with a Byzantine fault self-repairing capability. And a software resource scheduling weight and decline polling assignment mechanism ensures that the high-reliability and high-performance node bears a core service request.

The present disclosure provides for a system (100) and method (250) that facilitates a robust and effective solution to an entity or an organization by enabling ingestion and processing of a set of data packets corresponding to large streaming and batched data in a big data eco-system that may be fast and may provide optimal throughput, rate control, throttle and embedded fault tolerance. The system (100) is equipped with a data ingestion module (110) to receive the set of data packets and extract a set of attributes pertaining to the nature of data from the set of data packets received and then poll a plurality of second computing devices (104) based on the set of attributes extracted; and, upon establishing polling, store the set of data packets in the plurality of second computing devices (104).

The invention discloses a database blocking event processing method and device, equipment, a medium and a product, relates to the technical field of data processing, and aims at solving the problem that a blocking event processing suggestion is unreasonable and inaccurate due to the fact that a timing polling mode is large in delay. The method for processing the database blocking event comprises the following steps of: subscribing a lock and an original transaction event on an instance side in real time, and sequentially writing the lock and the original transaction event into an additional buffer area by adopting a local timestamp of an instance to form a continuous lock event sequence; dividing the continuous lock event sequence into a plurality of time slices by a sliding time window with a first time length, and constructing a three-layer space-time lock graph in each time slice; and on the basis of the three-layer space-time lock graph, a graph neural network is combined with heuristic search to generate a processing suggestion of the blocking event.

The invention provides a distributed lock control method and device, equipment, a medium and a product, relates to the technical field of computers, and can consider saving of computing resources and keeping of data consistency. According to the specific technical scheme, a plurality of service threads aiming at a target resource service are obtained; and adding a distributed lock to the target thread, and starting a monitoring thread configured for the distributed lock. And polling the completion condition of the target thread according to the polling interval duration of the monitoring thread, and prolonging the validity period of the distributed lock under the condition that the target thread does not complete service resource processing. And releasing the distributed lock of the target thread and ending the monitoring thread on the basis of the prolonged validity period under the condition that the target thread completes service resource processing. On one hand, the validity period and the polling interval duration are adjusted according to the service load rate to save computing resources, and on the other hand, the validity period of the distributed lock is automatically prolonged to keep data consistency.

The application discloses an instruction scheduling system and method, relates to the technical field of data processing, and comprises an instruction management module, which is used for receiving to-be-executed instruction sets corresponding to respective code blocks; an instruction polling module, which is used for polling and selecting to-be-executed instructions in the to-be-executed instruction sets; an instruction decoding module, which is used for decoding the polling-selected to-be-executed instructions to obtain decoding results; a scalar processing module, which is used for performing scalar calculation according to the decoding results to obtain scalar calculation results; and a neural processing module, which is used for calling the scalar calculation results from the scalar processing module and performing parallel calculation according to the scalar calculation results. The technical problems of extremely complex compiler design, large code size and poor compatibility are solved, the technical effect of not depending on an extreme optimization software compiler is achieved, and the code size is greatly reduced.

Disclosed is method for distributed decision-making in multi-role large language models (LLMs) architecture (mLLMa). Method comprises: receiving user query for initiating conversation between user and mLLMa; using graphical representation (GR) for identifying role(s) associated with user query, wherein role(s) is based on context thereof; assigning relevance score to each role; dynamically passing role(s), to mLLMa; generating role-specific prompt for role assumed by LLM; generating, by each LLM, role-specific response (RR) corresponding to role-specific prompt for each role and context; presenting RR from each LLM to peer LLMs; conducting polling process among LLMs for ranking RR therefrom; aggregating rankings from polling process to determine final ranking of RR for each LLM; selecting RR having highest final ranking as an action-inducing response (AR), and transmitting AR to user for providing user action; and updating GR based on AR and user action.

A method of processing packets in a computer, a computer and a computer program for the same. The computer comprises computing hardware and an OS (70) implemented on the computing hardware, the OS comprising a kernel (171), a ring buffer (72) managed by the kernel (171) in a memory space where the OS (70) is deployed, and a poll list in which packet arrival information is registered. The method executed in the kernel (171) comprises the steps of: starting a kernel thread that monitors arrival of a packet; and dequeuing a packet stored in the ring buffer (72) based on the packet arrival information. The kernel thread is dedicated to monitor the arrival of the packet by monitoring, according to a polling model, whether the packet arrival information has been registered in the poll list.

The application discloses a hybrid polling method, device, equipment and readable storage medium. The number of instructions to be sent is acquired from a host computer. It is judged whether the number of instructions is greater than a polling threshold. If yes, a polling mode is entered to process instructions generated by the host computer. If no, an interrupt mode is entered to process instructions. In the application, the number of instructions waiting to be processed on an I / O path is acquired to judge system load change, so that the timing of interrupt in hybrid polling switching is more accurate, and thus the system energy consumption is reduced while the performance advantage is maintained.

This application provides a command distribution method and control unit based on a routing table. The command distribution method includes retrieving a first command from multiple command distribution queues using a polling or weighted polling method. The command type of the first command includes management commands, read commands, and write commands. Based on the routing table, a first processor is determined to handle the first command. The first command is then sent to the first processor. The routing table is used to characterize the processor load. The load is characterized as a correspondence between the processor and the number of commands currently being processed, or a total load value. The total load value is the sum of the products of the load value for each command type and the number of commands, a load value calculated based on utilization, or a load value calculated based on power consumption. The correspondence between the processor and the number of commands currently being processed includes a correspondence between the processor and the total number of commands currently being processed, or a correspondence between processors categorized by command type and the number of commands currently being processed.

This invention discloses a CAN bus hybrid scheduling and fault-tolerant communication method and system for multi-module devices. The method includes: receiving and confirming host computer instructions in real time through a hardware interrupt mechanism and assigning them the highest priority through an event-driven task; performing serial polling communication on multiple lower-level modules based on a state machine model, and calling different fault-tolerant strategy parameters according to the instruction type, including the maximum number of retries and timeout; establishing an application-layer session verification mechanism based on a two-factor session verification mechanism of identifier and data field during the lower-level module response reception phase; and employing a limited number of retries and a fault node skipping strategy to achieve fault isolation and prevent system-level communication blocking when the lower-level module does not respond. Compared with the prior art, this invention significantly improves the real-time performance, determinism, and robustness of multi-device CAN bus communication, and is applicable to power control, embedded systems, and other fields.

The invention belongs to the technical field of memory performance testing, and particularly relates to a high-performance benchmark testing system and method for UFS equipment, testing equipment and a storage medium, and the high-performance benchmark testing system for the UFS equipment can accurately and stably measure the performance limit of the UFS equipment. Due to the fact that the test system is executed in a kernel mode, an uninterrupted pure IO polling engine model is adopted, and a software lock on an IO path is eliminated, the software stack expenditure is reduced to the maximum extent, performance data very close to the physical limit of a UFS storage hardware subsystem can be measured, and the precision of a test result is remarkably higher than that of a traditional test method; a strict environment locking mechanism is executed through the test environment management module, a highly consistent test environment is created, external interference is eliminated, and a test result has extremely high stability and repeatability.

The invention provides a vhost controller management method, device and equipment and a readable storage medium, and the method comprises the steps: responding to a vhost target process starting event, creating a plurality of reactor threads, and binding each reactor thread to a different calculation unit; the method comprises the following steps of: configuring a schedulable working thread of each reactor thread, and respectively configuring an independent storage engine client instance for each working thread; according to virtual annular queues vRing which are configured in a virtual machine in advance and correspond to the number of the working threads, different vRing is configured for each working thread; and respectively creating a polling task on each working thread. Through the technical scheme of the specification, a plurality of reactor threads and corresponding working threads are created for a single vHost controller, and a plurality of vRings of a virtual machine are respectively bound to different working threads, so that full utilization of a plurality of computing unit cores by the single controller is realized.

The invention provides a multi-thread fetch scheduling system, which relates to the technical field of computer system structures, and comprises a thread dynamic scheduling module, which can dynamically update the priority of threads according to the polling count and waiting time of the threads, so that each thread can obtain scheduling opportunities in the fetch process, and the scheduling efficiency is improved. The problem of thread starvation caused by the fact that low-priority threads cannot fetch instructions for a long time in a traditional fixed priority mode is avoided, and scheduling fairness is improved; meanwhile, when a certain thread is subjected to cache miss, the buffer management module can suspend the instruction fetching request corresponding to the thread, and the scheduling logic can be immediately switched to other threads to continue instruction fetching, so that the situation that the whole front-end assembly line is blocked by a single thread is avoided, and the service life of the whole front-end assembly line is prolonged. And the overall instruction fetching efficiency, the throughput rate and the effective instruction fetching bandwidth of the system are remarkably improved.

A method (800) for performing a polling process to collect information about the status, attributes, or settings of individual nodes in a connected network (100); the method (800) includes: sending a request (S801) from a first node (110) in the connected network (100) to a second node (121-12n) in the connected network (100), wherein the request includes an expected response about the status, attributes, or settings of the second node (121-12n); sending a response (S803) from the second node (121-12n) to the first node (110) when the actual status, attributes, or settings of the second node (121-12n) are different from the expected response (S802), wherein the response includes information about the actual status, attributes, or settings of the second node (121-12n); and avoiding sending a response (S804) from the second node (121-12n) when the actual status, attributes, or settings of the second node (121-12n) are the same as the expected response (S802).

The application relates to a task processing method and device, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring a first batch of tasks in a current polling process; performing initialization processing on the first batch of tasks in a task pool according to the first batch of tasks, and obtaining a plurality of first task groups; determining whether there is a to-be-processed task group according to a maximum thread number and polling identifiers of a plurality of second task groups of a second batch of tasks in a previous polling process; when it is determined that there is a to-be-processed task group, selecting the to-be-processed task group and at least one first task group from the task pool according to the maximum thread number; and calling a worker thread to process the to-be-processed task group and the at least one first task group. The maximum thread number is the number of the worker threads. The method can reduce a task processing failure rate.

The invention discloses an asynchronous task polling control method and device, equipment, a storage medium and a product. The method comprises the following steps: determining that a polling result of a current polling of an asynchronous task is a first result and a current accumulated polling time length is greater than or equal to a first time length threshold value and less than a second time length threshold value, and updating a polling interval time length of a next polling based on a first updating strategy of time length decreasing; and performing next polling based on the updated polling interval duration until the polling result is a second result or the current accumulated polling duration is greater than or equal to a second duration threshold, and quitting the polling control of the asynchronous task, the first result indicates that the asynchronous task is in an intermediate state, and the second result indicates that the asynchronous task is in a completed state; the first duration threshold value is smaller than the second duration threshold value, and the second duration threshold value is the maximum polling duration of the asynchronous task. On the basis of considering the polling efficiency and the resource consumption of the task, the response speed of the asynchronous task is improved, and the user experience is improved.