82 results about "Performance tuning" patented technology

The invention relates to the technical field of message queues, in particular to a message controllable optimization method and system based on kafka, and the method comprises the steps: obtaining a data attribute, and calculating a data priority value according to the data attribute; the method comprises the steps of establishing a position index cache mechanism based on priority data, persisting partition-offset data in a cache pool to an index table according to the position index cache mechanism, and performing priority consumption of high-priority messages based on persistent priority indexes, including setting a consumption control program architecture, performing priority consumption on the high-priority messages, and performing priority consumption on the high-priority messages based on the persistent priority indexes. Isolating the consumption queue and introducing a message de-duplication operation; according to the method, full-link performance adjustment and optimization are carried out on the premise of preferential consumption of high-priority messages, service scene extension is carried out based on the adjusted and optimized full links, the messages are pulled directionally according to the priorities through a consumption control program, the disorder of parallel consumption of Kafka partitions is broken through, the high-priority messages can directly jump to corresponding positions to be processed preferentially, and the service scene extension efficiency is improved. The consumption delay is obviously shortened.

An electronic device manufacturing system configured to adjust a virtual knob associated with a trained machine learning model. The virtual knob is configured to adjust a representative value of the trained machine learning model to control a performance of the trained machine learning model. Sensor data is provided as input to the trained machine learning model, the sensor data being associated with a substrate manufacturing process performed on a substate. An output value of the modified machine learning model is obtained, the output value being indicative of metrology data.

The embodiment of the invention provides a performance adjusting and optimizing method and device.The method comprises the steps that historical performance data generated in the process that a storage system bears a host task is obtained, and then the host task is set to be stored by a second node when a first node is abnormal, the method comprises the following steps: analyzing historical performance data and an initial performance index input by a user to generate predicted performance indexes after a first node returns to normal, then selecting a target performance index from the predicted performance indexes to determine performance tuning configuration parameters, and finally executing a data synchronization task by using the performance tuning configuration parameters. The difference data between the first node and the second node is synchronized, the purpose of intelligently adjusting and optimizing performance in a storage active-active fault recovery scene is achieved, the technical effects of optimizing resource allocation and improving data synchronization efficiency are achieved, and the problems that in the related technology, resource allocation mechanisms of host services and synchronization tasks are not coordinated enough, and data synchronization efficiency is poor are solved. And the resource utilization rate is difficult to guarantee.

A materials handling vehicle receives a wirelessly communicated performance tuning profile, and responsive thereto, adjusts at least one operating capability. The performance tuning profile can be updated dynamically as the materials handling vehicle is operated in a work environment based upon a number of different factors including operational, operator, environmental, other, or combinations thereof. Additional aspects provide a graphical user interface that allows an individual such as a supervisor to create a library of performance tuning profiles, and create and / or program a rules engine to automatically convey an appropriate performance tuning profile to a corresponding materials handling vehicle responsive to detecting a corresponding event associated with a programmed rule.

The invention discloses a bandwidth determination method and device for a host channel adapter, and relates to the field of computers.The bandwidth determination method comprises the steps that a preset interface of an operating system is adopted to drive and directly access a hardware register of the host channel adapter through the host channel adapter, and data invisibility of an operating system layer in a remote direct memory access framework is penetrated; original values of a receiving byte counter and a sending byte counter maintained by firmware in a hardware register are accurately captured, and the byte increment change is dynamically calculated in combination with a preset time interval, so that the aim of millisecond-level real-time monitoring on the network traffic is fulfilled on the premise of not modifying a kernel and not intercepting a data packet; therefore, the technical effects of high-precision bandwidth measurement, ultra-low performance overhead, full-flow penetration capability and cross-platform expandability are realized, and the fault diagnosis and performance tuning efficiency of the distributed system is remarkably improved.

The invention discloses a performance tuning method and system for a linux system. The method comprises the following steps: acquiring system information and generating a corresponding text file; extracting a key index from the text file, updating an index threshold in real time, and comparing the key index with the updated index threshold to judge whether an abnormal key index exists or not; if the abnormal key index exists, the system performance bottleneck and the tuning suggestion are matched according to the abnormal key index, and then the abnormal key index and the tuning suggestion are displayed; and waiting and obtaining the tuning file, and performing tuning operation according to the content of the tuning file. According to the method, the performance bottleneck can be automatically identified, the optimization strategy can be intelligently recommended, the tuning action can be efficiently executed, and the method has great significance in improving the stability, the response speed and the overall performance of the Linux system under high load.

Techniques for improving system performance based on data characteristics are disclosed. A system may receive updates to a first data set at a first frequency. The system selects a first storage configuration, from a plurality of storage configurations, for storing the first data set based on the first frequency, and stores the first data set in accordance with the first storage configuration. The system may further receive updates to a second data set at a second frequency. The system selects a second storage configuration, from the plurality of storage configurations, for storing the second data set based on the second frequency, and stores the second data set in accordance with the second storage configuration. The second storage configuration is different than the first storage configuration.

The invention provides a performance optimization method and system for aggregate communication operation, electronic equipment and a medium, and the method comprises the steps: executing a closed-loop optimization process taking meta-heuristic search as a core through determining a multi-dimensional parameter space for describing a hierarchical protocol strategy of the aggregate communication operation, a complex mapping relation between different strategy configurations in a parameter space and the actual performance of the system is fully explored; the target code of the candidate strategy is automatically generated and actually measured based on the historical performance data, the actually measured performance feedback is obtained through continuous iteration, and the target hierarchical protocol strategy of the set communication operation is obtained. According to the method, the centralization bottleneck problem in a direct protocol mode is effectively solved, and the communication efficiency and the resource utilization efficiency in a large-scale parallel system are remarkably improved. Moreover, the closed-loop optimization process gets rid of dependence on artificial experience, reduces the complexity and cost of manual adjustment and optimization in a multi-factor coupling environment, and improves the adaptability, performance and transportability of integrated communication operation in different platforms and tasks.

The invention provides a method for detecting the performance of an Android platform. The method comprises the following steps: configuring a performance index set, and injecting predefined multi-dimensional performance indexes based on operation characteristics of an Android system into an Agent; providing an execution tracker, wherein the execution tracker is configured to synchronously capture data generated in the execution process when the Agent executes the task; setting a plurality of test scenes based on the Android system, and generating a test instruction set according to preset equipment configuration parameters, network conditions and system load scenes; executing the test instruction set, and collecting test data in the test process through an execution tracker; and sorting the test data, processing and analyzing the test data in combination with the performance index set, and generating a performance analysis report. According to the method, the performance of the Agent at the mobile terminal can be comprehensively evaluated, the performance bottleneck can be accurately positioned, and the performance tuning direction of the optimized Android platform can be guided.

The invention provides a server control method, device and equipment and a readable storage medium, and the method comprises the steps: deploying a decision engine at a firmware layer of a system stack, and carrying out the data driving based on the current operation state of a server and a historical performance adjustment scheme to automatically generate a plurality of candidate performance adjustment schemes; the performance expression and the security risk of the server under the candidate performance adjustment scheme are further evaluated, and the decision engine combines the estimated performance and the security risk of the candidate performance adjustment scheme to select a target performance adjustment scheme considering both the performance and the operation security; the control parameters of the server component are directly written through the specified interface of the firmware layer, operation system scheduling and software stack transfer links are bypassed, full-process autonomous closed-loop control from performance optimization to safety protection is realized on the premise of ensuring safe operation of the system, the performance adjustment efficiency and the response speed of the server are improved, and the service life of the server component is prolonged. And the performance adjustment cost is reduced.

The invention discloses an application program resource monitoring method and device, electronic equipment and a storage medium, and relates to the technical field of computers, and the application program resource monitoring method comprises the steps that an out-of-band monitoring module is adopted to collect application program resource data, isolation of a monitoring channel and a service channel is achieved, and the service channel is monitored. Meanwhile, a resource use model of a time dimension is constructed based on real data to simulate production environment resource behaviors, and then performance pressure measurement and fault reproduction analysis are supported, so that the problems that in an existing monitoring and control method, due to the fact that an internal agent of a virtual machine is adopted and the virtual machine is not isolated from a service channel, monitoring data distortion is caused, application program operation is disturbed, and the service performance is affected can be solved. According to the technical scheme of the invention, the technical problems in the prior art are solved, and the technical effects of guaranteeing the accuracy of monitoring data, avoiding the interference on the operation of the application program, improving the performance optimization effect and effectively supporting the performance pressure measurement and fault reproduction analysis of the cloud computing system are achieved.

A technique for operating a device is disclosed. The technique includes recording log data for the device; analyzing the log data to determine one or more performance settings adjustments to apply to the device; and applying the one or more performance settings adjustments to the device.

The invention discloses a Ceph performance tuning method, and relates to the technical field of cloud computing distributed storage. Comprising the following steps of: 1, dividing Ceph-osd threads into a service processing thread, a network message transceiving thread, a message distribution thread, a heartbeat thread, a command interaction thread and other threads by analyzing a Ceph-osd system source code, a key service thread model and a monitoring condition during operation; according to the method, starting and exiting conditions of an osd process of a storage node are monitored in real time, a thread number of a Ceph-osd thread is obtained through an htop tool, the Ceph-osd thread is bound to a specified CPU set through a taskset tool according to configuration of a configuration item, and resources are prevented from being preempted by the thread.

The present application provides a device performance optimization method, apparatus, device, medium, and program product, which can be applied in the field of computer technology. The method includes: determining a comprehensive performance evaluation value for each of the multiple target devices based on multiple robustness indicators for multiple target devices in a server cluster; determining, for any target device among the multiple target devices, a performance reallocation value for the target device based on the performance deviation value of the target device and the comprehensive performance evaluation value of the target device when the performance deviation value of the target device is outside a preset threshold range; determining, based on a parameter mapping relationship, a performance parameter adjustment value for the target device based on a performance adjustment ratio for the target device determined using the performance reallocation value and the comprehensive performance evaluation value of the target device; and updating configuration parameters of the target device using the performance parameter adjustment value to optimize the performance of the target device.

The present invention provides a multimodal mechanical performance test management method and electronic equipment thereof, comprising: performing an expansion action on a target expanding head through a preset simulation loading device, and collecting performance test data of the target expanding head in real time; analyzing the performance test data to determine at least one corresponding abnormal test data; determining the corresponding performance adjustment parameters of the target expanding head during the expansion operation based on at least one abnormal test data; and adjusting the performance parameters of the target expanding head during the expansion operation based on the performance adjustment parameters. By performing real-time monitoring and analysis of key performance parameters during the expansion process through the above-mentioned method steps, it is possible to identify and locate the source of the problem when the expanding head has an abnormal operating state, and then dynamically adjust the expansion speed, expansion stroke, thrust setting or lubrication control strategy according to the identification result, thereby realizing closed-loop optimization control of the expansion process and improving the accuracy and efficiency of the expansion process.

The embodiment of the invention discloses an AI full stack performance tuning tool for top-down analysis. A front end, a middle end and a rear end are separated; the front end is used for visually displaying summarized information needing to be concerned on each level of the AI full stack and specific execution information of different granularities; the middle end is used for connecting the front end and the rear end and providing a plurality of configurable visual components; and the rear end is used for receiving the data files in various data formats, processing the data files and storing the processed data files in a plurality of mutually decoupled databases for the middle end to query and call after configuration. The method has the advantages that the working efficiency of developers of all levels can be improved; a developer can be assisted to realize more accurate and deep performance optimization; the personalized configuration requirements of different levels of development on the front-end interface can be met; the large model deployment and use cost can be reduced, and the utilization rate of the computing power cluster system and the computing power chip is improved.

The invention discloses a distributed storage engine performance optimization method and device, equipment and a storage medium, and relates to the field of data storage, and the method comprises the following steps: obtaining a target search domain and a particle swarm individual number; initializing a target particle swarm based on the target search domain and the particle swarm individual number, performing performance optimization on a target distributed storage engine by using a parameter combination scheme corresponding to each target particle in the target particle swarm, and determining a multi-target function value corresponding to each parameter combination scheme based on an optimization target; the optimization target comprises the storage performance and the CPU utilization rate; and based on the Pareto dominance relationship and the multi-objective function value corresponding to each parameter combination scheme, determining a target optimal parameter combination scheme so as to perform performance optimization on the target distributed storage engine based on the target optimal parameter combination scheme. According to the method and the device, the performance of the distributed storage engine is optimized.

A container-based orchestration system includes a master node and a plurality of worker nodes. The master node can receive, from each agent executing on a corresponding worker node, node characteristics associated with the worker node. The master node can determine, for each worker node, one or more parameters corresponding to the node characteristics associated with the corresponding worker node and a node profile of the worker node and provide the parameters to the agent executing on the corresponding worker node. The agent configures the worker node in accordance with the parameters. In response to receiving a request to deploy a pod to a worker node, the master node can select a worker node to receive the pod based on the node characteristics and the pod characteristics. The agent can configure the selected worker node to execute workloads of the pod in accordance with the one or more parameters.

The invention discloses a network performance adjusting and optimizing method and system after an operating system is upgraded, the method is applied to stock equipment after the operating system is upgraded, and the method comprises the following steps: acquiring a time delay value interval of a network card chip, and detecting a network connection state; if the network is connected, traversing all the time delay values in the time delay value interval to detect the network connectivity, adding the time delay values with normal network connectivity into an available time delay value set, and calculating network performance scores of all the time delay values in the available time delay value set; selecting the time delay value with the highest network performance score, writing the time delay value into a corresponding register, and associating the time delay value with a corresponding link connection event; and if the network is not connected, performing a single-machine loopback test on the time delay values with abnormal network connectivity, adding the time delay values with normal single-machine loopback test into an available time delay value set, selecting an intermediate value in the available time delay value set, writing the intermediate value into a corresponding register, and associating the intermediate value with a corresponding link connection event. According to the method and the device, network abnormity recovery or performance optimization after system upgrading can be completed without depending on an external test tool.

This invention discloses a software code optimization method and system based on big data. The method includes the following steps: constructing a code optimization knowledge base; acquiring the target code to be optimized and its corresponding estimated runtime data features; parsing the target code, extracting its structural features, and fusing the structural features with the estimated runtime data features to generate a target code fingerprint; performing similarity matching between the target code fingerprint and historical code fingerprints in the code optimization knowledge base to determine target historical code fingerprints that match the target code fingerprint; and determining a target optimization strategy for the target code based on the optimization strategy corresponding to the target historical code fingerprint. This invention can improve development efficiency, reducing the performance tuning work of senior architects, which can take hours or even days, to automated recommendations at the minute level, greatly improving software iteration efficiency.

A method may include obtaining multiple tunable parameters associated with a data transform accelerator operable to perform data transform operations. The method may also include configuring a resource configuration vector based on the multiple tunable parameters. The method may further include obtaining a target performance metric. The method may also include measuring one or more performance metrics associated with the data transform accelerator. The method may further include automatically tuning at least one tunable parameter of the multiple tunable parameters to obtain tuned parameters in response to a performance metric of the one or more performance metrics failing to satisfy the target performance metric. The method may also include updating the resource configuration vector in view of the tuned parameters.

A container-based orchestration system includes a master node and a plurality of worker nodes. The master node can receive, from each agent executing on a corresponding worker node, node characteristics associated with the worker node. The master node can determine, for each worker node, one or more parameters corresponding to the node characteristics associated with the corresponding worker node and a node profile of the worker node and provide the parameters to the agent executing on the corresponding worker node. The agent configures the worker node in accordance with the parameters. In response to receiving a request to deploy a pod to a worker node, the master node can select a worker node to receive the pod based on the node characteristics and the pod characteristics. The agent can configure the selected worker node to execute workloads of the pod in accordance with the one or more parameters.

Application performance tuning and benchmarking is performed to ensure that performance of an application being monitored is matching a target performance level. State of the art systems fail to consider various dynamic performance characteristic of the application. The disclosure herein generally relates to software performance management, and, more particularly, to a method and system for performance management of distributed software applications. The system determines a context (which captures static and dynamic characteristics of the application) of the application, and based on the context the application is subject to various load conditions to measure performance. Further, tuning is done if the measured performance is not matching the target performance, to improve the performance of the application.

This invention discloses a system performance optimization method and related equipment based on layered decoupling. When a kernel service node detects a cross-process service call, it obtains the call context information, filters target communication data that meets key service filtering conditions from the call context information to identify the current scenario and determine the current scenario identifier data. The hardware abstraction layer service performs performance parameter matching on the current scenario identifier data and performs business logic processing based on the parameter matching results to generate performance tuning data. The kernel service node then executes the performance optimization processing corresponding to the performance tuning data. This invention enables system performance optimization based on layered decoupling, significantly improving performance optimization efficiency. Through the layered decoupling architecture, it greatly reduces the difficulty and workload of code refactoring and adaptation, improves the maintainability and upgrade efficiency of the performance optimization framework, and can be widely applied in the field of system performance optimization technology.

The application discloses a kind of Xinxin terminal compatibility mutual authentication and performance tuning pedestal system, it is related to Xinxin terminal compatibility mutual authentication and performance tuning pedestal technical field, for solving the problem that address book module, service number module and workbench module under multiple Xinxin operating system and domestic CPU architecture are difficult to be uniformly carried, multi-security domain service routing error and terminal performance and compatibility are difficult to collaborative optimization;Around Xinxin terminal pedestal, collaborative work such as dependent reconstruction module, mobile application interface architecture rendering scheduling module is constructed, based on environmental compatibility score, Xinxin terminal pedestal construction and dependent reconstruction are automatically completed, under the premise that unlicensed personal information in public places is not collected, in combination with business module loading sequence optimization, interface scaling and rendering priority control and performance score and Xinxin compatibility test matrix closed loop optimization, improve the unified interface experience and running stability under multiple operating systems, multiple CPU architecture and multiple security domain deployment.

A system for incorporating user feedback in a large language model (LLM) based retrieval augmentation (RAG) tools includes an application for incorporating user feedback (UFA). The UFA receives an input to the LLM from a host device user, generates an LLM output as a response to the input, causes a system user to verify the LLM output and provide user feedback via a human-machine interface (HMI) of the host device, and in response to the user feedback, engages a performance tuning optimizer that modifies the LLM output by: revising domain knowledge, revising system prompts, and performing constrained regeneration of the LLM output. The performance tuning optimizer progressively reduces computational resource utilization, progressively increases computational efficiency, and progressively reduces reliance on human validators and user feedback over time, and wherein the LLM output is a command to one or more systems of the host device.

A system for incorporating user feedback into large language model (LLM)-based retrieval augmentation (RAG) tools includes a user feedback incorporation (UFA) application. The UFA receives input into the LLM from a user of a host device, generates an LLM output in response to the input, prompts a system user to verify the LLM output and provide user feedback via a human-machine interface (HMI) of the host device, and, in response to the user feedback, deploys a performance tuning optimizer that modifies the LLM output by revising domain knowledge, revising system requests, and performing constrained regeneration of the LLM output.The performance tuning optimizer progressively reduces the use of computing resources over time, progressively increases computing efficiency, and progressively reduces the reliance on human reviewers and user feedback, with the LLM output being a command to one or more systems of the host device.

The present application discloses a bandwidth determination method and apparatus for a host channel adapter, which relates to the field of computers. The method comprises: using a preset interface of an operating system to directly access the hardware registers of a host channel adapter via a host channel adapter driver, thereby penetrating the data invisibility of the operating system layer in a remote direct memory access architecture. Furthermore, by accurately capturing the original values ​​of a receive byte counter and a send byte counter maintained by firmware in the hardware registers, and dynamically calculating byte increments based on preset time intervals, the method achieves the purpose of millisecond-level real-time monitoring of network traffic without requiring kernel modifications or intercepting data packets. This method thus realizes the technical effects of high-precision bandwidth measurement, ultra-low performance overhead, full traffic penetration capability, and cross-platform scalability, significantly improving the efficiency of fault diagnosis and performance tuning of distributed systems.

The invention provides a multi-modal mechanical property test management method and electronic equipment thereof, and the method comprises the steps: executing a diameter expansion action on a target diameter expansion head through a preset simulation loading device, and collecting the performance test data of the target diameter expansion head in real time; analyzing the performance test data, and determining at least one item of corresponding abnormal test data; based on the at least one item of abnormal test data, determining corresponding performance adjustment parameters of the target expanding head in expanding work; and the performance parameters of the target expanding head in the expanding work are adjusted based on the performance adjustment parameters. The key performance parameters in the expanding process are monitored and analyzed in real time through the steps of the method, the problem source can be recognized and positioned when the expanding head is in the abnormal operation state, then the expanding speed, the expanding stroke and the thrust setting or lubrication control strategy are dynamically adjusted according to the recognition result, and therefore closed-loop optimization control over the expanding process is achieved, and the expanding efficiency is improved. And the precision degree and the efficiency in the expanding process are improved.