7 results about "Load generator" patented technology

The invention discloses a server non-perception resource scheduling method and system based on cluster telescopic self-adaptive reinforcement learning. The method comprises the following steps: step 1, using a monitoring assembly based on a time sequence database to collect various operation load indexes of a cluster server; and step 2, constructing a function resource scheduler by using a Kubernetes scheduling framework. And step 3, constructing a scheduling decision agent by using a cluster telescopic adaptive reinforcement learning algorithm, inputting a variable-length observation sequence, and outputting a node deployment action. And step 4, using a historical cloud service load data set to construct a load generator to simulate a production environment to train and strengthen an intelligent agent deep neural network. And 5, using the trained scheduling decision agent to make a production environment scheduling decision. According to the method, the existing reinforcement learning algorithm is subjected to elastic cluster-oriented algorithm optimization, the expansibility and robustness of the scheduling algorithm when the cluster executes node scaling are improved, and the method has practical application value.

The invention provides a performance testing method and device for a large language model and related equipment. The performance testing method comprises the steps that multiple testing platforms for testing, testing scene configuration parameters and a structured testing data set for testing are determined; on the basis of the test scene configuration parameters and the structured test data set, synchronous constraint parameters are generated and distributed to the multiple test platforms, and then a preset load generator is used for controlling the multiple test platforms to concurrently execute reasoning test tasks about the large language model; and when the test is executed, collecting multiple dimension performance indexes of each test platform in real time for data analysis, and generating a reasoning performance comparison evaluation report of the large language model in the multiple test platforms according to a data analysis result. According to the method, the cross-platform performance evaluation period is greatly shortened, and the waste of computing resources is reduced. Manual operation errors are avoided, comparability of multi-platform performance indexes is guaranteed, and more accurate and reliable multi-platform performance test results can be obtained.

A method and system for calibrating a simulator instance is disclosed. Specifically, the method and system disclosed herein may receive, by a workload generator, a trace file specifying operations; generate, by the workload generator, a workload based on the operations, wherein the workload includes one or more read or write requests corresponding to the operations of the workload; receive, by the simulator instance, the read or write requests from the workload generator; and process, by the simulator instance, the read or write requests based on a set of simulation parameters to generate simulation results, wherein the simulation instance is calibrated, at least in part, based on the simulation results.

The invention is suitable for the technical field of equipment testing devices, and provides a non-contact load rack device which comprises a base. The supporting box is fixed to the top of the base, a controller used for converting electric energy and feeding the electric energy back to electric equipment is arranged in the supporting box, and a dustproof net is arranged on the supporting box; and the disc type motor is mounted at the top of the supporting box and can be connected with an output shaft of the high-speed motor to be detected to serve as a load generator. According to the non-contact type load rack device, the disc type motor is arranged to provide a special test load for the high-speed motor, the interference of the product load on the durability test of a single motor is cut, the test accuracy and reliability are improved, energy recycling is achieved through the supporting box and the controller in the supporting box, the test cost is reduced, and the test efficiency is improved. And meanwhile, the device is convenient to use and operate, high in safety and capable of meeting test load requirements of various rotating speeds.

Various embodiments described herein relate to a test chamber device an associated methods and non-transitory machine-readable media including a test chamber, a system builder that is configured to build and apply an equipment load to the test chamber; a load maker that is configured to build and apply a predefined load to the test chamber; and a tester which measures action of the equipment load and the predefined load within the test chamber, producing a test state.

The invention discloses a software product performance test method, which comprises the following steps of: 1, planning load test, including the number of concurrent users, typical business processes and response time, and writing a performance test report according to related requirements of items; step 2, creating a vuser script which is divided into three parts of init, action and end, init being an initial part, action being a main body of a test, and end being an exit end; step 3, defining a scene: setting a test environment by using a loadrunnercontroller, and recording a complete script and adding the script into the scene; 4, setting a scene which comprises a static part and a dynamic part; compared with the prior art, the method has the advantages that hardware resources are reasonably distributed, a load generator carrier is selected, an efficient operation scene is set, software performance testing under the high concurrency condition is carried out, a staged and modular testing environment is set, and system performance bottlenecks are checked; optimization of system software is proposed, network hardware equipment resources are adjusted, and user experience is improved.

The application discloses a dynamic load test method and system of a short message communication system, and relates to data processing technology, which comprises the following steps: dividing a plurality of priority categories in advance for data messages; obtaining data messages under a plurality of historical emergency scenarios and classifying them according to the priority categories; clock aligning the classified data messages of each secondary node and selecting data message samples of a typical time period; extracting message features of each data message of any typical time period, assigning weights to the corresponding message features according to the priority of each data message, and constructing a dynamic load generator by using LSTM; extracting message features under a current emergency scenario and inputting them into the dynamic load generator; and comparing the output of the dynamic load generator with the load condition of the current communication system to complete dynamic load testing. The method of the application is used for dynamically allocating channel resources according to the result of dynamic load testing and exposing potential resource bottlenecks of the system in advance.