4results about How to "Dynamic Load Balancing" patented technology

The application discloses a real-time data caching and distributing method for industrial internet of things, comprising the following steps: S1, collecting real-time data of industrial internet of things heterogeneous devices, marking data priority and initial access heat after preprocessing; S2, building a three-level cache system, and clearly defining the storage capacity, range and time sequence data processing rules of each level; S3, constructing a time sequence convolution network access heat prediction model, combining real-time access heat value to form a two-dimensional evaluation system, and dynamically adjusting the cache storage strategy and data migration path; S4, monitoring the cache node and link load, and planning the data distribution path; S5, establishing an edge node cache cooperation mechanism to realize hotspot data cross-node sharing and three-level cache data synchronization; S6, real-time detection of the running state, triggering redundant backup and path switching when abnormal. The application significantly improves the data cache hit rate and distribution efficiency of industrial internet of things, and reduces the network bandwidth occupation and cloud load.

The present disclosure provides a traffic scheduling method, device, system, equipment and computer program product, relating to the technical field of network communication. The traffic scheduling method in the present disclosure comprises: obtaining a network connection request from a client, the network connection request comprising port information, the port information being determined based on the type of a request resource corresponding to the network connection request, the type of the request resource comprising a hot resource and a cold resource; determining a target processing process matched with the port information, so that network connection requests corresponding to the hot resource are distributedly allocated to multiple target processing processes, and different network connection requests corresponding to the cold resource are convergently allocated to at least one target processing process; and establishing a network connection between the target processing process and the client.

The application discloses a video decoding task scheduling method and device, wherein the method comprises the following steps: in response to a video stream to be decoded, a corresponding decoding thread is created, the decoding thread encapsulates each frame of video data of the video stream into a corresponding decoding task; a load value of a current encapsulated decoding task is calculated; current load states of all video decoding hardware units in a GPU are acquired; each video decoding hardware unit has a unique corresponding task queue, according to the current load states of the video decoding hardware units, a lightest-loaded video decoding hardware unit is selected from all the video decoding hardware units as a target decoding hardware unit of the current decoding task; and the current decoding task is dispatched to the target decoding hardware unit for decoding. The application can improve decoding efficiency in a multi-path video concurrent decoding scene.

The application discloses a data file uploading method, an energy storage power station site data management system and a medium, and relates to the technical field of data transmission. The method mainly comprises the following steps: combining network width, and dividing a data slice into a plurality of data slices; when the idle period of a browser rendering gap is greater than a first preset time length, performing data analysis on the data slices to generate structured metadata required for visual thumbnails, and providing the browser window with the structured metadata for visual display; and when a browser idle event is monitored and the idle period is greater than a second preset time length, selecting a preset number of data slices for uploading. The application scheme can eliminate the page lag phenomenon in the large data file uploading process, and can provide a zero-blockage uploading experience effect for users.