47 results about "Epoll" patented technology

The invention discloses an accompanying robot cloud service system and an accompanying robot cloud method based on a micro-service. The system comprises a terminal and a cloud service platform, wherein the terminal comprises a mobile terminal and an accompanying robot; the cloud service platform receives a connection request initiated by the terminal, extracts and stores a terminal protocol via aprotocol extraction model, then analyzes the connection request of the terminal and performs service processing, and at last feeds a processing result back to the terminal via a response interface; the service layer of the cloud service platform is implemented based on the micro-service, and comprises multiple fine granularities of bottom-layer micro-services, and a protocol extraction mechanism is designed at an interface layer; a data layer adopts a mixed storage mode of MySQL+MongoDB for storing data. According to the system and method, a multi-priority scheduling strategy based on resourcematching is provided from the aspect of scheduling, the utilization rate of system resources is improved, and from the aspect of development, the Mix-IO model of select+epoll is adopted for enhancingthe system concurrency.

The invention relates to a LoRa core network system and a realization method. According to functions of designed gateway, network management server, application management server, data distribution server and database server, a task of operating and managing a LoRa network can be finished. The designed application management server based on Nginx, TCP connection is taken over by Nginx, and parameters of request service are borne by a body in an HTTP message. A service request module is docked with an Nginx module. The realization complexity resulting from managing the TCP connection by an Epoll model is avoided, the system expansibility is improved, and the maintainability is relatively high. A request preprocessing mode accesses the rapid processing and blacklist filtering LoRa network based on a UDP, requests are rapidly transmitted to the data distribution server through the UDP, through blacklist filtering, illegal requests are prevented from entering the application management server through the data distribution server, and the system security and reliability are ensured.

The invention discloses a software license dynamic authorization management method and system based on a local area network and relates to the field of license dynamic authorization in the local network. Network sharing of a license is achieved. The software license dynamic authorization management system comprises a database, a license generating server, a license authorization management server and clients. Concurrent authorization is achieved through an epoll technology; license recycling of invalid networks is achieved through a heartbeat detection technology; compared with a traditional method, the authorization management method and system provide efficient and high-quality authorization services.

The invention discloses a method and a system for realizing access for a great amount of terminals for a helix server, which belongs to the technical field of the IPTV and is used for solving the technical problems in the prior art that when a great number of terminals are accessed, the treatment efficiency of the helix server is low, and the TCP short connection accessed with a great number of terminals cannot be reliably supported. The invention adopts an EPOLL / POLL event polling interface to perform event polling to the linked link circuit, thereby enhancing the capability of the system to access a great amount of terminals, and the invention adopts the socket file descriptor as the index entry of the polling list, thereby improving the retrieval efficiency of the link circuit during the process of accessing a great amount of terminals. The processing capacity of the helix server can be enhanced, so that the TCP short connection accessed with a great amount of terminals can be reliably supported by the helix server.

The invention discloses a TCP channel management method suitable for high-concurrency scenes. The method comprises an access gateway, message middleware, a data analysis service and a TSP business service, wherein the access gateway is connected with a plurality of clients; the access gateway s connected with the message middleware; the message middleware is connected with each of the data analysis service and the TSP business service; and the access gateway is used for providing services such as client access, link validity verification and overtime management. According to the method, an epoll asynchronous communication model is adopted, a task pool, link table management and a link detector are imported, and a task queue pool and a multi-thread technology are combined, and a communication link is decoupled. The method is suitable for high-concurrency application scene characteristics and network communication technology background, and sufficiently considers the advantages and disadvantages of various technologies so as to adopt the advantages and avoid the disadvantages.

The invention relates to the technical field of heartbeat packet control and especially relates to a heartbeat packet asynchronous control method and apparatus. The heartbeat packet asynchronous control method comprises the following steps: step A, performing server socket communication parameter initialization and server communication model initialization, and establishing a monitoring socket descriptor; step B, establishing and initializing a server dynamic list; step C, initializing and starting a system timer controlling a heartbeat packet; and step D, invoking a monitoring function, determining whether a client communication event is triggered, if so, a server main program exiting from the monitoring function, and the server traversing and processing the communication event, and if not, continuously waiting until the client communication event is triggered. According to the invention, by use of an Epoll server model of a linux system, the method and apparatus are simple, reliable and highly efficient, the problems of multithread programming and protocol complexity are overcome, system memory resources are fully utilized, heartbeat packet data is efficiently processed.

The invention relates to an electricity consumption data acquisition and pre-processing method and system. The method includes: 1, creating a file description through an epoll_create () function, and setting a number of socket descriptors to be managed; 2, creating a receiving thread related to epoll; 3, creating a listening socket descriptor listenSock, setting a non-blocking mode, and calling a Listen () function to listen to new connection requests on sockets; 4, in case of a new connection request occurring, calling an accept () function, adding a client socket descriptor to an epoll_data union, setting the client socket descriptor in the non-blocking mode, setting the type of an event to be processed in an epoll_event structure in read and write modes; 5, if the client socket descriptor has data available for reading, adding the client socket descriptor to a readable queue, informing the receiving thread of reading data, adding the received data into a linked list of received data, and adding a fed-back data packet into the linked list of received data to be transmitted by a transmitting thread.

The invention belongs to the network communication field and provides a network communication method and a system. The method comprises the following steps that: a network data transceiving thread and an interception interface are associated, and the interception interface is added to an epoll array; the interception interface intercepts events in the epoll array, and when the interception interface intercepts an event needing transceiving data, different events are processed and the network data transceiving threads associated with the different events are notified to transmit and receive data; and the network data transceiving thread transmits and receives network data according to notification of the interception interface. In the embodiment of the invention, the interception interface is established in the system kernel to intercept the events in the epoll array, to notify the network data transceiving thread to export a socket-associated function symbol table, and to transmit and receive the data by using a standard system call corresponding to the function symbol table in the kernel, so that the transceiving of the data in the kernel process is realized.

The present invention relates to IPTV technical field, and discloses a method and system for implementing access of a large number of terminals to a streaming media server to solve the technical problems of the streaming media server having low processing efficiency and not supporting the access of a large number of terminals by TCP short connections when there are a large number of terminals accessing to the streaming media server in the prior art. The present invention uses EPOLL / POLL event polling interface to poll the events in the established link, thereby improving the capability of the system to accept the access of a large number of terminals. A socket file descriptor is used as an index entry of the polling list in the present invention such that the link retrieval efficiency is improved when there are a large number of terminals accessing.

The invention provides a system for counting the number of online client sides based on a TCP / IP communication protocol. The system comprises a server and at least one client side connected with the server; the server is used for creating a dynamic link table; the dynamic link table comprises a head node and a plurality of child nodes; various nodes are sequentially connected in series through a pointer field next, so that the dynamic link table is formed; the server judges whether a communication event is triggered at the client side by calling an epoll_wait function; if so, the server quits the epoll_wait function, traverses and processes the communication event; and if not, the server waits for triggering of the communication event of the client side in a blocking manner. According to the system and the method provided by the invention, the number of the online client sides is counted in a dynamic link table manner by utilization of an epoll server communication model; furthermore, a client side heartbeat packet processing mechanism is added; in combination with the dynamic link table, a node, where an abnormally disconnected client side is, can be cleaned in time; and thus, the accuracy of counting the number of the online client sides can be improved.

The invention discloses a communication method, a device and a server for Internet of things. The method comprises the following steps: setting a plurality of first selectors so as to receive a connection request from a connection queue in a multithreads manner; determining one first selector to receive the connection request which is currently positioned at the utmost front end of the connection queue according to the amount of data received by the first selectors; establishing the connection to the side of equipment according to the connection request. According to the embodiment of the invention, the plurality of selectors are arranged, the selectors can receive the connection request from the connection queue in a multithreads manner, and further multiple connection requests are processed in parallel, so that the processing speed of the server is higher, and the problems that select and epoll models based on a single thread cannot satisfy the socket read and write requirements of the system, and the professional work thread of the server is blocked as packages are frequently transmitted by the equipment in the prior art are solved.

The invention discloses a method for processing high-concurrency IO based on PHP. The method comprises: Swoole creating one or more processes according to configuration; the Master process creating NReactor threads; creating a coroutine in each Reactor thread; and asynchronously processing the IO task through coroutine scheduling. According to the invention, a Swoole model is used, through the multi-thread Reactor model (based on epoll) and a multi-process Worker, a coroutine is created in each thread. Through coroutine scheduling, an IO task is asynchronously processed, namely, when meetingthe condition of IO time consumption, firstly processing other code logics through coroutine scheduling until IO is completed and then continuing to execute previous codes; according to the method, system resources are fully utilized, IO blockage waiting and system resource waste are reduced, the IO processing capacity under high concurrency is greatly improved, and the problems of Accept performance bottleneck and convulsion group can be solved.

The invention provides a method and device for testing pressure. The method comprises the steps that initialization is firstly conducted, wherein a reading connection pool and a writing connection pool of the epoll are initialized, application for a data area cache is conducted according to preset user customized parameters, and a data package is loaded to the cache from a pressure file; with regard to connection established with a module to be tested, data are read and written in a level triggered mode and trigger events of data reading and data writing are arranged and the connection is added to the reading connection pool and the writing connection pool; the pressure condition is counted at regular time through a counting thread and transmission of pressure data is stimulated at regular time through effective global variables which are continuously transmitted through set identification; after the trigger event of data writing is monitored through a pressure transmission thread, if the global variables are effective, the pressure data in the cache are transmitted to the module to be tested in a polling mode until the pressure number of the transmission period is achieved, and then the global variables are set ineffective. According to the method and device for testing the pressure, the high pressure testing requirement can be satisfied and the phenomenon that the tested pressure reaches the limit of a physical network card becomes possible.

The invention provides a write data monitoring method and system based on EPOLL. The method comprises the following steps: the Epoll server-side accepting the connection of a client; after the connection is successful, a socket data dictionary saving a socket object; creating an independent monitoring thread to monitor each socket object in the socket data dictionary at fixed time; checking the size of unsent cache area in the socket object at limited time intervals, if the size of the unsent cache exceeds a threshold value, processing the corresponding client. The invention further provides a write data monitoring system based on EPOLL. The system is used for monitoring the server source and immediately adjusting the server source.

The invention discloses a public network information high-efficiency transmission method which is used for achieving interregional communication among a front and terminals. The front is arranged on the safety region in an electric power integrated data network. The terminals are arranged in a public network region. When terminals need to send data to the front, a trigger mechanism based on the Epoll model informs the front. The front performs processing only related to terminals with data sending demands. Compared with the prior art, the public network information high-efficiency transmission method solves the transmission efficiency decrease problem when a large number of terminals are accessed. The method guarantees complete transmitted messages, high efficiency and stable performance when a large number of terminals are accessed.

The invention relates to the technical field of application servers, and particularly relates to a method for realizing a high-performance IP-SM-GW system. A monitoring interface and a processing business thread pool are established in the IP-SM-GW system, events in an epoll queue of the system are monitored, if a file describer event state in an epoll waiting queue changes, a kernel directly returns these events in an events array and calls an idle business thread to process the events, so sharp efficiency drop will not be caused by the increase of the number of the monitored events. By adoption of the method provided by the invention, the problem of low efficiency caused by the frequent calling of the thread for processing the events in the prior art is avoided.

The invention relates to a method and a device for reading and writing serial data, a computer device, and a storage medium. The method comprises: configuring a direct memory access resource; traversing direct memory access node names and comparing the direct memory access node names with preset node names; when the direct memory access node names are consistent with the preset node names, analyzing the direct memory access resource and configuring a direct memory access transmission mode; establishing a double-ring buffer region memory; establishing a device file descriptor and monitoring thedevice file descriptor; when dynamic conditions are monitored, performing writing-reading operation on a double-ring buffer region. Through a direct memory access transmission module, data rapid transmission is realized. Through establishing the double-ring buffer region to store and manage data, a data buffer storage problem is effectively solved. Through an epoll mechanism to monitor data of serial equipment, system resource cost is saved, data real-time transmitting-receiving efficiency is improved, so as to further solve problems of little overflow, loss, and usurpation in serial data transmission.

The invention provides a multi-server master and slave machine work method based on a TCP event. By using the method, in a system which has a high requirement to performance efficiency and stability, two or more than two servers are arranged, configuration of each server is the same, incremental data is synchronized and the servers can replace each other to carry out work in the system. At one moment, only one server is served as a master machine (which is the aftermentioned master), other servers are served as slave machines (which are the aftermentioned slave), and incremental data of the slave machines and the master machine is synchronized. When the master machine can not work normally, the system selects the slave machine with the highest priority level as the new master machine. In the invention, an I / O multiplexing model epoll is used to manage the TCP event, efficiency is high and the method is easy to realize, and an incremental data synchronization event can carry out incremental data synchronization between the master machine and the slave machines so that a requirement of mutual replacement among the servers is satisfied.

The invention discloses a concurrent and asynchronous task processing method and equipment. Through a multi-thread Reactor and multi-process Worker architecture, an epoll based reactor can process anunlimited number of connection requests, so that the concurrent processing capability of the server side in response to client requests can be improved; two processes are created at the same time, wherein one process is a common worker process, the other is a task worker process, the task process is used for processing a request with short time consumption, the task worker process is used for processing a request with long time consumption, all asynchronous tasks can be averagely distributed to the task process to be executed, and the task response processing speed of the server side is further improved.

The invention belongs to the network communication field and relates to a system for supporting multiple platform network communication processing and a method thereof. The system comprises a network database module which is connected with an upper application system. The network database module is provided with the following modules: a channel create module, which is used to request the network database module to create a new channel and execute creating the new channel; a channel closing module, which is used to request the network database module to close the channel and execute closing the channel; a set user data and callback function module, which is used to arrange and store the user data and a callback function pointer and execute calling the callback function; a network event processing module, which is provided with a thread pool and an epoll module and is used to process an epollin event, wherein the thread pool and the epoll module are connected with each other. The channel create module, the channel closing module and the set user data and callback function module are respectively connected with the network event processing module. The set user data and callback function module is connected with the upper application system. By using the system and the method, TCP and UDP can be integrated and portability is strong.

The invention provides a wifi networking multi-hotspot control communication method, comprising the following steps of designing and making an AP-AC discovery protocol; establishing a TCP and binding the TCP with a port; establishing a UDP and binding the UDP with the port; establishing an epoll monitoring triggering event, and performing corresponding command operation; establishing a two-dimensional form, mobile phone PING and DATA protocol related data, wherein the data comprises all current information of APs and a PC; displaying current device state information in real time by a response foreground; and modifying a configuration issuing and reboot upgrading function in real time. The method has the beneficial effects that situations that a network is delayed, or even completely crashed, the machine bearing number of routers is limited, and channels are scrambled among the routers are improved; uniform management is carried out by an AC; and the APs are plug and play and are balanced in load. The method has the advantages that the network is fast, reliable and stable, and the power and channels are automatically adjusted and are deployed flexibly.

The invention discloses a video acquisition method and device based on Linux multi-core environment, and the method comprises the following steps: transmitting the instruction for acquiring video data to corresponding cameras; starting up a corresponding thread to receive the video data transmitted from the cameras; and saving the received video data in a data storage base. Prior to the step of receiving the video data from the camera, the method comprises: initiating an epoll network model with which a corresponding thread responds to process data when readable events are monitored. The invention also discloses a video acquisition device based on Linux multi-core environment. With one thread corresponding to a plurality of cameras to acquire the video data of a plurality of cameras and to cooperate and use the epoll network model, the video acquisition efficiency of cameras is increased.

The invention discloses a network processing request system, and relates to the field of Web services. The system comprises a session index unvarnished transmission module, a message metadata item order-preserving module and an active event notification module. The session index unvarnished transmission module and the message metadata item order-preserving module are executed in parallel; the session index unvarnished transmission module takes the session index as a new file descriptor used by the NGINX, and the NGINX takes the original file descriptor of the NGINX as the new session index tofind and access the session; the message metadata item order preserving module comprises a token number and an enqueue number, and checks the consistency of the token number and the enqueue number when the message metadata item is enqueued; the active event notification module comprises a VPP working thread which actively awakens an NGINX working process and an NGINX main process which actively awakens the NGINX working process, and actively inquires a mark in a user mode shared memory to confirm whether an AF_UNIX domain socket epoll event arrives or not. According to the method, the networkrequest processing performance is improved, and when no network request exists or the request amount is small, CPU resources are saved, and energy consumption is reduced.

The invention provides a data processing method based on a distributed file system. The method comprises the steps of performing connection with at least one initiator; starting an epoll network model, and when detecting transmitted network data, processing the data in response to corresponding threads, wherein multiple threads correspond to one initiator; starting the corresponding threads to receive a request message sent by the initiator and packaging the request message into SCSI command words; processing the received request message and storing metadata corresponding to the processed request message into a data memory pool. The invention also provides a data processing device based on a distributed file system. The method and the device combine multiple threads with iSCSIs, so that each thread is corresponding to one or more initators and the processing properties of a distributed file system block device is improved.

The invention provides a JCTP implementation system and method based on an EPOLL model and a medium, and the JCTP implementation system comprises any one or more of the following modules: an SOCKET communication connection pool module which performs connection pool initialization, obtains a connection from a connection pool, and releases the connection after the connection is used up; a stream coding and decoding processor module which is used for coding and decoding the data according to the private protocol; a data protocol constructor module which constructs a data protocol and transmits the constructed data protocol to a corresponding processor; and an encryption and decryption security processor module which is used for carrying out security authentication on the communication request sent to the server by the client and carrying out encryption or decryption processing according to the data flow direction. According to the invention, all hardware and operating system platforms are supported, so that a user has a greater degree of freedom in selection of an operating system and the hardware, the technical specification is more comprehensive, the support for enterprise-level applications is stronger, and the method has the advantages of once writing and everywhere running.

The invention discloses a task processing method and device. A task processing method is characterized in that a producer thread creates a thread pool comprising at least one consumer thread in advance, and the method comprises the steps that the producer thread monitors whether a new connection request is generated or not, and a new task is created based on the Epoll technology under the condition that the new connection request is monitored; The producer thread determines any consumer thread conforming to a preset task distribution rule from the thread pool according to the preset task distribution rule, and distributes a newly built task to the consumer thread based on an Epoll technology; The consumer thread determines whether a task distributed by the producer thread is received or not based on the Epoll technology, and executes the received task after receiving the task; And under the condition that the consumer thread finishes the received task, determining whether the connection corresponding to the task is closed, if so, ending the task, and if not, keeping holding the task.

The invention discloses a platform for automatically realizing video format conversion processing, which is characterized in that a set of high-performance streaming media service framework is constructed, and multiple streaming media protocols of rtmp / rtsp / hls / http-flv are supported. Linux, macos, windows, ios and an android platform are supported, and the supported coding formats comprise H264, AAC and H265; an adopted model is multi-thread IO multiplexing non-blocking programming (epoll is adopted under linux, and select is adopted by other platforms); the framework is developed based on C++11, use of a bare pointer is avoided, memory copying is reduced, codes are simple and reliable, concurrency performance is excellent, and under a linux platform, the advantages of a multi-core CPU can be fully utilized by a single process; the CPU and network card performance can be maximized; and the performance limit of the 10-gigabit network card can be easily and meanwhile, extremely low time delay can be achieved while high performance is achieved, and a picture is opened in seconds.

The invention discloses a multithreading high-performance http agency realizing method and system. According to the scheme, an epoll event notification mechanism and the multithreading technology are combined to be used, through reasonable modeling and design, the concurrent processing capacity of http agency is greatly improved, and the problem that an existing http agency realizing scheme can not effectively process especially-large concurrent requests is effectively solved.

A system and method for controlling a request response rate within an application. The systemcomprises:a convergent, EPOLL Manager, Request Statistics device, Sending queue, receiving queue, Queue load statistics, service module connecting pool and multiple service modules.When the user accesses the background system through the access gateway system, if the request data sent by the client is toomuch or too fast, the access layer will not be able to process the request data, and the data will be piled up in the connection management module of the access layer, resulting in data loss and eventhe process crash of the access layer. At the same time, the processing capacity of each service module is different. One service request is sent too fast at the access layer, and other service requests should not be affected. The invention aims at designing a method and a device for controlling the request response rate among the sub-modules in an application program, avoiding mutual influence between different service requests, avoiding dropping data packets because the application program cannot be processed, and avoiding collapse of the application program because the request rate is too fast.