54results about How to "Guaranteed concurrency" patented technology

The invention discloses a large-scale container scheduling system and method based on a cloud platform. The system comprises a control node for receiving and analyzing a request sent from a client and then sending the analyzed request to the cloud platform; the cloud platform for managing the container and creating underlying resource used by the container, and providing the virtual resource of the underlying cloud platform for the container. The method comprises the following steps: acquiring a sending request of the client; forwarding the request by the container resource scheduling module; forwarding the request through an API gateway of the control node and creating a container cluster; forwarding the request through an application cluster API gateway; sending an instruction to a proxy through a scheduler; executing the instruction by the proxy; and constructing a container set through a container engine, and realizing the elastic telescopic through a transcript controller; creating a corresponding server and port; and dynamically adjusting in running. The system disclosed by the invention not only can manage the container, but also can automatically provide the virtual resource of the underlying cloud platform.

The invention discloses a realization method of a memory database on a home gateway. The method of the invention comprises the steps: 1. the storage structure and queue of data sheets and records in the memory database are initialized; 2. on the basis of the establishment of the storage structure of the memory database, the business data in a business database are uploaded to the memory database on the home gateway, while before the business data are uploaded, the business data are filtered; 3. data synchronization: when the data in business database are changed, the given data in the memory database are backed up; data synchronization is carried out by adopting a synchronous module, and if the synchronization succeeds, the backup data are cleared; 4. data access: a unified access interface of the memory database is provided for inquiring, adding, deleting and modifying, etc. Compared with the realization method of the existing memory database, the realization method of the invention requires few resources, has high data storage efficiency and is especially suitable for the business application that is based on a home gateway and has high requirement for data real time.

The invention discloses a dispatching method and a system of a distributed system. The method comprises the following steps: a dispatching request comprising data reading operation is received by a data dispatcher, and data is stored in at least one entity providing data service in the distributed system; status of each data service corresponding to the dispatching request is determined by the data dispatcher, dispatching response consenting to conduct reading operation on each data service is sent by the data dispatcher when each data service is in an idle / reading status; and each data service is at least one same data service. The invention can effectively solve the defects of resource waste, low task processing efficiency and the like in the distributed dispatching method in the prior art, realizes reasonable resource utilization, and improves task processing efficiency.

The invention discloses a digital video processing and cataloguing system and method based on a cloud edition technology. The system comprises a cloud edition client, a cloud edition service center, a cloud media processing center and a cloud storage platform, wherein the cloud edition client is communicated with the cloud edition service center through an internet; the cloud edition client comprises a cloud editor and an intelligent media stream buffer; and the cloud edition service center comprises a WMS (web media service) cloud edition media server, an ML (markup language) cloud edition material library, a cloud transmission accelerator and a CMPC (cloud media processing center). According to the digital video processing and cataloguing system and method based on the cloud edition technology, the processing and cataloguing area of digital video materials can be enlarged, and the processing range can be extended to the internet; the edition client can encrypt and buffer an acquired material stream into a memory or a local disk, so as to ensure the security of the materials; after the client edits the materials, the cloud center cluster media processing service can generate new programs, so as to ensure the concurrency between the ''cloud-client'' works; the materials need not to be downloaded to the local disk first, so that the timeliness can be enhanced; and the material stream can be collected and catalogued simultaneously.

The invention discloses a method for achieving power control and a user terminal. The method comprises the following steps: after monitoring that a first communication mode is in a communication state, that a second communication mode requires communication, and judging that an interference scene exists between the first communication mode and the second communication mode, selecting a communication mode with a low priority level from the first communication mode and the second communication mode; setting the communication mode with the low priority level as a power backoff state; if the communication quality of the communication mode with a high priority level is lower than a preset threshold, reducing an upper limit of transmitting power of the communication mode with the low priority level; and adjusting the transmitting power of the communication mode with the low priority level, to prevent the transmitting power of the communication mode with the low priority level from exceeding the corresponding upper limit of transmitting power. Therefore, the time delay problem caused by service interruption / restart can be effectively avoided, and the concurrence of the communication modes can be guaranteed to the utmost, and the overall throughput of the system can be effectively improved.

The invention discloses a real-time tracking system and method of a moving target. The real-time tracking system comprises a video monitoring subsystem, a camera subsystem, a GIS (Geographical Information System) subsystem, an image recognition subsystem and a tracking control subsystem, wherein the GIS subsystem is used for providing a peripheral monitoring point list of the current position of a tracking target; the tracking control subsystem is used for extracting a target characteristic information list and a preset characteristic matching strategy; the image recognition subsystem is used for obtaining each path of video stream in the peripheral monitoring point list and carrying out real-time calculation on the monitoring camera of the current position of the tracking target according to the target characteristic information list and the preset characteristic matching strategy; the video monitoring subsystem is used for switching the main video into the monitoring camera of the current position of the tracking target. The real-time tracking system and method disclosed by the invention have the advantages that by integration of a GIS technology, a cloud computing technology and an image recognition technology, the real-time performance and the automation for tracking the moving target are realized, so that the labor cost and the personal errors are reduced to greatest extent and the tracking efficiency is improved.

The invention relates to a high-real-time concurrent testing method based on a multi-task mechanism of VxWorks system. The method comprises the following achieving steps that test tasks are planed according to a test case, a task attribute is configured for each test task, the task attribute comprises a task name, a task period, a task priority, starting time and finishing time; the tasks and semaphores corresponding to the tasks are established; a timer is registered for multi-task dispatching, and a task information table is established. According to the high-real-time concurrent testing method based on the multi-task mechanism of the VxWorks system, calculation, logic processing and interaction with the tested system can be completed in determined time, and the multiple concurrent test tasks can be completed in the same time period, so that high-real-time concurrent test requirement of embedded software is met.

The invention discloses a mass data real-time acquisition and processing method under the supercomputing environment. Firstly, the message source of the data source side generated by the supercomputing cluster data is collected through the source side of the flume software, the collected message source is then gathered into Kafka software through the flume software, and the message sources is stored after being buffered by the Kafka software, the message source to be processed is extracted from the Kafka software through Spark software for data processing, so that the real-time collection andprocessing of massive data under the supercomputing environment is realized. The Kafka is used as an intermediate buffer, and the reliability of the data is improved. A distributed message subscription system based on the Kafka can have multiple message producers and multiple consumers, and the high availability of the system message is guaranteed. The method combines with the docker containerization technology and the load balancing technology to realize container orchestration and management, the method can be applied to real distributed data collection, real-time processing and scalable super-large cluster environment.

The invention relates to a data cleaning and repairing method, which comprises the following steps of: for data errors generated in acquisition and data transmission generated by power distribution network terminal equipment, setting an access module to carry out data cleaning and repairing; For data errors generated by information equipment in the power distribution network information system, setting a conversion module for data cleaning and repairing; For data interaction among a plurality of business information systems of the power distribution network, data badness generated in data processing, setting a rule base cleaning module; For cleaning of historical data stored in a warehouse, setting an offline module for data cleaning and repairing, and carrying out logic checking on the whole historical data. According to the method, data errors and bad data conditions can be efficiently judged, data missed judgment and wrong judgment are prevented, the repair efficiency is ensured, and the upper-layer business application data requirement is met.

The invention discloses an Android based audio and video streaming push method and device as well as an Android based audio and video streaming playing method and device. The Android based audio and video streaming push method comprises steps as follows: a push terminal collects audio data and video data through an audio collection unit and a video collection unit respectively; the push terminal encodes and compresses the collected audio data and video data through an own encoding unit; the push terminal sends a data push request to the background; the background acquires an identification code of the push terminal, judges whether the push terminal has a push permission and returns permission information to the push terminal; if the permission information indicates 'allowed', the push terminal transmits the compressed audio data and video data to the background for storage. The audio and video streaming collected by an Android mobile terminal can be pushed or played, sharing of audio data and video data is conveniently realized, and the audio data and the video data collected by the Android mobile terminal are shared and subjected to live broadcasting, so that individual needs of users are effectively met.

The invention discloses a method and device for improving calculating ability of a virtual machine in a resource monopolizing and exclusive mode. The method comprises the steps that firstly, a task is issued through a task issuing system, and whether the virtual machine needs high-performance calculating ability is described in the task; secondly, after a task executing system receives virtual machine building information, a virtual machine building module is called, and the virtual machine is built according to front-end input data; thirdly, after the virtual machine is built, according to requirements, if an appointed CPU needs to be monopolized, a new request for calling a CPU binding module is generated, and binding of a virtual machine VCPU is completed through the CPU binding module. By means of the method and device, in some specific scenes, for example, an oracle data base operates in the virtual machine, IO and concurrence ability of the virtual machine needs to be guaranteed. According to the method and device, by changing a CPU processing mode, the processing ability of the virtual machine can be improved.

The present invention relates to a replacement method of thread sharing code cache in binary translator and a system thereof. The replacement method comprises the following steps: step 10, dividing local code space into N replaceable units with same magnitude and a never-replaceable unit; step 20, allocating local code from the present replaceable unit until the volume capacity of present replaceable unit obtains a high limit, then allocating local code from the next replaceable unit; and step 30, when the space of local code obtains the high limit of volume capacity, executing local code replacement with the replaceable unit as unit according to a first-in first-out strategy. In the steps N is a natural number. The replaceable unit is used for allocating the common local code and actualizing replacement strategy. The invention presents and realizes the replacement plan of multithread code cache. The replacement plan can be widely suitable for the binary translator on various architectures, and guarantees the concurrency of threads with less expenditure.

The invention discloses a local multi-query optimization method based on predicate statutes and cost estimation and belongs to the field of big data query optimization. The method comprises the steps that firstly, an existing optimizer of a data query system is utilized to optimize queries in a query set, and the queries are shown in the form of query trees to obtain an optimized query tree set; then, the local multi-query optimization method is combined to conduct equivalent or statute processing on same or similar sub-tasks among the queries through multiple times of iteration to generate a global multi-query plan tree; finally, the statute relation between generated global multi-query plans and the sub-tasks is combined to estimate intermediate result reuse expenditure according to a price model, whether the sub-tasks are directly executed or intermediate results are reused is judged, and the global multi-query plans are optimized. The method fully considers the balance between intermediate result utilization and query concurrence, repeated operations are decreased, and the query performance is effectively improved.

An embodiment of the invention provides a data rollback processing method and a data rollback processing device. The method comprises the following steps of determining a first rollback section distributed to a first event from various rollback sections stored in various lists of a database system when the database system receives the first event; recording using conditions of the first rollback section according to the first event; and storing the used first rollback section in the corresponding list according to using conditions of the first rollback section. The rollback sections are used for storing rollback data generated by event access data files. By using the data rollback processing method and the data rollback processing device, the various rollback sections can be reasonably managed, namely, when rollback data of one of events are stored, another event can normally acquire rollback data, the concurrence characteristic of various events is guaranteed, and the work performance of the database system is improved.

The invention discloses a share-based mobile payment management unit, terminal and service method. Existing functions and resource of a mobile payment service module are fully used. A module financial payment function includes that the communication functions of different interfaces and certain service functions requiring safety module resource are achieved through increasing of safety modules provided with various types of access interfaces. If a mobile payment service module needs to be accessed simultaneously while instruct data of the different interfaces are processed, scheduling and management are performed in a resource sharing and competing mode, ordered access of the instruct data of two communication interfaces to the mobile payment service module is also guaranteed based on that the concurrency of applications operating on different interfaces can be guaranteed, and therefore the instantaneity, the concurrency and the stability of two core functions, including a telecommunication function and a payment function, of mobile payment are guaranteed. The share-based mobile payment management unit, terminal and service method can be applied to the field of the mobile payment.

The invention discloses a dynamic scheduling method and a system for collaborative computing between a CPU and a GPU based on two-level scheduling. The method includes: forecasting the processing capacity of each node in the system, a global scheduling module dynamically distributing data to each node according to the processing capacity of each node according to the batch number of the request ofthe node scheduling module in each node, when the node scheduling module finds that the data queue used to place the processing data is empty, requesting the next batch of data to be processed from the global scheduling module, and dynamically scheduling the tasks according to the CPU and GPU processing capacity in the node. According to the heterogeneity of system resources, the invention allowsweak nodes to share fewer tasks and strong nodes to process more tasks, which can improve the overall concurrency degree of the CPU / GPU heterogeneous hybrid parallel system and reduce the task completion time.

The invention relates to a crowd-sourcing competition platform framework system and a crowd-sourcing competition platform framework method based on crowd intelligence. The method comprises the following steps: arranging a data labeling platform access module, a crowd-sourcing community access module and an interactive training space module at a server, and presenting the modules to a user in a Webpage form for interaction. According to the invention, the defect that the existing machine learning competition system does not fully utilize the swarm intelligence is solved. The data utilization efficiency and the model training efficiency of an existing machine learning competition system are improved. The advantage of group intelligence can be fully utilized. The competition holding cost ofcompetition organizers is reduced to a great extent. Meanwhile, the competition participation cost of competition participants is saved.

The invention relates to a mobile information writing and reading method and system. The mobile information writing method comprises the following steps: step a1: receiving mobile information, and parsing the mobile information into a group of complete array; step a2: separately allocating an uniquely corresponding inquiry index for all the mobile information of the array; storing registration information of the mobile information in a core information field; according to the inquiry index, respectively writing all the mobile information of the array into an unique file location, and completing a writing file of the array; and and step a3: after writing successfully, writing the inquiry index information and the core information field into a database. A data storage manner provided by the invention adapts to the mobile information system, and meets an industry application standard; the information storage not only needs to satisfy an uplink and downlink transmission request of the information, but also satisfy a persistent zero failure request that ensures the information from being lost under abnormal conditions of network disconnection, shutdown and the like.

The invention discloses a data writing method server and a computer-readable storage medium. At least one performance index of a distributed document storage database is monitored, and when the performance index reaches a preset threshold value, a Redis service is enabled, and the received data sent by a terminal is written into the Redis. As such, when that server is low on distributed document storage database capability, a user can write data through Redis. The invention bypasses the distributed document storage database with insufficient capacity, thereby ensuring the writing performance of the server, ensuring the concurrency of the server side, and ensuring that the server can process the data sent by the terminal in time, returning the corresponding data to the terminal, and ensuring the fluency of the application of the user side.

The invention relates to a host RAID management method and device based on a solid state disk array, a computer device and a storage medium. The method comprises: acquiring a host RAID management request based on the solid state disk array; creating a virtual address mapping table from a host logic address to an SSD array according to the request; obtaining an access request initiated by a host end application program through the host logic address; querying or allocating virtual address information of a corresponding SSD array in the mapping table according to the host logic address; and initiating a read-write request to the corresponding SSD according to the virtual address information of the SSD array. By introducing the virtual address mapping table from the host logic address to theSSD array, the SSD array management module can sequentially allocate logic spaces of different SSDs to store write-in data of the host, so that the random write-in performance of the host is effectively improved, the write amplification is reduced, and the service life is prolonged.

The invention discloses a hotspot account transaction system and method. A transaction receiving module receives each account transaction request; a buffer scheduling module caches the account and theaccount balance in the borrowing database and the account and the account balance in the creditor database into a distributed memory; a transaction processing module processes each account transaction request, updates the account balance corresponding to the borrower and the account balance corresponding to the creditor in the distributed memory, and records each transaction detail of the borrower account and each transaction detail of the creditor account; a buffer processing module gathers each transaction detail of the borrower account and each transaction detail of the creditor account inbatches, gathers the transaction amount with the successful transaction state of the lender and the transaction amount with the failed transaction state of the lender, and updates the transaction amount with the successful transaction state into the lender account database; and summarizes the transaction amount of which the transaction state is successful and the transaction amount of which the transaction state is failed, and updates the transaction amount of which the transaction state is successful to the credit account database.

The invention discloses a message order preserving method based on a serial number mechanism, which comprises the following steps that: S1, when asynchronous message sending operation is carried out each time, a sending end packages a current sending serial number into a packet header, and adds 1 to the sending serial number; S2, a receiving end compares the current receiving serial number with the sending serial number carried in the packet header every time the receiving end successfully receives the asynchronous message; S3, after the message requests arriving in sequence are successfully processed, 1 is added to the value of a message serial number generator at the receiving end, and whether elements matched with the current receiving serial number exist in the elements arriving in a out-of-sequence queue or not are determined; and S4, if it is determined that the asynchronous message arrives out of order once, the message buffer area which finishes receiving this time is added into an out-of-order arrival queue, and processing is performed after the subsequent message arrives and the receiving serial number is updated. Concurrency of communication paths can be effectively guaranteed, the problem of message sequences caused by a bottom-layer interconnection framework can be effectively solved, and meanwhile influences of serial number generation and matching on performanceare very low.

The invention discloses a multi-thread event processing component, which realizes unified management of events by unifying external event definition basic types, packaging high-concurrency lockless queues, unifying semaphore and IO multiplexing wake-up mechanisms, and realizes low-delay and high-concurrency access of thread communication while reducing development difficulty and cost. According tothe technical scheme, through a unified event model, underlying event management and thread notification management mechanisms are decoupled, the cost of application developers for multi-thread eventdevelopment is reduced, and the event processing capacity of applications is improved. Compared with an existing event processing method, on one hand, the consistency problem of underlying network communication and internal event communication is solved; by means of the embedded memory pool model, the lock-free queue and other mechanisms, the event processing efficiency of the application programis greatly improved, the complexity of the system is lowered, and meanwhile high concurrency and low delay are guaranteed.

The invention belongs to the technical field of intelligent energy management, and particularly relates to a system for providing remote service and intelligent management for a battery. The system comprises four subsystems, namely a service data acquisition system, a service data calculation system, a service data analysis system and a service data presentation system, the system is deployed on the cloud, is used as required, has infinite expansibility, and meets the requirement of massive terminal expansion under the framework of the Internet of Things; the system also has a function of inheritance operation management of data exchange with exchanged battery equipment, a firmware remote upgrading function and an electronic fence function, and is also provided with a security mechanism inthe aspects of data encryption, access authority management and the like. A distributed modular design based on a resful framework is adopted, a visual web page, business data service and business data processing are realized, the business management capability of system Internet of Things equipment is realized through end-to-end full-dimensional visual data, and the system has strong universality, expandability and safety.

The invention discloses a software-defined wide area network routing calculation method and system based on kubernetes scalability, and belongs to the technical field of software-defined wide area networks. The method comprises the steps that S01, when a routing computation scheduling service module monitors that a client side equipment routing network segment or a client side equipment networkingmode changes, the routing computation scheduling service module initiates a task to a routing computation module, and the scale of the routing computation module is expanded and contracted accordingto a computation load; S02, after receiving the task, the routing computation module requests a client side equipment routing network segment from a client side equipment network segment cache servicemodule, then performs routing computation according to a client side equipment networking mode, and returns a routing computation result to a routing computation scheduling service module; and S03, the routing computation scheduling service module performs difference computation on the current routing computation result and the previous routing computation result, and distributes the difference result to each piece of client side equipment. According to the invention, the routing computation capability and the connectivity between enterprise branches can be improved.

The invention relates to an intelligent multi-communication system and method for an electromobile alternate current charging pile in a mountainous environment and belongs to the technical field of new energy and communication. The intelligent multi-communication method is combined with various communication modes including 230 MHz wireless private network communication, PLC communication and GPRS / CDMA wireless public network communication, and mainly comprises a master station layer, a communication channel layer and an alternate current charging pile communication unit layer, wherein the master station layer is in charge of data management and application; the communication channel layer performs message identification and environment judgment and selects the corresponding communication modes; the alternate current charging pile communication unit layer is in charge of data sending and command execution in the different communication modes. The intelligent multi-communication method comprises the following steps: S1: message triggering and initializing; S2: message identification: the master station layer identifies power grid transaction information and platform monitoring data messages from acquired object information; S3: communication environment judgment: a communication environment value is calculated and judged; S4: selection of the communication modes according to judgment results.

The invention relates to a WAF flow guide device, system and method. The WAF flow guide device comprises a WAF computing cluster module, configured to: receive a forwarded access request, perform analysis and judgment on the access request based on a preset WAF judgment rule, and send an analysis and judgment result; and a WAF access layer module, configured to: receive the access request of a client, forward the access request to the WAF computing cluster module, receive and judge the validity of the analysis and judgment result and execute a corresponding operation according to the judgmentresult, wherein the WAF computing cluster module and the WAF access layer module are located at different servers. By separating the WAF judgment rule that occupies a large amount of CPU resources from a WAF access layer and calculating the WAF judgment rule singly in a computing cluster mode, access and computational logic are separated to ensure the concurrency of the WAF access layer.

The invention relates to a host end mapping method and device of an SSD array, a computer device and a storage medium. The method comprises the following steps of S1, creating a mapping table; S2, acquiring a request from a host; S3, performing read / write processing on the request of the host according to the mapping table to obtain a processing result; and S4, feeding back a processing result tothe host. By creating the mapping table, the logic spaces of different SSDs can be sequentially allocated for storing the write-in data of the host, so that any write-in data of the host can be written into each SSD of the SSD array in a balanced manner, the wear balance is guaranteed, the concurrency of the write-in performance is guaranteed, and the requirements can be better met.

The invention provides a virtual resource allocation method, a virtual resource allocation device, a computer readable medium and electronic equipment. Relating to the technical field of data processing. The virtual resource allocation method comprises the following steps: acquiring configuration parameters for configuring virtual resources so as to acquire a resource average value and an allocable total number through the configuration parameters; generating the allocable total number of resource values in combination with a random number and the resource average value so as to obtain a resource queue through the allocable total number of resource values; and allocating the virtual resources to users according to the resource values in the resource queue so as to complete virtual resourceallocation. According to the system generation method, the problem that accurate control cannot be achieved during resource allocation can be solved to a certain extent, and then the controllabilityof resource allocation is improved.