31 results about "Business cluster" patented technology

The embodiment of the invention provides a business routing method and a business network. The method comprises the following steps: receiving a business interaction message including a target business address and a target business type; and routing the business interaction message to an access business router of the target business through at least one of a sub-network, a global peer-to-peer network layer and a business cluster according to the address and the type of the target business. The method can access the business into the business network according to the business ownership so as to facilitate the business management and control. The business interaction cross the sub-networks is routed by using the global peer-to-peer network layer to ensure the self-organization ability of the entire business network. The business cluster can construct the shortcut for the interaction of similar businesses among a plurality of sub-networks, thereby reducing the routing hops for interaction of the similar business, fastening the business response and reducing burden of the global peer-to-peer network layer.

The invention discloses a business cluster processing method and a cluster system, wherein the cluster system comprises a plurality of business processors, a plurality of authentication data servers and an association database; the authentication data servers are respectively connected with each business processor so as to store authentication data for authenticating the identity legality of the triggering party of business process and / or the legality of message contents; the association database is connected with the business processors so as to store association data generated during business processing. The invention can improve the processing performance of the association database and raise the efficiency of business processing.

The invention discloses full path monitoring methods, equipment and device and a memory medium. A method comprises the steps of obtaining network stream data, and extracting network data packets fromthe network stream data; carrying out message analysis on the network data packets, and obtaining target business data; carrying out data clustering on the target business data according to preset business keywords, and obtaining business clusters corresponding to the preset business keywords; computing target business quantity and target response time of the business clusters; and displaying thetarget business quantity and the target response time. According to the methods, the equipment, the device and the memory medium, data clustering and displaying are carried out on real-time massive network stream data, refinement is carried out to any process sub-node, a business detailed topological graph is automatically constructed, and whole network monitoring is carried out on current business quantity and business state.

The invention belongs to the technical field of cloud computing, and relates to a load balancing algorithm. The invention discloses an API gateway load balancing algorithm for a small-scale micro-service architecture. The API gateway load balancing algorithm comprises the following steps: (1) judging whether a micro-service identifier in a request source is contained in an API gateway routing table or not; (2) determining a service cluster to which the micro-service identifier belongs according to the micro-service identifier; (3) retrieving a rear-end micro-service container list meeting conditions according to the business cluster information; (4) obtaining the millisecond-level last digit of the instant time of the API gateway, taking the remainder of the number of the back-end micro-service containers, taking the remainder as the index of a back-end micro-service container list, and selecting the back-end micro-service containers meeting the condition; and (5) enabling the API gateway to carry the micro-service request parameters to initiate service calling to a specific back-end micro-service container. According to the API gateway load balancing algorithm oriented to the small-scale micro-service architecture, the high availability of an API gateway system is greatly improved, and meanwhile, the concise load balancing algorithm is easy to implement and high in balance.

The embodiment of the invention discloses a centralized control system comprising a centralized control server, database equipment, a system background management terminal and a plurality of business clusters. Each business cluster includes a cluster management terminal and a plurality of integration units. The centralized control server is used for managing the business of each business cluster. The database equipment is used for providing a data storage function for the centralized control server. The system background management terminal is used for configuring a management account number for each business cluster. The cluster management terminals are used for managing the multiple integration units in the clusters. And the integration units are used for receiving instructions from the cluster management terminals and completing the operations corresponding to the instructions. Therefore, a problem of centralized control and management of schools can be solved.

The embodiment of the invention discloses a service scheduling method of a service cluster, equipment and the service cluster. The method is applied to the service cluster comprising load balancing equipment and multiple service processing equipment. The method comprises the steps that through detecting the Java virtual machine JVM running state parameter of the service processing equipment, when the JVM running state parameter reaches a set threshold, the service processing equipment sends a service pause request to the load balancing equipment, wherein the service pause request is used for allowing the load balancing equipment to suspend the sending of a service request to the service processing equipment. Through the invention, a response timeout problem when each of the service processing equipment carries out JVM old garbage collection is avoided, and the overall stability of a system is improved.

The invention relates to a micro-service configuration method, device, equipment and system and a storage medium, and relates to the technical field of computers, in particular to the fields of cloud calculation, Internet of Things and the like. The specific implementation scheme of the method is as follows: receiving configuration data from a main cluster; moreover, under the condition that it is determined that the service data in the main cluster is changed, sending a service data acquisition request to the main cluster, and receiving the service data from the main cluster; and executing a service governance task according to the configuration data and the service data. According to the scheme provided by the invention, the service cluster can acquire the service data by adopting an active acquisition mode. The method has the advantages that the load of the main cluster can be reduced, and the stability of the main cluster can be greatly improved especially in a scene with a large number of service clusters.

The embodiment of the invention provides a method and device for managing and consuming distributed messages . The management method includes the following steps: a health evaluation center checks whether each business cluster of a server is available; and if the health evaluation center checks that at least two business clusters are available, cluster priorities are marked for the available at least two business clusters. In the management method provided by the embodiment of the invention, the health evaluation center checks availability of the business clusters of the server, and marks thecluster priorities of the available business clusters, so that business clusters with the highest cluster priorities are provided for a consumption side to perform connection and consumption, therebyimproving high availability of the business clusters.

The invention discloses a method and device for solving knowledge base cold start in intelligent customer service. The method comprises the steps of: calling manual customer service log data for cleaning and preprocessing to obtain a candidate sentence set; carrying out coarse clustering on candidate sentences in the candidate sentence set by using business word information to generate coarsely classified business clusters; performing fine-grained semantic aggregation on the coarsely classified business clusters to generate finely classified business clusters; and dividing sentences in each finely classified business cluster into a standard sentence and similar sentences corresponding to the standard sentence according to the sentence occurrence frequency from most to least, thereby obtaining a final knowledge base for solving the knowledge base cold start. According to the method, an enterprise can be helped to quickly build a knowledge base of the enterprise, and the obtained knowledge base is higher in specialization degree, more accurate in problem and wider in problem coverage compared with an existing knowledge base.

The embodiment of the invention discloses a distributed access service processing method and device, and the method comprises the steps: carrying out the analysis on the load condition of each NEA ina current NEA business cluster through a service gateway, determining a business process of a first NEA to process a service request, sending a command generated by the service process of the first NEA to a target network element through a southbound gateway, analyzing the load condition of each NEA in the current NEA service cluster, determining the service process of the second NEA to process the message, and sending the data processed by the business process of the second NEA to an upper-layer service through the service gateway. According to the embodiment of the invention, message forwarding is carried out between the southbound gateway and the network element, the service gateway executes communication between the upper-layer service and the NEA service cluster and determines the proper NEA for processing according to the load condition of each NEA, so that on one hand, the manual operation complexity is reduced, on the other hand, the resource utilization rate of the server is improved, uneven busy and idle conditions are avoided, and the probability of problem occurrence under a high-load condition is reduced.

The invention discloses a universal enterprise management system which is characterized in that the system includes a cloud server and the user terminal; the user terminal includes a material management module, a sales management module, a payment and collection module, a human resource module, and a POS management module; and the modules are connected and communicate with one another. The beneficial effect is that the enterprise management system provided by the invention forms a system scale based on the enterprise data analysis and decision management application and a clear business cluster.

The invention provides an order estimation method and device, and can realize that a neural network still has good robustness in a business clustering deployment scene. For example, the constructed order pre-estimation neural network can pre-estimate the price trend of the product by analyzing the order, so that an early warning effect on the abnormal change of the order price can be achieved. The method comprises the following steps: acquiring multi-dimensional data of a historical order; acquiring a first basic matrix and a second basic matrix, wherein the first basic matrix is determined according to a first reference matrix of each device in the device cluster, the second basic matrix is determined according to a second reference matrix of each device in the device cluster, the first device belongs to the device cluster, and both the first reference matrix and the second reference matrix are matrices pre-trained by corresponding equipment; combining the first basic matrix and the second basic matrix to construct an order estimation neural network; and processing the current order through the order estimation neural network, and estimating the price trend of the order.

The invention discloses a service module capacity adjusting and optimizing method and device for a service cluster and a server. The method comprises the following steps: when a server call request for the same service module is received in a service cluster, judging whether module capacities among servers are different or not; and if the module capacities are different, adjusting the called weights of the servers to balance the capacities between the servers. The technical problem of lack of a capacity tuning party for the service module with unhealthy capacity is solved. The adjustment and optimization processing on the unhealthy service module is realized, so that the utilization rate of servers with different performances under the module can be maximized, and the capacity between theservers can be balanced.

The embodiment of the invention provides a remote multi-activity distributed message consumption method and device. The method comprises the following steps that: monitoring whether at least two business clusters of a service side are available or not; if a monitoring result shows that at least two business clusters are both available, selecting any one of the at least two business clusters as a main business cluster; and connecting with the selected main business cluster for consumption. In the embodiment of the invention, the availability of all business clusters on the server side can be monitored, a main business cluster is selected from the available business clusters, then, connection is established with the main business cluster, and consumption is carried out. In other words, in the embodiment, a consumption end selects the main business cluster from the available business cluster so as to improve the high availability of the business cluster, in addition, after the consumptionend is switched to the new business cluster, consumption is carried out according to reset consumption schedule, and the repeated consumption of messages can be reduced to a large extent.

According to a method, a reliable multicast mechanism is adopted, data is distributed from a source business cluster to a target business cluster, at any time, the target cluster only has one node (named as a main node) which is responsible for distributing processed business data to the outside again, and other nodes are named as standby nodes. All the nodes in the target business cluster join the same multicast group and process the received service data at the same time, but the standby node does not participate in external data distribution, and the processing speeds of all the nodes are consistent, so that the conversion sequence of internal state machines can be kept consistent, and the strong data consistency among the nodes is ensured. If the main node fails, the reasonable standby node can be selected to be promoted to the main node, and data processing is continued, so that distributed data consistency is achieved, namely, all data can be accessed from any node in the cluster. According to the invention, the original TCP point-to-point transmission mode is replaced by connectionless reliable multicast, so that the number of nodes accommodated by a service cluster is obviously increased.

The disclosure proposes a data transmission method, device, routing device, server, and storage medium based on an organizational structure, and relates to the technical field of the Internet. The method is applied to routing nodes in the data transmission system. The data transmission system also includes at least one business cluster, each business cluster includes multiple business nodes, and each business node included in the same business cluster corresponds to a different business organization type, wherein, At least one service cluster is connected to the routing node through a service bus, and the method includes: receiving a service request sent by the first service node, the service request carrying the identifier of the second service node, and sending the service to the second service node in the at least one service cluster Request, receiving the service data fed back by the second service node according to the service request, and sending the service data to the first service node. The present disclosure can realize data intercommunication between various service clusters and service nodes in a data transmission system, improve resource utilization, and reduce costs.

An embodiment of the present invention provides a method and device for consumption of distributed multi-active distributed messages in different places. The method includes: monitoring whether at least two service clusters of the server are available; if it is detected that the at least two service clusters are available, Then select any one of the at least two business clusters as the main business cluster; connect to the selected main business cluster for consumption. In the embodiment of the present invention, by monitoring the availability of all business clusters on the server side, a main business cluster is selected from the available business clusters, and then a connection is established with the main business cluster and consumption starts. That is to say, in this embodiment, the consumer side selects the main business cluster from the available business clusters, thereby improving the high availability of the business cluster, and after the consumer side switches to a new business cluster, according to the reset consumption progress Start to consume, greatly reducing the repeated consumption of messages.

The invention discloses an all-path monitoring method, equipment, storage medium and device, the method comprising: acquiring network flow data, extracting network data packets from the network flow data; performing message analysis on the network data packets , to obtain target business data; perform data clustering on the target business data according to preset business keywords, and obtain business clusters corresponding to the preset business keywords; calculate the target business volume of the business clusters and target response time; displaying the target traffic volume and the target response time. In the present invention, data clustering and display of real-time massive network flow data can be refined to any process sub-node, and the business detailed topology map can be automatically constructed to realize the whole network monitoring of the current business volume and business status.

The invention discloses a method and a system for managing kubernetes at a cloud edge. The method comprises the following steps of: deploying a kubernetes proxy and a set of kubernetes meta cluster; a kubernetes business cluster is deployed by using the kubernetes meta cluster; the method comprises the following steps: acquiring life cycle information through a kubernetes proxy end; according to the technical scheme, the life state information of the kubernetes service cluster is acquired in real time by the kubernetes meta cluster, and the kubernetes service cluster is driven to reach a target state based on the life cycle information, so that the cloud edge kubernetes nano-management of kube-on-kube is adopted, manual intervention is not needed, and the edge kubetes are managed through the kubernetes.

The invention provides a cluster data management method and a storage medium. The method comprises the steps of configuring a routing cluster connected with a service system; configuring interfaces, including write-in, update and read-only types, supported by each service cluster in the routing cluster; enabling the service system to determine a service cluster to be accessed according to the current configuration of each service cluster returned by the routing cluster; when the current capacity of the database corresponding to one service cluster reaches a preset threshold value, modifying the configuration of the service cluster in the routing cluster into an interface only supporting update and read-only types; when the modification exceeds a preset duration, modifying the configurationof the service cluster to only support an interface of a read-only type; and migrating all data in a database corresponding to the business cluster. According to the method, the multi-cluster independent database can be conveniently subjected to capacity expansion operation; recovery of database resources can be supported; load balancing of the service cluster can be realized; and a historical data query function can be provided.

This application discloses a system for managing distributed business clusters, including: a management subsystem and at least one node control subsystem connected to it; the management subsystem is used to manage operations on resource objects and manage node control subsystems The execution of the node control task of the system; the node control subsystem is used to execute the node control task issued by the management subsystem. The system provided by this application can quickly and efficiently implement various management tasks for distributed business clusters. The present invention further provides a method for creating a distributed service cluster, a method for adjusting a distributed service cluster, a method for submitting jobs to a distributed service cluster, and a method for expanding the service software environment of a distributed service cluster.

The present invention relates to a video image information collection method based on a distributed cluster, the steps are as follows: Step 1, the data in the collection device / system is interconnected with a private network, the data is interconnected and stored in the cluster area through the private network, and the data stored in the cluster area It is provided to the business information system through private network interconnection again; step 2, the cluster area is composed of a cluster management layer and an interface service layer, and the cluster management layer in the cluster area provides a fixed IP port to the private network; step 3, the cluster area in the cluster area The cluster management layer has a master management node and a standby management node; the master management node and the standby management node in the cluster management State interface nodes, the interface nodes use HTTP long connection mode to connect the main management node and standby management node in the cluster management layer; step 5, the interface nodes in the interface service layer are interconnected with information queue / cache and cloud storage service respectively; step 6 , the information queue / cache area is interconnected with the main business cluster, and the main business cluster is interconnected with the cloud storage service. The video and image information collection method based on distributed clusters is adopted to improve the video and picture information processing capabilities and save cost and time.

The invention discloses a traffic grooming and distinction importance protection method based on clustering, and belongs to the technical field of communication networks. The method includes the following steps that a network is divided into clusters, and services in the clusters are directly groomed through a layered auxiliary picture; the shortest paths between the clusters are looked for by means of a cluster convergence layer for the services between the clusters and then are mapped onto the layered auxiliary picture for grooming; according to the situation of multiple priority services in an independent wavelength after grooming, importance of wavelength links is distinguished; by means of an intelligent telescopic P ring, grading protection is effectively carried out. By means of the method, dynamic services are quickly and efficiently groomed, and the utilization rate of network resources is greatly improved; meanwhile, different service quality guarantees are achieved through distinction of the importance of the link wavelengths, the situation that resources are seriously wasted due to the fact that the services are protected one by one in a traditional method is avoided by means of P-Cycle comprehensive protection, arriving and leaving of the services can be perceived, the P ring is dynamically extended and retracted, current network intelligent requirements are met, protection resource cost is greatly reduced, and the network capacity of bearing the services is improved.

A client system of enterprise management software includes a material management module, a sales management module, a payment and collection module, a human resource module, and a POS management module, all of which are connected and communicate with one another. The material management module is used for coding and classifying stock materials, setting master data of virtual warehouses, and transferring and allocating materials among different warehouses. The sales management module is used for processing orders of various sales services during product selling and warehousing out a product in the system when the product is delivered to a customer. The payment and collection module is used for handling various payment and collection services generated in the operation of a company. The human resource module is used for managing the basic information of employees and recording attendance information of employees. The POS management module is used for a central warehouse to ask for replenishment and for the system to give replenishment and distribution suggestions. The enterprise management system provided by the invention forms a system scale based on the enterprise data analysis and decision management application and a clear business cluster.

The method includes steps: setting up command of ending up group calling of cluster in cluster service in advance in use for dispatch terminal to end group calling; in process of group calling, cluster service exchange control center determines whether the received information is identical to the command of ending up group calling setup in advance; if yes, based on the received information, the control center ends up group calling. The invention discloses method for ending up group calling specially for dispatch terminal to use. When receiving off line command from the dispatch terminal, the exchange control center carried out off line treatment for the dispatch terminal only, thus, other communication terminals in the cluster can carry group calling continuously.

The application relates to a method, system, computer equipment and storage medium for electing master and backup copies. The method includes: each copy in the business cluster that satisfies the election condition elects a candidate master copy from the business cluster; the candidate master copy sends an arbitration request to the management cluster; the arbitration request is used to instruct the management When the dynasty of the candidate master copy is the same as that of the candidate master copy, update the dynasty of the business cluster, and determine the candidate master copy as the master copy of the updated dynasty; after determining the master copy of the updated dynasty, remove the master Each replica other than the replica sends a dynasty join request to the master replica; when the corresponding replica meets the dynasty join condition according to the dynasty join request, the master replica determines the replica that satisfies the dynasty join condition as the backup copy of the updated dynasty. This method can ensure the reliability of the master and backup replica election without increasing the replica deployment cost.

The present invention provides a service processing method, node and system based on policy sharing. The method includes: the management node receives the data message sent by the processing node, the data message includes service type information, and the processing node communicates with the management node A processing node in one of the at least two connected service clusters; the management node determines whether there is a shared policy component corresponding to the service type information; The policy component processes the user service. Since the policy component with the same business processing function shared between at least two business clusters is set in the management node as a shared policy component, each processing node in each business cluster does not need to configure the shared policy component, which greatly reduces the need for processing The occupancy of processing resources of the node.