52results about How to "Solve the overload" patented technology

The invention relates to a computer HDFS field and discloses an improved method aimed at small files of an HDFS. Partial authority of a Namenode is delegated to a Datanode which caches element data information of part of small files and the Datanode is used for processing the writing and reading request of most small files. Burdens of the Namenode are reduced to the utmost. The new processing method is provided for solving the problem that the efficiency in processing the small files by the HDFS is low. The improved method aimed at the small files of the HDFS can effectively solve the problem that burdens of a single node, namely the Datanode are excessively large and pressure of the small files is distributed to the Datanode. Thus, large file and small file processing efficiency and performance of the integral large data processing colony are quite ideal.

The application discloses a data copying method and system based on different databases; the method and system build a shared memory area with a preset size in advance in an application server memory, and the memory area serves as a data transfer channel (or channel file) between a source database and a target database for data copying; the channel is employed to carry out data transfer between the source database and the target database in a dataflow mode, thus copying data between the source and target databases; the method and system need not to make to-be copied data as local files on the application server, so intermediate data needs not to be localized, thus solving the overloading problems of an application server disk; the method and system use the channel file to copy data, the channel file hardly occupies space, and the dataflow mode is used to transfer the to-be copied data, thus realizing high concurrency massive data copying.

The invention provides a high-concurrency scene processing method and device based on a Redis and Kafka online snap-up system; the method comprises the following steps: generating a login token basedon a user login request, writing the login token into caches of front-end Cookie and rear-end Redis, checking whether the front-end login token is matched with the rear-end login token; if not, returning an operation failure result; if yes, judging whether the user login token is valid; if not, returning an operation failure result; if yes, judging whether a commodity lock of a corresponding commodity is obtained or not; if not, returning an operation failure result; if yes, pushing request data of snap-up success to a Kafka information queue. According to the method and the device in the invention, loading pressure of a server and a database is effectively reduced through current-limiting peak-clipping processing; a large number of concurrent requests can be borne in short time; the throughput capacity of the cluster is greatly increased; through front-rear end check on the number of user access times, a cheating user can be quickly found out, so that the event is guaranteed to be carried out stably and fairly.

The embodiment of the invention provides a data reading method and device and a distributed storage system. The data reading method comprises the steps that a server where a primary copy is located receives a data reading request of a user; the server where a copy with the smallest load is located is selected according to load information of servers where all copies are located; if the server where the primary copy is located is not the server with the smallest load, the data reading request is sent to a server where a standby copy with the smallest load is located, so that data reading operation is carried out; data read by the server where the standby copy with the smallest load is located is received and shown to the user. According to the technical scheme, the server where the copy with the smallest load is located is selected for data reading operation, data reading operation is not all carried out by the server where the primary copy is located, therefore, load balance of the servers where the copies are located can be achieved, data reading efficiency is improved, and the problems of resource waste of the server where the standby copy is located and too heavy loads of the server where the primary copy is located in the prior art are solved.

The embodiment of the invention discloses a database synchronization method and device, a storage medium and a terminal. According to the invention, when the archived log in the source database is updated, an incremental log record in the archived log file is obtained, a plurality of SQL statements are generated according to each incremental log record, the plurality of SQL statements are sent toa target database, then, the target database is operated based on the SQL statement; in order to solve the problem that in the prior art, the target database is overloaded due to the fact that the target database inquires the source database in a polling mode, the incremental log in the archived log file is obtained from the source database and sent to the source database on the basis of the incremental log, so that the inquiry expenditure of the source database is reduced.

The invention provides a performance testing system and method for an MS (Media Server) clustered system. The performance testing system comprises an MS clustered system, a set top box simulating device, IPQAM (Internet Protocol Quadrature Amplitude Modulation) equipment and a playing terminal; the set top box simulating device is used for simulating a test case and analyzing relevant performances of the MS clustered system according to time information on signalling interaction between the set top box simulating device and the MS clustered system; and the playing terminal is used for receiving and playing a media stream which is selected and sent by the MS clustered system according to a request of the set top box simulating device, and obtaining performance data of the MS clustered system according to a playing result. According to the performance testing system provided by the invention, a media stream playing process and a signalling processing process can be separated, the problem of overweight load of a testing machine in the traditional testing system is solved, and the analysis and the processing to the media stream are facilitated; and the set top box simulating device in the system can be used for simulating a plurality of set top boxes, so that the number of user loads contained by the test case is greatly increased, and the testing cost is greatly saved.

The invention discloses a query method based on a distributed search engine. A master node distributes a query request to a main node in a server, so that the main node can resend the query request toslave nodes in the server where the main node is located. The problem of processing bottleneck caused by the fact that the total node distributes the query request to all the nodes in the server cluster in the prior art is solved. The load balance on each node is realized. The processing overhead on the master node can be reduced, so that the query efficiency is improved. The query response timeis shortened.

The invention discloses a resource scheduling method in a Mesh network. The Mesh network comprises a master node and a plurality of slave nodes. The method comprises the following steps that the master node acquires service demand information of the master node and the slave nodes; the master node allocates time domain resources to the plurality of slave nodes according to the service demand information; and the master node sends time domain resource allocation information of each slave node to the corresponding slave node respectively, so that each slave node determines the time domain resources which are obtained by allocation respectively and carries out frequency domain resource allocation on the respective time domain resources. The invention further discloses a corresponding Mesh network node and the resource scheduling method in the Mesh network. According to the resource scheduling method in Mesh network and the Mesh network node, a problem of overlarge network load of the master node is effectively solved, and the stability of network performance is improved.

The multi-motor oil-free dry frame head and frame tail drafting drive system for ring spinning frame includes front roller motor, front roller gear box, middle and rear roller motor, middle and rear roller gear box, and three roller trains. The front roller motor is connected via a synchronizing belt to the front roller gear box, and the two output shafts of the front roller gear box are connected via synchronizing belts to the right and left front rollers separately. The middle and rear roller motor is connected via a synchronizing belt to the middle and rear roller gear box, and the four output shafts of the middle and rear roller gear box are connected via synchronizing belts to the right and left middle rollers and right and left rear rollers separately. The present invention adopts triphase permanent magnetic sync motor for driving and open-loop control, and has simple structure, smooth driving, high precision, easy regulation in spinning process and expanded technological application range.

The invention discloses a big data security framework system based on identity authentication, and relates to the technical field of big data security. The system comprises at least one big data server, the big data server is utilized to distribute the intelligent cards to the clients; the big data server creates a public key, defines operational rules, the big data server acquires an identifier and a client password of the client; a secret key needing to be authenticated of the client is encrypted by utilizing the public secret key and the identifier, the identifier and the client password are encrypted by utilizing an operation rule, the random parameter is simultaneously encrypted, and the secret key needing to be authenticated of the client, the encrypted identifier and the client password, the encrypted random parameter and the operation rule are put into the intelligent card to be sent to the client; and the big data server and the client are used for mutual authentication through the intelligent card so as to establish a session.

The invention discloses a data gathering method with privacy protection and load balancing in smart grid. Sensitive data generated by a user is disturbed in the smart ammeter end by means of a randomized response technology, the data is compressed and transmitted by a compressed sensing technology, and a reconstruction recovery algorithm is used to gather privacy data. Compared with a common centralized data gathering privacy protection method, the method of the invention can ensure the data effectiveness and improve the privacy protection degree. In the realization end of the method, computing in the local end includes simple mathematical operation but not complex operation as encryption and decryption, and power consumption of the smart ammeter is reduced greatly.

The invention relates to a walking mechanism of a bridge detection vehicle and the bridge detection vehicle. The walking mechanism of the bridge detection vehicle comprises a frame, a first oil cylinder, a second oil cylinder and a walking part; one end of the first oil cylinder is hinged to the frame, the other end of the first oil cylinder is hinged to the walking part, one end of the second oil cylinder is hinged to the frame, the other end of the second oil cylinder is hinged to a cylinder barrel of the first oil cylinder, the second oil cylinder drives the first oil cylinder to be obliquely and upwards lifted or to be downwards vertical to the ground, and the first oil cylinder drives the walking part to be upwards retracted or downwards grounded. Accordingly, the phenomenon that the problem that the on-vehicle rotary space is influenced due to the fact that the stroke of a single oil cylinder is too long is generated for meeting the rising height requirements can be avoided, and the problems that too large loads are generated at the hinged joints, and a local structure is cracked and damaged can be solved.

The embodiment of the invention provides a cloud platform monitoring data system, and belongs to the technical field of Internet. The cloud platform monitoring data system comprises a monitoring database used for storing monitoring data of a cloud platform; a monitoring cluster server comprising a plurality of monitoring servers and used for receiving the operation request for the monitoring database and executing corresponding operation on the monitoring database according to the operation request; a load balancing component used for receiving operation requests for the monitoring database from each virtual machine and each physical machine of the cloud platform, and sending the operation requests to one monitoring server in the monitoring cluster servers according to a preset load balancing strategy; and a sub-library component used for performing sub-library processing on the monitoring database and distributing part of data in the monitoring database to one sub-library.

The invention provides a safe scheduling method. The method is characterized in that enabled safe filtering functions defense attacks sequentially; under-attack information of each enabled safe filtering function is acquired; and an attack defending sequence of the safe filtering functions is regulated according to the under-attack information of each safe filtering function. The invention also provides a safe scheduling system. By the method and the system, the load of a safe defense system can be effectively reduced, so that the safety of a cloud computing system in a large-scale cloud computing environment is reliably ensured, and the problem of overweight load of a large-scale safe defending attack system in the cloud computing environment is solved.

The invention relates to spraying equipment, in particular to a series connection nine-degree-of-freedom spraying robot. The spraying robot comprises a work execution mechanism, a work positioning mechanism and an auxiliary supporting mechanism, wherein the work execution mechanism is connected with the work positioning mechanism in series. According to the series connection nine-degree-of-freedomspraying robot, a form in which a 9R main body structure and an auxiliary supporting device are combined is adopted in the aspect of the structure, so that the operation flexibility is met, and the problem that part of the joint load is too large is solved; a special 9-shaft series connection design is adopted in the aspect of motion space so that the robot ca enter a semi-closed space for spraying operation, and a larger flexible working space is achieved; in the aspect of driving control, the joint control is easy, and different modes of control schemes can be realized; the working positioning mechanism and the auxiliary supporting mechanism are controlled in a cooperative mode, and the work execution mechanism is independently controlled; and fixed region detection and trajectory prediction are combined on the anti-collision design.

The embodiment of the present invention discloses a data acquisition method and device, which are applied to WLAN equipment. The WLAN equipment includes an EPON module and a WiFi module, wherein the EPON module collects the user's HTTP data and forwards it to the WiFi module; After the HTTP data is filtered, the filtered HTTP data is stored, and finally the stored HTTP data is sent to the big data server, so that the big data server uses the HTTP data to analyze the user's online information. The method adopts a dual-mode design, that is, the EPON module collects the user's HTTP data, and the WiFi module processes the HTTP data and reports it to the big data server, thereby solving the problem of excessive CPU load caused by the traditional method and avoiding the phenomenon of WLAN equipment disconnection Occurs, thereby ensuring the stability of the user's network connection.

The invention belongs to the technical field of automobile bus automatic diagnosis and relates to a fault detection system and method for a bus to cyclically send fault codes of a controller, and mainly aims to cyclically send the fault codes. The system comprises a fault detection module and a fault sending module. The system is characterized in that the fault detection module is used for detecting fault states of all fault nodes of the controller; the fault sending module is used for sending fault information of all fault nodes of the controller to a serial bus according to a detection result of the fault detection module; and the duration of occupying the bus by the fault information of fault nodes with the faults is longer than the duration of occupying the bus by the fault information of fault nodes without faults. According to the method and the device, the fault codes are circularly sent, the problem that the network load is too large when multiple pieces of fault information are sent is solved, and the time of sending the fault information of the fault nodes occupying the serial bus is longer than the time of sending the fault information of the fault nodes without the faults.

In order to solve the problems of chip failure, local overload and resource fragmentation of the brain-inspired computer, the present invention proposes a task migration method for the brain-inspired computer operating system. By performing task migration between computing nodes, computing nodes and chips, In order to realize the reasonable allocation and scheduling of running tasks.

The embodiment of the invention provides an adjacent residential district management method and device. The method comprises the following steps of receiving a measurement report reported by a base station; sending the measurement report after the base station receives the measurement report reported by a mobile terminal and the condition that a target residential district in the measurement report is not configured is determined; judging whether the target residential district in the measurement report belongs to a current access server or not, wherein the current accessed server is an accessserver corresponding to the current located residential district of the mobile terminal; if the target residential district belongs to the current access server, configuring the target residential district in the measurement report by the current access server; if the target residential district does not belong to the current access server, notifying the base station of reconstructing measurementconfiguration so as to configure the target residential district in the measurement report by a main server. The problem of excessive load of the main server is solved; the beneficial effect of reducing the load of the main server is achieved.

The present invention provides a method and device for processing high-concurrency scenarios of an online snap-up system based on Redis and Kafka. The method includes: generating a login token based on a user login request, writing the login token into the front-end Cookie and the back-end Redis cache, Check whether the front-end login token matches the back-end login token, if not, return the operation failure result, if so: determine whether the user login token is valid, if not, return the operation failure result, if yes, then: determine Whether the product lock of the corresponding product has been obtained, if not, return the operation failure result, if so: push the request data of the successful snap-up to the Kafka message queue. The present invention effectively reduces the load pressure on the server and the database through current limiting and peak shaving processing, can withstand a large number of concurrent requests in a short period of time, and greatly increases the throughput of the cluster; , can quickly find out cheating users, ensuring stable and fair activities.

The embodiment of the invention discloses a method and system for realizing load balancing and switching of uplink and downlink data of a mobile node (MN). The method includes the following steps that: a bidirectional tunnel is established between each local mobility anchor (LMA) and mobile access gateways (MAG); each local mobility anchor (LMA) automatically checks a load status at preset time intervals and sends the load status of the local mobility anchor (LMA) itself to a route management entity (RMS); the route management entity (RMS) receives the load status of each local mobility anchor (LMA) and stores the load status to a load list maintained by the route management entity (RMS) itself; the route management entity (RMS) traverses the load list maintained by the route management entity (RMS) itself, and adjusts data forwarding paths of the mobile access gateways (MAG) in real time according to the load status of each local mobility anchor (LMA); when the mobile node (MN) is switched to a current mobile access gateway (MAG), the current mobile access gateway (MAG) completes standard proxy binding update (PBU) / proxy binding acknowledgement (PBA) switch, and then, forwards routing update signaling to each local mobility anchor (LMA) through the route management entity (RMS); and each local mobility anchor (LMA) adjusts corresponding routing items respectively and forwards obtained data of the mobile node (MN) to the current mobile access gateway (MAG) through a tunnel portal. The embodiment of the invention also discloses a system for realizing load balancing and switching of uplink and downlink data of a mobile node (MN).

The invention provides a safe scheduling method. The method is characterized in that enabled safe filtering functions defense attacks sequentially; under-attack information of each enabled safe filtering function is acquired; and an attack defending sequence of the safe filtering functions is regulated according to the under-attack information of each safe filtering function. The invention also provides a safe scheduling system. By the method and the system, the load of a safe defense system can be effectively reduced, so that the safety of a cloud computing system in a large-scale cloud computing environment is reliably ensured, and the problem of overweight load of a large-scale safe defending attack system in the cloud computing environment is solved.

The invention discloses an SRAM (Static Random Access Memory) unit. A first transmission transistor comprises a first source and a first gate; a first pull-down transistor comprises a second source; a second transmission transistor comprises a third source and a second gate; a second pull-down transistor comprises a fourth source electrode; a first pull-up transistor comprises a fifth source; a second pull-up transistor comprises a sixth source; in first and second interconnection line layers, one interconnection line layer comprises a word line and a power supply connection line, the other interconnection line layer comprises first and second bit lines and a grounding line, or one interconnection line layer comprises the word line and the grounding line, and the other interconnection line layer comprises the first and second bit lines and the power supply connection line; the word line is electrically connected with the first and second gates; the first bit line is electrically connected with the first source; the second bit line is electrically connected with the third source; the power supply connection line is electrically connected with the fifth and sixth sources; and the grounding line is electrically connected with the second and fourth sources. According to the scheme, the number of peripheral control circuits arranged on the first interconnection line layer is the same as the number of peripheral control circuits arranged on the second interconnection line layer; and supported lead tasks are basically balanced.

The invention discloses a network remote control system comprising a remote control base terminal and a control terminal, wherein the control terminal controls the remote control base terminal throughwireless transmission. The invention further provides an implementation method of the network remote control system for quickly and flexibly performing remote network control through the remote control system. According to the network remote control system disclosed by the invention, a user can control a mobile phone and other smart devices to move ( perform displacement or action) through a network by additionally mounting a base for the mobile phone and other smart devices. By additionally mounting the device for the mobile phone and other smart devices, the mobile phone and other smart devices can have brand new functions such as remote visit, remote display and remote commentary. Compared with the remote control devices currently on the market, the device disclosed by the invention has the advantages of low cost, high adaptability, convenient application, diverse functions or the like. Furthermore, by means of the remote visit / display / commentary functions, the manpower and the travel expenses of personnel can be greatly reduced, and the efficiency is higher.

An embodiment of the present invention provides a cloud platform monitoring data system, which belongs to the technical field of the Internet. Among them, the cloud platform monitoring data system includes: a monitoring database for storing the monitoring data of the cloud platform; a monitoring cluster server including multiple monitoring servers for receiving operation requests for the monitoring database, and performing corresponding operations on the monitoring database according to the operation requests. Operation: The load balancing component is used to receive the operation request for the monitoring database from each virtual machine and each physical machine of the cloud platform, and send the operation request to a monitoring server in the monitoring cluster server according to the preset load balancing strategy; The sub-database component is used for sub-database processing of the monitoring database, and assigns part of the data in the monitoring database to a sub-database.

The invention provides an intelligent power grid consistency privacy protection method and system based on state decomposition, and belongs to the field of information transmission and privacy protection. The method mainly comprises the processes of user-side electric meter data preprocessing and data distribution. Firstly, measured intelligent electric meter data are decomposed, one part is used for conducting data transmission with adjacent users, the other part is used for conducting internal calculation, and internal and external weights needed when information is transmitted are generated; in the data distribution process, an improved push-sum algorithm is adopted, a user carries out aggregation operation on data sent by adjacent users, information data of the user is weighted and sent to neighbors, and a final consistent value is achieved. According to the method, on the basis of ensuring that each user obtains an average consistency result, the personal privacy of each user can be protected.

The invention discloses a low-orbit satellite internet of things data service transmission method based on sectional connection, which comprises the following steps: S1, the low-orbit satellite authenticates the basic information of the terminal, then registers the identity information, and establishes a connection; S2, the LEO satellite notifies a terminal that the attachment is successful; 3, the terminal sends a data packet to the low-orbit satellite; 4, whether that labelled data packet in the terminal needs to send confirmation information is checked, and the low-orbit satellite carries out corresponding operation according to the inspection result; 5, the LEO satellite labels the terminal information of the data packet and transmits the data packet with the completed terminal information label to an earth station; S6, the earth station simulates the instruction interaction between the terminal and the base station by adopting a deception mechanism, and establishes the connectionbetween the simulation terminal and the simulation base station; S7, the earth station accesses the core network to complete the data transmission. The invention has excellent use effect and high useand popularization value.

The invention discloses a switch-based load balancing system and method. The system adopts a switch-based load balancing architecture, and determines the same next-hop address assigned to a request data packet from a user end according to a preset balancing condition, and receives the same next-hop address from the switch. The received request data packet is sent to the front-end scheduling server according to the determined same next-hop address, and then forwarded to the application server, thereby realizing load balancing, wherein the number of front-end scheduling servers and application servers is greater than 1. In the present invention, since the forwarding of the switch is pure hardware forwarding, and the ports are all wire-speed, the performance of the load balancing of the switch is better than that of LVS, and the complicated active-standby contact mechanism during failover is avoided, and the utilization of resources is improved. High rate, strong scalability, and also solves the problem of excessive load of traditional active and standby LVS architecture service clusters.