31results about How to "Save query time" patented technology

The invention discloses a cloud computing-based service providing method and device. The method includes the following steps that: a service reservation request containing reservation information, which is transmitted by a user end, is received; at least one first candidate service provider corresponding to the reservation information is inquired in a cloud storage server according to the service reservation request; the first candidate service provider is transmitted to the user end; a target reservation request containing a target service provider identifier, which is transmitted by the user end, is received; the target reservation request is transmitted to a target service provider end corresponding to the target service provider identifier; and reservation result information transmitted by the target service provider end is received and transmitted to the user end. Compared with the prior art in which manual query of the information of service providers is required, and with the method and device of the invention adopted, the manual query of the information of the service providers is not required, and therefore, query procedures and query time can be saved, and service reservation efficiency can be improved.

The invention relates to a simple network management protocol SNMP agent multivariate group query method, which defines the data structure of the simple network management protocol agent management information database according to the management information database file, analyzes the simple network management protocol message, analyzes the variable body, and according to the simple network management protocol message The operation type processes each variable, sets the variable state and the working state of the variable, groups unprocessed variables in the variable body, calls the method of obtaining the value of the object in the group, and queries data according to the object information and variable index value information in the management information base , query the value of each variable in the successfully bound group, and mark the variable in the group as processed; if unsuccessful, process each variable in the group according to the simple network management protocol message operation type, set the variable state and working state, and repeat the preceding steps Until all variable working statuses are processed. The method of the invention realizes high-speed completion of multi-variable query and can greatly improve query efficiency.

The invention discloses a method for high-speed storage of high-fidelity continuous-frame queries and an image output method thereof. The method comprises the following steps of: acquiring free memory array information of a data server; recombining the free memory array information, and writing the free memory array information in a free storage unit group of the data server; changing a data structure of storage data by using the free storage unit, and writing the storage data of which the data structure is changed in a target storage unit group; recombining the target storage unit group to obtain a target storage unit array; and forwarding the storage data in the target storage unit array to a network memory. The method has the advantages that: (1) the performance of a storage system can be improved, and various kinds of data are massively stored; (2) the storage speed of the data server and the efficiency are improved; and (3) the data can be inquired and read at a high speed, and the effectiveness requirement of data burst is satisfied.

The invention relates to an FPGA-based (Field Programmable Gate Array-based) IO (Input / Output) bus device with an automatic recognition function, and belongs to the technical field of buses. The IO bus device comprises a CPU (Central Processing Unit), an FPGA, slot position plates, a parallel bus, a serial bus and a field bus, wherein the CPU and the FPGA are used for carrying out data interaction through an external bus, and the FPGA and the slot position plates are interacted through the parallel bus or the serial bus; the CPU and the slot position plates are directly interacted through the field bus; the FPGA is used for providing a plate address for the CPU and helping the CPU to send and receive data, a state machine is arranged in the FPGA, and the state machine can be used for selecting plate interfaces in a circulation mode to read plate information and data signals. The FPGA-based IO bus device has a self-recognition plate information function, so that the reliability of data interaction is enhanced. The communication among plates is more flexible, and upgrading of a device platform is facilitated. The FPGA serves as a bridge of the CPU and external data, and the interaction of plate data is carried out through the serial bus and the parallel bus.

The invention discloses a cache management method of a distributed memory columnar database, comprising: establishing a cache queue in a cache master control node; taking each physical task as a root node to cut its physical execution plan to obtain each physical task Corresponding cache calculation track; according to the cache calculation track corresponding to each physical task, a cache feature tree is built in the cache master node; when a query request arrives, the query execution engine parses the SQL statement into a physical execution plan; from the physical execution plan The root node starts to traverse each node in the physical execution plan hierarchically, and judges whether the cache calculation trajectory corresponding to each physical task matches the cache feature tree; if it matches, directly reads the cached actual data of the physical task from the cache slave node, otherwise Compute the physics task. The cache management method of the distributed memory column database provided by the present invention quickly detects whether the cache is hit through an efficient cache matching algorithm, thereby improving query efficiency.

Embodiments of the present invention provide a medical information query system and method. The system includes: a rule engine module, a decision support module and a graph database module; the rule engine module is used to receive clinical data and obtain query rules matching the clinical data; the decision support module is used to Query the medical information corresponding to the query rule acquired by the rule engine module in the graph database stored by the graph database module. The embodiment of the present invention can reduce the load of the rule engine module, so that it can quickly respond to the task requirements of the decision support module, and the efficiency is high; at the same time, the rule engine module is not coupled with the graph database module, which simplifies the development and maintenance of the system complexity, saving storage space and query time; in addition, using graph database to store medical information also makes query more convenient.