30results about How to "Avoid read and write operations" patented technology

A method for managing Virtual Machine (VM) storage space is provided. In the method, a Storage Balloon Agent (SBA) module deployed in a VM is adopted to directly acquire virtual storage free block information and deliver the acquired virtual storage free block information to a Storage Balloon Daemon (SBD) module deployed in a Virtual Machine Monitor (VMM) layer; and the SBD module releases a part or all of physical storage space corresponding to the virtual storage free block information, and marks virtual storage blocks corresponding to the released physical storage space as unavailable. A corresponding system and a physical host are further provided in the present invention. Through the method of an embodiment of the present invention, use condition of virtual storage space can be acquired in real time, and a large number of read and write operations of a storage system can be avoided.

The invention discloses an SAR (Synthetic Aperture Radar) imaging signal processing data transposing method based on an FPGA (Field Programmable Gata Array), which is mainly used for solving the problems of high power consumption, low speed, low efficiency and no radiation resistance existing in a real-time imaging processing CTM (Communication Terminal Module) based on a DSP (Digital Signal Processor). The method comprises the following implementation steps of: storing signal processing data in a DDRII SDRAM (Double Data Rate II Synchronous Dynamic Random Access Memory) through the FPGA; performing matrix partitioning on matrix data in the DDRII SDRAM; transposing a diagonal mode matrix block and a symmetrical non-diagonal mode matrix block in sequence; when the quantity of distance direction sampling points is unequal to that of azimuth direction sampling points, transposing the an asymmetrical non-diagonal mode matrix block; resetting after transposition; and reading out the transposed data in the DDRII SDRAM as required by using the FPGA. Due to the adoption of the transposing method, the working efficiency of the DDRII SDRAM is improved greatly, and a transposition memory CTMsystem runs more stably, has lower power consumption, higher speed and higher efficiency and is resistant to radiation; and the transposing method can be applied to the design of an SAR imaging system.

The invention relates to a high-precision riverway flood inundated area generation method for mass DEM (Digital Elevation Model) data. The method comprises the following steps of: performing stripe division on the mass DEM data; performing run length compression storage on a plurality of continuous inundation units on grid lines in stripes to reduce a data size; and finally extracting inundation ranges by virtue of a compression unit boundary tracing method, thereby generating an inundated area under complex terrain conditions. The method changes the conventional inundation source seed point filling method into grid compression storage and grid boundary tracing calculation, so that a large grid data size is processed with small memory configuration; and at the same time, a large amount of recursion judgment in the seed point filling method is avoided, and thus the computing speed is improved. The method fully utilizes digital terrain analysis and data compression storage technologies, avoids the problems generated by mass DEM data inundation analysis in the seed filling method, can be used for well performing connected domain analysis on complex terrain environments with rivers and lakes, and generates the wide-range and high-precision river and lake flood inundated area.

The invention discloses a writing and block granularity compressing and combining method and system of a key value storage system based on an OCSSD, and belongs to the field of computer storage. Through selective compression and combination of data block granularity, read-in and write-back of non-overlapping data blocks are avoided, and the problem of read-write amplification in the LSM-tree compression and combination process can be reduced; the data blocks of the same file are uniformly distributed in all channels of the OCSSD, so that the multi-channel parallel characteristic of the OCSSD can be explicitly utilized, and the throughput of a key value storage system is improved; by directly managing a bottom storage medium at a user layer and bypassing intermediate software layers such asa file system, consumption of host resources is reduced, and semantic features of application software are conveniently transmitted. Generally speaking, the problem of read-write amplification of thekey value storage system based on the LSM-tree can be effectively reduced, semantic isolation between upper-layer application software and bottom-layer storage equipment is eliminated, and the throughput of the whole key value of the system to operation is improved.

The invention discloses a data processing device used for saving storage space, which comprises a memory module, a bit-reverse module, a comparison module and a regulation module, wherein the memory module is used for saving each data required to be reordered and each reordered data; the bit-reverse module is used for obtaining an index value of each data required to be reordered, and can carrying out bit-reverse operation to each reordered data, so as to obtain a corresponding new index value; the comparison module is used for comparing the index value of each data with the corresponding new index value, so as to obtain a comparative result; and the regulation module is used for exchanging respective corresponding data of the index value and the corresponding new index value which are compared in the memory module only when the comparative result is determined to accord with a predetermined condition, and keeping the corresponding data of the index value in the memory module invariant when the comparative result is determined not to accord with the predetermined condition. By adopting the device, the storage space can be saved during the operation, and the processing efficiency of the operation is improved. Meanwhile, the invention discloses a data processing method for saving the storage space.

The invention discloses an intelligent road passing RFID (radio frequency identification device) card reader-writer. The card reader-writer comprises a wireless positioning module, a vehicle passing card reader-writer and a lane charging system. The vehicle passing card reader-writer is provided with a microprocessor, a built-in power module, a toll station information read-write module, a path identification module, a positioning and managing module and a communication module inside. Compared with the existing highway passing card reader-writer, the card reader-writer is provided with the vehicle passing card reader-writer and the lane charging system, the reader-writer can only be used in specific positions by locking read-write of the reader-writer, specific position coordinates are automatically overlapped during each read-write, updating of the existing reader-writer for management is achieved, by the intelligent card reader-writer, charging can be performed to specific coordinates at an entrance and an exit, a charging managing function is improved, passing card read-write outside a toll station can be reduced or prevented, and toll avoidance can be reduced.

A hard disk test method under an extensible firmware interface (EFI) includes the following steps: the system files of the EFI of the hard disk are backuped to the storage area in a memory; the backup area of the system files is imaged to be a real hard disk; the storage area of a non-system file is set up in the memory, which is used for storing the non-system file in file area of hard disk test; the information of the system files and the non-system file of hard disk test are obtained, and a system file list and a non-system file list are generated sequentially; the tested non-system files of hard disk test are backuped to the storage area of the non-system file when the hard disk test is executed; the system file list and/or the non-system file list are/is accessed to load the system files and/or the non-system files of hard disk test directly from the storage area in the memory. The invention can not only accelerate the seed of the test, but also can protect the system from destruction, which can realize the comprehensive test on hard disk.

The invention discloses a method and a system for user action collection based on cache memory and asynchronous processing technology. The method includes S1, recording user request data into a cache memory, S2, judging whether contents in the cache memory excess a specified size or not, writing the user request data into log files in batches if so, and carrying on recording the user request data into the cache memory if not, S3, activating a data collecting module when the system is idle, reading user visiting logs recorded by a log recording module in busy hours, analyzing user request data, matching user operation, and storing analyzed data into a database. The system comprises the log recording module and the data collecting module. The log recording module is used for recording the user request data in batches; and the user request data are cached in an internal memory and excess the specified size. The data collecting module is used for analyzing user request data in logs when the system is idle, matching the user operation, and then storing the analyzed user request data into the database.

The embodiment of the invention provides a machine learning reasoning coprocessor. The machine learning reasoning coprocessor comprises a reasoning unit and a control unit. The reasoning unit is usedfor reading and calculating a task instruction, data and parameters so as to realize reasoning operation corresponding to the task instruction; and the control unit comprises a main control unit and aplurality of channel control units. Wherein the main control unit is used for realizing global control on the machine learning reasoning coprocessor; and each channel control unit in the plurality ofchannel control units is used for respectively controlling a channel corresponding to the reasoning unit in the reasoning unit to realize response to a single task or a user according to the task instruction. According to the embodiment of the invention, multiple users and multiple tasks can be supported, and flexible customized design can be realized.

A data protection method, apparatus and system are disclosed. The system includes a host, a first storage device, a second storage device, and at least one other storage device. Wherein the first datais stored in the first storage device and the second data is stored in the second storage device, and the second data is a parity result of the first data and at least one third data. The host is configured to trigger a first instruction and a second instruction, the first instruction carrying an association identifier for indicating the second instruction. The first storage device is used for acquiring the first instruction and the fourth data, performing exclusive OR operation on the first data and the fourth data after acquiring the first instruction to obtain the fifth data, and sending adata message to the second controller, wherein the data message comprises the fifth data and the associated identifier. The second storage device is used for acquiring the second instruction and thedata message, and performing exclusive OR operation on the fifth data and the second data according to the second instruction to obtain the sixth data.

The invention provides a key value data storage method, device and equipment, and a storage medium, and the method comprises the steps: obtaining the number of first physical blocks in a first physical block set in a three-dimensional flash memory; Judging whether the number of the first physical blocks reaches a first preset number or not; if the judgment result is no, obtaining the position information of the first physical block written in the previous time; obtaining the position information of the currently written first physical block according to the position information of the first physical block written in the previous time; wherein when the data blocks in which the key value data is not written exist inside and outside the three-dimensional flash memory at the same time, the regional information of the currently written first physical block is different from the regional information of the previously written first physical block; and storing the key value data to the currently written first physical block according to the position information of the currently written first physical block, so that the currently written first physical block is far away from the previouslywritten first physical block, and the situation that the temperature is sharply increased due to frequent operation on the physical blocks in the same region is avoided.

The invention discloses a local object management method. The method comprises the steps of creating an invalid data chain table used for recording attribute information of each aggregation file and position information of invalid data on the aggregation files; judging whether a writing instruction for writing the target data is received or not; and if yes, writing the target data needing to be written into the target invalid data segment corresponding to the target invalid data according to the position information recorded in the invalid data link table so as to cover the target invalid data. Therefore, the data segments of the aggregated file can be reused in time for storage, a large number of read-write operations generated by a disk in the process of transferring valid data are avoided, and the influence on the actual read-write operations of front-end services is reduced; extra load brought to the disk is avoided while quick release of the storage space is realized. In addition,the local object management device and equipment and the storage medium provided by the invention correspond to the method.

The invention discloses a decoding method, which is based on a TDMP (Time Division Multiplexing Protocol) scheduling strategy, and comprises the following steps of: updating all variable nodes after updating a part of check nodes, then updating all variable nodes after updating the other part of check nodes, and repeating the process. According to the method, the operation of updating a part of check nodes is divided into a plurality of sub-operations, and the sub-operations and the operation of updating all variable nodes are organized into a two-stage assembly line, so that the utilization rate of a component decoder is improved; according to the method, control logic of an assembly line is constructed by combining parallelism degree requirements, so that all-zero pseudo blocks in a check matrix are skipped, and iteration time consumption can be reduced; particularly, the method performs row-column exchange on the check matrix to reduce the number of invalid operations, significantlyreduce the iteration time consumption and improve the throughput. The invention also discloses a decoder.

The invention relates to the field of household appliances, and provides a water outlet control method and device and drinking equipment. The water outlet control method comprises the steps that a label code of a water cup label is read; determining a check code based on the tag code; checking the check code based on the target check code to obtain a matching detection result of the water cup and the water drinking equipment; under the condition that the matching detection result is effective matching, target physical parameters are obtained based on reading operation on the water cup label; based on the target physical parameters, a water outlet control instruction is generated, and the water outlet control instruction is used for controlling the drinking equipment to discharge water according to the target physical parameters. According to the water outlet control method and device and the water drinking equipment, the complexity of user operation can be reduced through the reading operation on the water cup label, simple, convenient and rapid water taking is achieved, the intelligent degree is improved, the water outlet efficiency is improved, meanwhile, the water cup is verified through the verification code, read-write operation on the water cup label of an unmatched brand is avoided, and the user experience is improved. The reliability of the drinking equipment can be improved.