147results about How to "Reduce the number of reads and writes" patented technology

The invention discloses a Key-Value local storage method and a Key-Value local storage system based on a solid state disk (SSD). The method comprises the following steps of: 1, performing read-write separation operation of a memory on data by adopting a memory snapshot B+ tree index structure; 2, performing first in first out (FIFO) queue management and caching on the indexed data aiming at a B+ tree; and 3, performing read-write operation on the data. Mapping management of logical page number and physical position is realized through an empty file mechanism in the log type additional write-in data.

The invention relates to a WEB-based database system, which comprises a database server with disk mode read-write and a database server with distributed memory, wherein the database server with distributed memory carries out data synchronization by a data synchronization module and data read and written in the database server; the database server with distributed memory also comprises a Manager server of a memory database and more than one Agent server of the memory database, the Manager server of the memory database finishes united dispatching of memory resource, realizes data synchronization between the database sever and the database with distributed memory, and provides a data visit interface for the web application server, and the Agent server of the memory database stores specific data. The WEB-based database system adopts the database with distributed memory, shares memory resource of the server, realizes the load balancing of the memory database, and improves the timeliness ofdata response and the expansibility of an application program.

The invention is applied to the technical field of storage, and provides a method and system for data caching of a solid state disk. The method for the data caching of the solid state disk comprises the following steps: setting a part of memory of a system memory as a data buffer cache of the solid state disk; caching read and write data of the solid state disk into the data buffer cache of the solid state disk. Therefore, the method and system of the data caching of the solid state disk reduces the visit times of the solid state disk and lengthens the service life of the solid state disk.

The invention provides a network television direct broadcasting system. The system comprises a streaming media signal acquisition server, a streaming media signal receiving server and a streaming media content transmitting server, wherein the streaming media signal acquisition server is used for acquiring the original streaming media signal from a program signal source in real time and generating a streaming media direct broadcasting file by coding; the streaming media signal receiving server is used for acquiring the streaming media direct broadcasting file from the program signal source in real time and buffering streaming segments; the buffering of the streaming segments is to establish an index for each key frame of the streaming media direct broadcasting file; and the streaming media content transmitting server is used for acquiring the streaming media direct broadcasting file from the streaming media signal receiving server and transmitting the streaming media direct broadcasting file to a network player for playing in real time. The system has the functions of direct broadcasting, playback, multi-screen signal acquisition, code optimization, multi-rate coding, automatic cutting, multi-packet transmission and the like.

The invention discloses a method and a device for acceleratively starting an operating system and a method, a device and a terminal for generating prefetched information, and belongs to the field of operating systems. The method for acceleratively starting the operating system includes acquiring at least one piece of prefetched information, which contains file paths of data blocks required by a process, deviation values and length values, of the to-be-accelerated process in a starting procedure of the operating system; and setting positions which are indicated by the file paths and the deviation values as starting positions according to the prefetched information, reading the data blocks with the reading lengths equal to the length values into a cache of the system, and completing the starting procedure of the process by the aid of the data blocks in the cache of the system. By means of prefetching the data blocks required by the integral or partial process in the starting procedure, frequency of page faults in the starting procedure and read-write frequency for a low-speed memory are greatly reduced, starting time of each process corresponding to a program can be effectively shortened, and the effect of remarkably increasing the starting speed of the operating system is achieved.

The method includes procedures: (1) establishing hash table on software layer; the hash table records indexes for list items of each route; (2) looking up the hash table to obtain needed index of list item; (3) based on having looked indexes of list items of routes to search route table of hardware; based on searched result to carry out deletion or insertion. Maintaining a hash table of recording indexes for list items of each route through software, when looking up, inserting, or deleting list items of route, the invention first obtains indexes for list items of the route in corresponding TCAM table through looking up hash table in software; and after having looked the corresponding list item, the invention prevents reading list item of hardware route item by item in each time. Through reducing time to read / write route item of hardware, the invention raises performance of system, and lowers cost.

The present invention discloses a management method of magnetic disk array buffer memory for reading / writing request. The management process for reading request includes the following steps: when the reading request of main machine is reached, judging that the reading request data is stored in the magnetic disk array buffer memory or not, if said reading request data is stored in said magnetic disk array buffer memory, obtaining reading request data from said magnetic disk array buffer memory and returning said reading request data to main machine, otherwise, obtaining reading request data from magnetic disk, returning it to main machine and executing step b 1; b 1 includes the following contents; judging that in the magnetic disk array buffer memory a free field is existed or not, if said free field is existed, storing reading request data in said free field, otherwise, executing step C 1; step C 1 includes the following contents; selecting reading data block in magnetic disk array buffer memory, storing the reading request data in position place of selected reading data block. For writing request of main machine said invention adopts similar method to make management.

The invention discloses a method and device for processing geographic information vector data. The method includes the steps of converting the vector data in an external format into vector image layer data, wherein the vector image layer data comprise image layer basic information, image layer metadata information and geographic target data, and the geographic target data comprise geographic target attribute information, geographic target index information and geographic target vector data; using the vector image layer data as a storage unit, and computing an outer covering frame range of the storage unit; storing the image layer basic information, the image layer metadata information, the geographic target attribute information, the geographic target index information and the outer covering frame range of the storage unit into a relation database; storing the geographic target data in the storage unit into a distribution key value database. According to the method and device for processing the geographic information vector data, the geographic information vector data storage efficiency and the geographic information vector data inquiring efficiency can be effectively improved.

The invention provides an acquiring and coding service system of a streaming media, comprising a signal acquiring module, a coder and a communication service module, wherein the signal acquiring module is used for acquiring an original audio-video stream; the coder is used for coding the original audio-video stream into audio-video data, providing the audio-video data to the communication service module and adjusting the probability of a key frame in the audio-video data in real time according to a network wideband in the coding process; and the communication service module is used for splitting the audio-video data into a network transmission packet and transmitting the audio-video data to a request end. In addition, the signal acquiring module can acquire various paths of original audio-video streams from each video source in real time; and the coder can be used for starting various paths of codes from each video source in parallel, wherein each path corresponds to a code rate. The system can ensure the smooth playing of image frames and provide multiple code rates for users to selectively switch.

The invention discloses an LDPC (low density parity code) layering BP (belief propagation) decoding algorithm based on a least-mean-square-error criterion. In the algorithm provided by the invention, the complexity of algorithm implementation is reduced through constructing a linear equation to approach to the operational formula at a check node in the BP algorithm, and the decoding performance of the algorithm is improved by utilizing the least-mean-square-error criterion to solve the optimum solution of a linear equation coefficient. Simultaneously a layering decoding idea is introduced into the algorithm, the convergence of the algorithm can be improved, and the requirement on the memory space is reduced. Based on the algorithm, the invention provides a LDPC decoder structure provided with a partial parallel structure and used for a multimedia wireless sensing network. The decoder supports the decoding of the LDPCs with various code lengths and code rates, and has the advantages ofless system resource occupation, high data throughput, low system power consumption, and the like.

The embodiment of the invention discloses GNSD (Global Navigation Satellite System) solid-state drive equipment and a data read-write method based on the same, and the solid-state drive equipment comprises a flash memory for storing first data, an SSD DRAM used for storing the mapping table and the second data; a DRAM translation layer (DTL) used for controlling access to the SSD DRAM; a controller used for determining that the read data is stored in the SSD DRAM and / or the flash memory according to the mapping table when the host reads the data and also used for storing the written data intothe SSD DRAM or the flash memory according to the number of the available DRAM entries in the mapping table when the host writes the data. By adopting the method and the device, the writing and erasing times of the flash memory can be effectively controlled, so that the durability of the flash memory is improved.

The invention relates to the field of data storage and provides a distributed file system and a method for storing lots of small files to solve the problem of relatively low efficiency for storing lots of small files at present. The system comprises a client, a resource location module and a data storage module. According to the technical scheme provided by the invention, the small files are combined into a large file to be stored and the small files are accessed by positioning a specific position of the large file, so that the file read-write frequency is reduced, the IO access process is simplified, and the processing performance of a file system is improved.

Disclosed in the embodiments of the present invention are a data reading and writing method, a memory controller and a computer, relating to the field of storage, reducing flash chip loss and efficiently improving storage performance of a system. The specific scheme is as follows: a first write command sent by a software layer controller is obtained through a first reading and writing interface, wherein the first write command comprises a characteristic value of first data and the first data to be written; according to the characteristic value of the first data, the first data is cached into a first storage unit using a predetermined write policy, and a reference count value of the first data in the first storage unit is updated, wherein the reference count value of the first data records the read and write times for the first data in the first storage unit; when a writing indication for the first data is detected, all the first data cached in the first storage unit is written to a second storage unit. The present invention is used in the process of data storage.

The invention discloses a coding and decoding circuit for a system of ultra-high frequency radio frequency identification (UHF RFID). Every time when error codes are found, the coding and decoding circuit based on the prior cyclic redundancy check (CRC) circuit resends, thus frequency of communication between a reader-writer and a tag can be increased in the situation of long distances of reading and writing and serious noise interference. If the speed of reading and writing is definite, time for successfully reading once is certainly extended, thereby influencing practicability and reliability of the UHF RFID system in the situation of reading and writing at high speed. To solve the problem, the invention is provided with a convolution error correcting code, adopts concatenation connection with the CRC circuit to reduce frequency of reading and writing, so as to improve the successful reading rate of the system in certain time. The CRC circuit of the invention is a parallel circuit with compact conformation and fast arithmetic designed based on FPGA, thereby further reducing time of communication between the reader and the tag, and increasing the amount of tag identification in unit time.

The invention provides a digital-analog hybrid storage and calculation integrated chip and a calculation device. The digital-analog hybrid storage and calculation integrated chip comprises a pluralityof flash memory processing arrays and an on-chip memory connected with the plurality of flash memory processing arrays; each flash memory processing array reads the data in the on-chip memory and performs operation processing on the data to obtain an operation result; and the on-chip memory is used for storing the operation result and input data of the digital-analog hybrid storage and calculation integrated chip, so that a digital-analog hybrid storage and calculation integrated function is realized. The digital-analog hybrid storage and calculation integrated chip provided by the inventioncan effectively reduce the design complexity and the manufacturing cost, is beneficial to large-scale industrialization, improves the design flexibility, prolongs the service life of a flash memory, and can be applied to occasions such as neural network operation.

The invention discloses a tax disk which comprises a tax disk hardware system and a tax disk software system, wherein the tax disk hardware system comprises a main control chip, a clock module, a power supply module, a storage, a register and a USB (Universal Serial Bus) interface protector which are connected with the main control chip, wherein the power supply module is also connected with the clock module; the tax disk software system comprises a hardware interface layer, a file system management layer, a safety management layer, a tax application layer and a communication interface layer which are connected with the tax disk hardware system; and a three-level index technology is adopted in the file system management system to index invoices. Compared with the traditional tax control device, one tax disk can complete the services of invoice purchasing, invoice making, invoice cancellation, tax copying and the like without other card readers or storage card devices, and can simultaneously support national tax and regional tax, wherein each tax application supports dozens of invoice types.

The invention discloses a method for managing solid-state hard disk mapping tables. The method is characterized by comprising the following steps of designing a functional structure of a solid-state hard disk, dividing translation pages for storing the mapping tables into in-order storage and dynamic storage, establishing an index for dynamic translation pages, and designing a buffer area writeback strategy, the search flow of mapping records and combination of the dynamic translation pages and the like. A second chance is provided for dirty data in caches to stay in a buffer area, and the writeback operations are reduced. When the dynamic translation pages are hit, the mapping records of one translation page is read every time and all loaded to the buffer area, and the cache hit rate is increased. Compared with traditional mapping table management methods, the method has the advantages that the frequency of reading and writing of the solid-state disk is reduced, the service life of the solid-state disk can be prolonged at the same time, the hit rate of the buffer area is increased, and the performance of the solid-state disk is greatly improved.

The invention discloses a magnetic disc data protection method, which comprises: receiving a magnetic disc operation request, and counting the degree of access heat of a requested vector; when the degree of the access heat of the requested vector meets a hot point requirement, copy is unavailable and the free space of the magnetic disc is greater than a preset free space threshold, generating a copy of the requested vector; and when the data in the requested vector and the data in the copy of the requested vector are consistent during reading and writing operation, selecting a vector which is at the shortest distance to the current position of a magnetic head for operation, or when the data in the requested vector and the data of the copy of the requested vector are inconsistent, selecting the vector with the latest data from the two vectors for operation. When the hot vector or the copy of the hot vector has a fault, data in a mirror sector are used for restoring the failed vector. In the invention, the fault probability of the hot vector is reduced by reducing the reading and writing times of the vector and the reliability of the data in the hot vector is guaranteed; meanwhile, while the system request is responded to, the data at the shortest distance to the magnetic head are fed back, so higher reading and writing performance are provided.

The invention discloses a method for processing data in a linux system. The method comprises the following steps that S1, a data processing progress is started when the linux system is started; S2, configuration files are loaded in internal storage; S3, the data processing request of an application program is received; S4, data in the current file are processed according to a hash algorithm, and a processing result is returned to the application program. The invention further discloses a system for processing the data in the linux system. The data can be processed fast due to the fact that configuration data in a storage device (such as a flash) are read when the system is started, and in addition, data reading is carried out in a hash table in the internal storage under a general condition, values corresponding to corresponding parameter names are searched for faster with the assistance of the hash algorithm, and meanwhile due to the fact that the number of times of flash reading and writing is reduced, the data are not prone to being lost.

The invention discloses an index combining method and an index combining device. The index combining method comprises the following steps: determining whether not less than n sub-indexes of which the sizes meet preset combining conditions based on a preset reference exist in an index base or not, wherein n is an integer and is greater than or equal to 2, and the preset reference is related to thesizes of the preset sub-indexes; if the sub-indexes exist, taking the determined sub-indexes as a sub-index set; if the sub-indexes in the sub-index set are combined, computing the sizes of new sub-indexes to be acquired; searching not less than m sub-indexes of which the sizes meet the preset conditions based on the reference related to the sizes of the new sub-indexes in residual sub-indexes, wherein m is the integer and is greater than or equal to 1; if the sub-indexes are searched, adding the searched sub-indexes into the sub-index set, and executing the computing step and the searching step to the new sub-index set; and if the sub-indexes are not searched, combining the sub-indexes in the sub-index set. The method and / or the device of the invention can be used for obtaining relatively stable optimized system performance under dynamic text set environment.

The invention provides an original and accompaniment switching method and device. The method comprises the following steps: after a first switching command input by a user through a touch screen in a KTV (Karaoke Television) is received, audio equipment is controlled to only play a first audio track corresponding to the first switching command, and the audio equipment is controlled to only play a second audio track corresponding to a second switching command after the second switching command input by the user through the touch screen in the KTV is received to switch an original and an accompaniment. Since a KTV set top box only obtains one music file, i.e. a video file, corresponding to a current playing task, the original and the accompaniment can be switched through switching of different audio tracks, so that the original and accompaniment switching method only needs to read and write a hard disk where a song library is stored for one time, hard disk reading and writing frequencies are reduced, and the service life of the hard disk is prolonged. In addition, since one song file is read from the hard disk where the song library is stored, the occupancy rate of transmission bandwidth is lowered so as to lighten the intranet transmission burden of the KTV.

The invention discloses a data management method, an acquisition station and an equipment monitoring system. The data management method comprises the following steps: receiving one or more data records acquired through monitoring equipment by at least one acquisition device; determining whether the length of each data record is less than a first threshold according to the length type of the data record; when the length of the data record is less than the first threshold, storing the data record in a first storage unit; when the length of the data record is greater than or equal to the first threshold, storing the data record in a second storage unit; detecting whether the length of the data, stored in the first storage unit, of each acquisition device is greater than a second threshold; and when the length of the stored data is greater than the second threshold, transferring the stored data into the second storage unit, wherein the second threshold is less than the first threshold.

The present invention relates to the field of communications. Disclosed are a cache management method and device for solving the problem of how to improve cache resource utilization. The method comprises: receiving an access request and determining data to be accessed by the access request (301); judging the strength level of the spatial locality of the data to be accessed (302); and according to the strength level of the spatial locality of the data to be accessed, allocating a cache subunit corresponding to the level for the data to be accessed (303). The present invention is used for cache management in the field of communications.

The present invention provides a blocking storage system of video monitoring data. The system is characterized by comprising: a memory used for storing video monitoring data and provided with a continuous storage space; a data block conversion unit, used for dividing each-path data of multi-path video monitoring data into at least one data block, and communicating with the memory in a unit of data blocks; and a serial communication unit, used for storing the data blocks into the continuous storage space sequentially in a serial communication manner or reading the data blocks from the continuous storage space. According to the storage system of the video monitoring data provided by the present invention, the problems of lowering read / write efficiency of a disk and reducing service life of the disk, which are caused by keeping moving a head of the disk to perform a positioning write operation by a parallel processing mode of the disk when multi-path videos are written, are avoided.

The invention discloses a terminal, an application data processing method, a data processing device and a storage medium. The terminal comprises a memory, a non-volatile memory NVM buffer, an exchangepartition, a memory controller MMU and an I / O controller. The memory includes a plurality of memory pages, the NVM cache includes a plurality of cache pages, and the swap partition includes a plurality of page slots. The MMU is used for controlling the application data of the memory page and / or the application data of the buffer page to be transmitted between the memory and the NVM buffer throughthe memory bus. The I / O controller is configured to control application data of a page slot and / or application data of a cache page to be transmitted between the switching partition and the NVM cachethrough an I / O bus. The method provided by the embodiment of the invention improves the data exchanging speed in the memory, reduces the response time of the application program, prolongs the life ofthe exchanging partition, and reduces the page jitter caused by the application data frequently moving in the memory page and the page slot in the prior art.

The invention aims to provide a method and equipment for optimizing storage efficiency of a storage device of an embedded terminal in peer-to-peer (P2P) application. The method comprises the following steps of: receiving downloaded data of P2P downloading application by the embedded terminal; storing the downloaded data in the embedded terminal and a downloading cache corresponding to the P2P downloading application; detecting whether a first trigger condition for writing the downloaded data in the downloading cache into the storage device of the embedded terminal is met; and when the first trigger condition is met, writing the downloaded data in the downloading cache into the storage device to release the downloading cache. Compared with the prior art, by introducing the cache, the read-write frequency of the storage device of the embedded terminal is reduced, the storage efficiency of the embedded terminal in the P2P application is optimized, the damage to a hard disk is reduced, and the service life of the embedded terminal is prolonged.

The embodiment of the invention relates to the field of computer technology and discloses a disk read-write method and device and a computer readable storage medium. The disk read-write method comprises the steps that storage positions of to-be-accessed resources corresponding to a resource access request are determined through a process, wherein a corresponding read-write lock is arranged at thestorage position of each to-be-accessed resource; and read-write lock operation is performed on the read-write locks corresponding to the storage positions of the to-be-accessed resources through theprocess, corresponding permissions are acquired, and disk read-write is completed. Through the embodiment, under high-concurrency connection or disk IO operation, read-write speed of disk IO is increased, and the performance of a server is improved.

The embodiment of the invention discloses a data equalization method, device and a computer-readable storage medium. The method includes counting the reading and writing times of each solid-state harddisk in the cycle time; detecting whether the reading and writing times of each solid-state disk meet the preset conditions; migrating data from a solid state drive that meets preset conditions to amechanical hard drive. Read and write times reflect the performance of SSDs, and each SSD has its corresponding performance limit. By detecting the read and write times of SSDs, we can determine whether the performance limit of SSDs has been reached or not. The preset conditions contain the basis for evaluating the performance of a solid-state drive, When the reading and writing times of a certainsolid-state disk meet the preset conditions, the data in the solid-state disk can be migrated to a mechanical hard disk to reduce the reading and writing times of the solid-state disk and realize thereasonable control of the reading and writing times of the solid-state disk, so that the solid-state disk can reach a better performance state.

The invention discloses an IO processing method in a storage system, which comprises the following steps: merging data of overlapping IO in a chain list to form merged data; Performing a down brush operation on the merged data; After brushing the merged data, the corresponding response function is called to complete the IO processing. The method fully utilizes the existing linked list of overlapping IO, merges the data of overlapping IO in the linked list first, and then performs down-brush operation on the merged data, that is to say, merges the data of overlapping IO in advance, so that themerged data can be brushed down once and returned in batch; Further, the number of reads and writes to the back-end disk can be reduced, the delay of IO processing can be reduced, and the IOPS of thestorage device can be improved as a whole; The invention also discloses an IO processing device, a terminal and a computer-readable storage medium in a storage system, which have the beneficial effects mentioned above.

The invention discloses a data fragmentation transmission method and device and computer equipment, and the method comprises the steps: respectively distributing a unique identification code for each target fragmentation file; according to the unique identification code of each target fragmentation file, obtaining meta attribute information of each target fragmentation file; determining an original arrangement sequence of each target fragment file according to the meta-attribute information; when a fragment merging request is received, generating a merged meta-attribute information list according to the original arrangement sequence and the meta-attribute information of each target fragment file. By implementing the method and the device, the problem that the throughput capacity of the server is poor due to the fact that the data read-write frequency is high in the related technology is solved, logic combination of the fragmented files is realized by combining the unique identification codes allocated to the fragmented files and the generated combined element attribute information list, a new file does not need to be generated by splicing, the seriality during file uploading can be ensured, the operation steps are simplified, and the data throughput capacity of the server and the user experience are improved.