146results about How to "Reduce the number of writes" patented technology

The invention provides a high-performance reliable solid-state disk realizing method. The method includes: (1), dividing all flash memory chips in a solid-state disk into groups and forming an RAID (redundant array of independent disks) 4-level flash array in each group by N successive flash chips; (2), receiving and storing data through a cache; (3), judging whether the cache is filled up or not, if yes, entering the step (4), and if not, returning to the step (2); (4), extracting N-1 data blocks from the cache and computing check values of the N-1 data blocks; making up the N-1 data blocks and the check values into filled stripe data and writing back the flash array; returning to the step (2). The flash chips in the solid-state disk are used for establishing the RAID4-level physical array to assure data reliability. Faults at different levels including page level, block level, or even chip level can be processed. Besides, writing-in performance is improved by writing of filled stripes and sequence, and spatial and performance loss resulted from data errors can be reduced to the utmost.

The invention discloses a method for prolonging the service life of a solid-state disk, which comprises the following steps of: (1) adding a write request into a write request queue in a buffer area of a solid-state disk; (2) selecting a data page in the write request as a sampling page; (3) calculating the fingerprint of the sampling page and also comparing with fingerprints in a fingerprint library so as to carry out matching; (4) if no matching fingerprint is found, writing the rest data pages in the sampling page and the request into a flash memory of the solid-state disk directly; and (5) if a matching fingerprint exists, carrying out fingerprint calculation on each of the rest pages respectively and also comparing with the fingerprints in the fingerprint library respectively so as to carry out matching: updating a corresponding mapping table directly for the data page in which the matching fingerprint is found and writing the data page in which the matching fingerprint is found into the solid-state disk. According to the method for prolonging the service life of the solid-state disk, the actual physical occupation of data in the solid-state disk on the flash memory is reduced, the redundant space of a system is indirectly increased, and the frequency of the garbage recovering operation of the system is reduced, so that the service life of the solid-state disk is enhanced.

The invention discloses a caching method and a caching device. The caching method comprises the steps of determining a to-be-deleted data block needing to be deleted from an NV (Non-volatile) storage medium in a caching space according to the access situation of data blocks; selecting and deleting the to-be-deleted data block from the NV storage medium according to a preset data recovery condition. According to the caching method and the caching device, which are disclosed by the invention, the writing times of an NV storage device which is unbalanced in writing and poorer in duration can be reduced, and the service life of the NV storage device can be prolonged.

The embodiment of the invention provides a cache replacing method, a cache controller and a processor. The method comprises the following steps that: the cache controller determines an associated cache pool of a cache line to be replaced, wherein each associated cache row in the associated cache pool and the cache row to be replaced belong to the same memory row; a cache row to be written back is further determined from the associated cache pool according to the access information of the associated cache row; and data in the cache row to be replaced and the cache row to be written back are simultaneously written into a memory. The cache row to be replaced and the cache row to be written back belong to the same memory row, so that the hit rate of the cache region can be improved; and the memory access performance is improved. The cache controller further determines the cache row to be written back from the associated cache pool according to the access information of the associated cache row, and only the cache row to be written back in the associated cache pool is written into the memory, so that the number of the memory writing times can be reduced; and the service life of the memory is prolonged.

The invention relates to the technical field of electronics, in particular to a method, device and system for data storage. The method, device and system for data storage are used for solving the problems that a storage device is low in writing performance and short in service life. The data storage method provided by the embodiment includes the steps that a storage controller receives data to be stored and is internally provided with a first cache and a second cache, the first cache is used for caching the first type of data, the second cache is used for caching the second type of data, and the updating frequency of the first type of data is larger than that of the second type of data; the storage controller recognizes the first type of data in the data to be stored and writes the recognized first type of data into the first cache; if the first type of data written in comprise updated data, the data, which correspond to the updated data and are not updated yet, in the first cache are set to be invalid; when the storage controller determines that the writing-in triggering condition of the first type of data is met currently, the valid data in the first cache are written into a storage medium.

The invention discloses a shared storage message queue-based implementation method for high availability of virtual machines and relates to the field of cloud computing. The shared storage message queue-based implementation method for the high availability of the virtual machines comprises the following steps: initializing a piece of contiguous space on shared storage by a server program to serve as a logical volume; distributing a plurality of contiguous sectors from the logical volume to serve as a sector pool of a client program; acquiring the configuration information of the sector pool and the number of a virtual machine, of which the high availability is enabled, by the client program; sending a heartbeat message by the client program, and updating the timestamp of a virtual machine control block maintained by the client program; receiving the heartbeat message by the server program, and updating the timestamp of the virtual machine control block maintained by the server program by utilizing the timestamp carried in the heartbeat message. According to the shared storage message queue-based implementation method disclosed by the invention, the unnecessary stop time of the virtual machines is avoided; system resource consumption is low by adopting a heartbeat detection mechanism; the performance is high; the service life of the disk of the shared storage is prolonged.

The invention provides a block-level-data-based directional allocation method for hierarchical storage. The method includes: establishing a data directional allocation architecture comprising a data block popularity judgment module, a data block directional allocation and migration module, and a hierarchical storage and organization module. The data block popularity judgment module is used for monitoring part of fixed-size block data in a volume and statistically analyzing data popularity so as to complete hierarchical management of the block data. The data block directional allocation and migration module is used for directionally allocating the data according to data popularity state and equipment features. The hierarchical storage and organization module is used for integrating the storage resources to form a storage resource pool capable of hierarchical storage. Compared to the prior art, the block-level-data-based directional allocation method for hierarchical storage allows hierarchical storage management to be more efficient and is highly practical and easy to popularize.

The invention discloses a method for resuming transmission from a break-point for a reliable file receiving system on the basis of a unidirectional broadcast network, which comprises the following steps: firstly, slicing files to be transmitted according to the fixed size; secondly, attaching offset information and check information to each sliced file; thirdly, submitting a combined application data unit to a reliable multicast layer to split the application data unit into data blocks at the layer; fourthly, coding the data blocks so as to generate corresponding redundant data blocks; and fifthly, transmitting all the data blocks in a scrambled mode by a file transmission service terminal. When a user terminal receives the file, an information index file of the file is created for recording the receiving situation of the file data blocks. When the file receiving is resumed after break, information record files are checked to determine whether the transmission of the file needs to be resumed, and the transmission of the file is resumed according to the attached information. By adopting the method, the file can be received continuously from the break-point and the redundant and time-consuming write operation is skipped over to reduce writing times of a disk to protect the disk, thus the aims of reducing the I / O times and the CPU occupancy rate and saving hardware resource of a receiving end is achieved.

The invention discloses a disk array caching method for a virtual SSD and an SSD isomerous mirror image. The disk array caching method comprises the steps that a disk is logically divided into two parts, wherein the upper half part forms an RAID0 or RAID5 structure to serve as the virtual SSD so as to back up dirty data in a real SSD, the virtual SSD is in a log writing mode, a ram writing cache is set, and the lower half portion is assembled to be of different RAID structures according to needs of users, and is used as a data array so as to store cold data; an address mapping Hash locating table between a data disk and the cache of the real SSD is built; a timer is set so as to regularly write data cached in the ram writing cache of the virtual SSD into the virtual SSD; a request from an upper layer file system is received; for a reading request, when the reading request is not hit in the real SSD, data are read from the data array. The disk array caching method gives consideration to performance, reliability and cost.

The invention discloses an I / O (input / output) scheduling method based on internal parallelism of a solid state disk (SSD). The method comprises the steps of dividing a logical address space of the SSD into N areas in the same size, conducting serial numbering on the areas, assigning a subqueue used for managing each area for each area, setting an area counter i=1, judging whether a read request FIFO (First Input First Output) queue of the ith area contains a read request or a write request FIFO queue contains a write request, if so, judging whether the last service request in the ith area is the write request, and if so, judging whether the read request FIFO queue and a read request red-black tree contains the read request, wherein each subqueue comprises the read request FIFO queue, the write request FIFO queue, the read request red-black tree and a write request red-black tree. The method can reasonably and moderately utilize the internal parallelism of the solid state disk and avoid mutual interference between the read request and the write request, so that the I / O performance of a whole solid state disk system can be improved, and the service life of the whole solid state disk system can be prolonged.

The invention discloses a solid state disk (SSD) prefetching method for mixed caching and the SSD. The method comprises the following steps: when the SSD stores a data prefetching request, searching an idle storage block pointed by an address pointer in an idle block linked list; storing sequence flow data in the same sequence flow in the data prefetching request into an idle page of the same idle storage block; and storing the address pointer of the storage block which stores data into an occupied block linked list. The data in the same sequence flow is stored in the same block during sequence flow prefetching based on a spatial locality principle, so that the page in the same block is replaced at the same time or at the imminent time, the writing times of the SSD is reduced and the service life of the SSD is effectively prolonged.

The invention discloses a method for establishing an RAID (Redundant Arrays of Independent Disks) in an SSD (Solid State Drive). The method mainly comprises the following steps: abstracting Nand Flash Blocks into Super Blocks, abstracting Pages into Super Pages, and performing RAID establishment in the Super Pages by taking the Super Pages as RAID stripe units; making the Super Blocks equivalent to the Nand Flash Blocks by an FTL conversion layer: performing writing in sequence from small to big according to Super Page numbers during writing operation and erasing the Nand Flash Blocks in the Super Blocks at the same time during erasing operation; making the Super Pages equivalent to Nand Flash Pages by the FTL conversion layer: performing writing in sequence on the Nand Flash Pages in the Super Pages from small to big according to die numbers during writing operation; in addition, recording bad dies and RAID dies in the Super Blocks by the FTL conversion layer, and skipping dies where bad pages are when data is written for RAID establishment; and after all user data is written in RAID stripes, determining final RAID check data of the stripes and writing the RAID check data into the RAID dies of the Super Pages. According to the method, required DRAM resources are reduced, the cost is low, and the efficiency is high.

The invention provides a data updating method and device, a storage medium and equipment. The method comprises the steps: acquring at least one data updating request sent by front-end equipment; Storing the data updating requests with the same data identifiers into a queue; processing data updating requests in the queue sequentially, wherein the processing comprises the steps: reading to-be-updated data corresponding to the data updating requests are read according to data identifiers; In response to the data updating request, processing to-be-updated data; Storing the processed to-be-updateddata to a cache; Judging whether the processing time reaches a preset time period or not; And if yes, updating the to-be-updated data in the cache to the database, and pausing processing until the operation of updating the to-be-updated data to the database is completed.

The invention discloses a method and a device for writing data in a non-volatile memory. The method comprises the following steps of: backing up all data of a memory cell at an address of data to be written-in in a data region to a data buffer region in a random access memory (RAM) when determining that data is to be written in the data region of the non-volatile memory during execution of an operating instruction; backing up original data at the address of the data to be written-in in the data region to a backup buffer region in the RAM and recording address information of the original data, which is backed up, in the data region in the backup buffer region; writing the data to be written in at a corresponding position of the data buffer region according to the address of the data to be written-in in the data region; and writing the data in the data buffer region back to the data region and clearing the backup buffer region when determining that the execution of the operating instruction is finished. By the method and the device, the data is quickly written in the non-volatile memory, write-in frequency of the non-volatile memory is reduced, program running speed is increased and service life of the non-volatile memory is prolonged.

An embodiment of the invention provides a cache writing method, a caching device and a storing system. The method includes that at least one metadata in an internal storage is updated; at least one metadata set where the metadata belongs is obtained; if a writing state of a first metadata set in the metadata set is a non-writing state, and a plurality of writing input / output (IO) request commands waiting for executing metadata writing operation exist under the first metadata set, the plurality of writing IO request commands are moved to a combination writing metadata queue of the first metadata set; the metadata in the first metadata set is written into a metadata block corresponding to the first metadata set in Flash; according to the plurality of writing IO request commands under the combination writing metadata queue of the first metadata set, the state showing whether the first metadata set is written successfully is returned. By means of the cache writing method, the caching device and the storing system, writing times of metadata of the Flash can be reduced to some extent.

The embodiment of the invention provides a method and device for storing data, relating to the field of data storage and solving the problems that the service life of an SSD (Solid State Drive) becomes short and the storing efficiency of the SSD is reduced because of frequently writing data to the SSD. The scheme comprises the following steps of: respectively counting the visiting volume of any data in the device for storing data in N periods, and N is greater than 0; determining whether the data meet the features of the solid state drive SSD according to the visiting volume of the data in the N periods, wherein the features of the SSD are used for reflecting the liveness of the data in the N periods; storing the data in the SSD if the data meet at least one feature of the SSD.

The invention provides a storage system with a second-level cache structure. The storage system comprises a host internal memory and a solid state hard disk, wherein the solid state hard disk comprises a main control chip, an NAND chip and an MRAM; the host internal memory comprises a read cache; and the MRAM comprises a write cache. The invention furthermore provides a reading / writing method which adopts the storage system with the second-level cache structure. According to the second-level cache structure and the reading / writing method which adopts the second-level cache structure, the reading operability is better; the MARM is used as the write cache, so that the data writing speed is greatly improved while the security of the data under the condition of unexpected outage is ensured; by using the read cache in the host internal memory, the limited MRAM space can be used as the write cache as much as possible, so that the cost is optimum and an expensive outage protection system is saved; and by using the MRAM as the write cache, the write frequency of the NAND is decreased, so that the NAND is protected and the service life of the NAND is prolonged. According to the storage system and the reading / writing method, a scheme with good cost-benefit ratio is provided.

The invention discloses a page management method based on an embedded system mixed main memory. The embedded system mixed main memory is an embedded system PCM / DRAM mixed main memory, a CPU of an embedded system sends an access page request, an access of the main memory is performed if request data or an instruction is not in a cache, and the page management method is executed at the moment. The page management method comprises building a CLOCK linked list existing in a page of the mixed main memory and an LRU linked list stored data of which are metadata of the page of an internal memory removed from the CLOCK linked list, determining whether the page accessed by the request is stored in the mixed main memory of the embedded system, accessing the CLOCK linked list if the page accessed by the request is stored in the mixed main memory of the embedded system, determining a type of the page in the CLOCK linked list to perform change operation of page identification bit or page migration operation, entering the next step if the page accessed by the request is not stored in the mixed main memory of the embedded system, obtaining a free page as a storage space of the accessed page, accessing the LRU linked list, and calling a page insertion algorithm to insert the accessed page into the mixed main memory.

This invention relates to a large volume nonvolatile alarm data storage device including a microcontroller, a nonvolatile RAM storage, a FLASH storage, a CMOS low power loss linear regulator and an interface, in which, the FLASH storage and the nonvolatile RAM storage are combined to store the alarm data, the microcontroller controls the read and write of the two storages and receives the alarm data via the serial interface or receives instructions to output the alarm historical records advantaged that the FLASH is taken as the primary storage to store large volumes of historical data and the nonvolatile RAM storage is used to buffer alarm data without limit of writing times to reduce the writing times of the FLASH storage and increase the life time of the device.

The invention relates to data storage method and equipment. The method comprises the following steps of: receiving a write request; determining an itemization unit SU of the write request in the layout of a virtual logical unit number LUN which is generated in advance; if judging that the SU already writes data and the write times of a hard disk unit DU which belongs to the SU and has a given size exceed the preset write times, redistributing at least one new DU for the SU from a DU resource pool; and writing the data in the write request and the original data on the SU into at least one new DU which is distributed for the SU. By abrading and balancing the DUs of all hard disks in a storage system, the embodiment of the invention enhances the service life of a solid-state hard disk greatly.

The invention provides a server cluster which comprises a login server, a database server and a game server, wherein the login server is used for carrying out login authentication on users, the database server is used for storing game data, and the game server is used for carrying out synchronous control on game data and carrying out logic control on the game play, the server cluster also comprises multiple gateway servers and a message routing server, the multiple gateway servers access the game server by users, and the message routing server is used for carrying out message routing between each gateway server and the game server, the gateway servers are connected with the game server through the message routing server, and the login server is connected with the game server through the database server. Accordingly, the invention also provides a server system. According to the technical scheme of the invention, under the premise of not affecting the stability and smoothness of games, the system load is improved and the online number of people playing games is increased.

The invention discloses a merging method based on non-aligned update data in a solid-state disk cache system. Based on a FTL page level mapping module, a page level mapping module structure is redesigned, a merging page mapping module structure is added, an effect of storing the non-aligned update data in the cache system, which is from a plurality of logic blocks, into a solid-state disk after being merged is achieved, and the maximum utilization rate of a physical available storage space is ensured; meanwhile, aiming at write update read loss brought in by a non-aligned update operation, performance loss of an additional read operation request, which is brought in by a write update operation, is optimized; and in the other aspect, due to a case that by merged writing, actual writing times of the solid-state disk are sufficiently reduced, the internal available storage space is utilized to the greatest extent, so that a trigger rate of an internal garbage recovery operation is reduced, thereby indirectly prolonging the service life of the solid-state disk.

The present invention discloses an NAND flash memory-oriented page replacement method, belonging to the field of data storage. According to the method, a cold clean page linked list and another hybrid page linked list respectively manage cold clean pages and other cache pages, and a reserved ghost page linked list records metadata information of recent obsolete cache pages; before page visit processing, initialization is performed first, if the page is hit in another hybrid page linked list or the cold clean page linked list, the page is marked a hot page and is moved to a recently visited terminal of another hybrid page linked list; if the page is hit in the ghost page linked list, record of the page in the ghost page linked list is deleted, a new cache page is allocated to the page and is marked a hot page, and the page is moved to the recently visited terminal of another hybrid page linked list; and finally, if it is determined that the page is read or write request, the data is read or written in the cache page and is returned. Through adoption of the method, flash memory write operation is reduced, and a higher cache hit ratio is maintained as far as possible.

The invention provides a solid-state drive control chip integrating an MRAM (Magnetic Random Access Memory). The solid-state drive control chip comprises a CPU, the MRAM, a host interface and an NAND controller, wherein the host interface adopts a standard memory reading and writing interface. The invention furthermore provides a solid-state drive and a corresponding reading and writing method. According to the solid-state drive control chip integrating the MRAM and the solid-state drive, the solid-state drive control chip integrating the MRAM is adopted and the host interface adopts the standard memory interface, so that the reading and writing speed of the solid-state drive is no longer limited by the speed of a serial interface; the reading and writing speed of the solid-state drive is increased; NAND management software is run by the CPU of the solid-state drive control chip integrating the MRAM, so that the technical scheme is easier to apply; due to the adoption of a POP package technology or a 3D SIC technology, the control chip can be used in a mobile phone and a light and thin notebook; and due to the adoption of the MRAM, the writing frequency of an NAND chip is reduced, so that the service life of the NAND chip is prolonged.

The invention discloses a method of optimizing a page-level flash memory transformation layer and belongs to the field of data storage. By means of the characteristic that an SLC flash memory supports part of writing, a mechanism of writing an in-situ log is adopted to reduce expenditure of access of a flash memory mapping page. The method comprises the following steps of: reserving a small part of the flash memory mapping page as a log region; when the flash memory mapping page is updated, establishing a log unit by all mapping items belonging to the mapping page in cache and adding the log region written to the mapping page by employing a part of writing mechanism; when the log region is written fully or the writing number of times exceeds a certain threshold value, combining and operating the mapping items in the flash memory mapping page and the items of the log page, and writing the same to a new flash memory mapping page; and at the moment, the log region in the new mapping page is empty, receiving an updating operation of the new mapping page and employing an in-situ log writing mechanism again. By reducing the expenditure of access of the flash memory mapping page, purposes of improving the performances of a flash memory solid disc and prolonging the service life are achieved.

The invention discloses a design method of a page-level flash memory conversion layer of a solid-state hard disk. The invention divides the NAND flash memory into two parts of a data block area and atranslation block area, wherein the data block area is used for storing conventional user data, the translation block area only stores mapping information, and the whole flash memory adopts the page-based mapping. The RAM is divided into W-CMT, R-CMT and GTD three parts. The W-CMT is used for storing write data mapping information with high access frequency. The R-CMT is used to store the read data mapping information with high access frequency, and the GTD is used to record the address mapping entries in the mapping block. The method of the invention realizes the high-flexibility page-level mapping, and at the same time, greatly reduces the frequent updating of translated pages caused by data updating, thereby improving the overall system performance and prolonging the service life of thesolid-state hard disk.

The invention relates to a dynamic strip system design method and a dynamic strip strategy realization method based on a RAID5 solid-state hard disk array, and relates to the technical field of data storage. Data updating modes of reading, modifying, writing, reading, reconstructing and writing in the RAID5 solid-state hard disk array are replaced by using the mode of constructing dynamic stripe.The dynamic stripe design system of the invention comprises a stripe recombination module and a metadata module. The strip recombination module reconstructs the data blocks of local strips to form dynamic strips. The metadata module records and manages the mapping state and storage state of the data. The combination of the above two modules and a RAID5 array changes the request processing mode ofa traditional RAID5 array. The system separates the writing of full band from the writing of local band, so that data blocks in the local band are constructed to be dynamic strips through recombination. The dynamic and static stripes are mapped and managed by the metadata, which reduces the writing times of the solid-state disk, prolongs the service life of the array, reduces the pre-read delay before writing, and improves the writing performance of the array.

The invention is applicable to the technical field of solid state disks, and provides a solid state disk and a method for optimizing a write request of the solid state disk. The method includes the steps: A, detecting the small-range write request of the solid state disk; B, caching the write request in a memory; C, issuing the write request when a preset condition is met or the write request is finished. Therefore, the write frequency of the solid state disk is effectively decreased, and the service life of the solid state disk is prolonged.

The invention discloses a cache selecting strategy based on classification forecasting. According to the strategy, objects are divided into two types including 'only one-time access' objects and 'more than one-time access' objects by using naive bayesian classification thinking. The 'only one-time access' objects are prevented from being placed in a cache, and the price for classification time and space based on naive bayesian is small. In a caching process, pictures which are not forecasted accurately are compensated through a historical information sheet, and the error probability is reduced. In addition, the forecasting weight of 'only one-time access' is dynamically adjusted according to real-time conditions of the cache, and therefore, the cache can be utilized effectively. Pre-judgment before a replacement algorithm is cached, and cache selecting strategy can be used with an optional cache replacing algorithm. While the complexity of the algorithm of the cache is not changed, the cache hit rate is increased, SSD writing is reduced, and the service life of SSD is prolonged. Meanwhile, machine learning thinking is guided into the cache, and a new consideration is provided for research of the cache algorithm.