CN105930356A - Method for implementing log type heterogeneous hybrid memory file system - Google Patents
Method for implementing log type heterogeneous hybrid memory file system Download PDFInfo
- Publication number
- CN105930356A CN105930356A CN201610218095.3A CN201610218095A CN105930356A CN 105930356 A CN105930356 A CN 105930356A CN 201610218095 A CN201610218095 A CN 201610218095A CN 105930356 A CN105930356 A CN 105930356A
- Authority
- CN
- China
- Prior art keywords
- file system
- hash table
- memory
- page
- internal memory
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/0223—User address space allocation, e.g. contiguous or non contiguous base addressing
- G06F12/0292—User address space allocation, e.g. contiguous or non contiguous base addressing using tables or multilevel address translation means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/10—File systems; File servers
- G06F16/18—File system types
- G06F16/1805—Append-only file systems, e.g. using logs or journals to store data
- G06F16/1815—Journaling file systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/061—Improving I/O performance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0638—Organizing or formatting or addressing of data
- G06F3/0644—Management of space entities, e.g. partitions, extents, pools
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Human Computer Interaction (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a method for implementing a log type heterogeneous hybrid memory file system. A step of partitioning a hybrid memory: performing hierarchical partition on the hybrid memory of the file system, wherein the hybrid memory is partitioned into four different logic blocks including a zone, a section, a segment, and a page, and logical numbering is performed on the four logical blocks. A step of mapping the logical blocks of the hybrid memory: establishing mappings between the logical blocks of the file system and underlying SCM physical addresses, and implementing dynamic allocation and recovery of a page. A step of establishing a multilevel hash table: establishing a multilevel directory structure based on the hash table, wherein the multilevel directory structure is used for searching a file. According to the method, the file system can reasonably allocate memory space according to a hardware character of a heterogeneous memory media, the utilization of the hardware can be improved, the service life of the hardware is prolonged, and the storage efficiency and the access efficiency of the file can be improved.
Description
Technical field
The present invention relates to file system technology field, in particular it relates to a kind of log type isomery mixing memory file system
Implementation method.
Background technology
The rise of novel memory technology goes out with different SCM (Storage Class Memory stores class internal memory) device
Existing, cause people to mixing memory system framework and the concern of way to manage and research.SCM device has non-volatile
Property, read to write soon slow, quiescent dissipation is low and dynamic power consumption high, differs from traditional internal memory.The expert of lot of domestic and foreign
Scholar proposed multiple memory organization mode and optimization method also based on SCM device.But numerous organizational form and excellent
Change method has respective limitation and the scope of application.
Along with the development of the technology such as cloud computing, Internet of Things, data are just constantly increasing with unprecedented speed and are accumulating, as
What manages and organizes these data, is a challenge for traditional file system.From the point of view of upper layer application, although working as
Front file system is that application provides a kind of easy-to-use Method of Data Organization, but the growth of data and Information Complexity expand
All increase resource management and the difficulty of retrieval data, and among data are in and are continually changing, application is difficult to accurately obtain
Potential relation between the overall picture of data and data;Additionally from the perspective of bottom hardware, traditional Method of Data Organization
Also cannot give full play to the effectiveness of DRAM and SCM mixing memory architecture, the most effectively organize data, substantially
The performance improving mixing internal memory is also a challenge.
Summary of the invention
For defect of the prior art, it is an object of the invention to provide a kind of log type isomery mixing memory file system
Implementation method.
According to the implementation method of the log type isomery mixing memory file system that the present invention provides, comprise the steps:
Mixing internal memory partiting step: the mixing internal memory of file system is divided by different level, be i.e. divided into zone (territory),
These four different logical blocks of section (district), segment (section), page (page), and four logical blocks are patrolled
Collect numbering;
Mixing internal memory logical block maps step: build the mapping to bottom SCM physical address of the file system logic block, and
Realize the dynamically allocation and reclaiming of the page;
Multi-level hash table construction step: build multi-level bibliographic structure based on hash table, be used for carrying out ff.
Preferably, the mixing internal memory of file system is divided into four layers, specifically, one by described mixing internal memory partiting step
The mixing internal memory of individual file system comprises multiple zone;One zone comprises 2*n1 physical address continuous print section;
One section comprises 2*n2 physical address continuous print segment;One segment comprises multiple physical address space
Continuous print page;Page is the ultimate unit of memory storage space distribution, and n1, n2 are positive integer.
Preferably, sequential write step is also included: use log-structured (log type structure) and copy-on-write technique pair
Isomery mixing internal memory carries out sequential write operation;Specifically, when a certain blocks of data revised by needs, do not modify in original place,
But after original blocks of data reading DRAM completing amendment, amended data block is written to new SCM block
In.
Preferably, in described mixing internal memory logical block maps step, use node address tree by the logical number of internal memory and the end
Layer SCM physical address maps.
Preferably, the dynamically allocation and reclaiming of the page refers to: one logical number of predistribution is to bottom SCM physical address
Mapping table, when file system need distribute storage unit stores data time, the actual physical address being assigned to is inserted and reflects
One list item of firing table;When wanting freeing of page, only need to be labeled as corresponding mapping item losing efficacy, by mapping table pipe
The distribution of the page of reason bottom SCM and recovery.
Preferably, described multi-level hash table construction step includes: use multiple hash table, each hash table to represent one
Level, there is difference in the entry (item) that every level hash table bucket (bucket) comprises, but same hash table bucket
The entry comprised is the same number of;The TOC level of file is mapped with the level of hash table, when at n-th layer mesh
During record one file of write, attempt starting to map from the n-th hash table, when the n-th hash table is full, reattempt mapping
To (n+1)th hash table, carry out certainly increasing of n, until mapping successfully.
Compared with prior art, the present invention has a following beneficial effect:
1, the implementation method of the log type isomery mixing memory file system that the present invention provides enables file system according to different
The ardware feature rational management distribution memory space of structure memory medium, improves the utilization rate of hardware, extends the use longevity of hardware
Life, improves storage efficiency and the access efficiency of file.
2, the implementation method of log type isomery mixing memory file system that the present invention provides at different grain size to file system
It is managed;Uniformly write strategy for what prolongation isomery mixing internal memory was taked, use log-structured structure and write recurrence of disease at the same time next year
Technology processed, is ensureing while file system consistency, is improving system read-write performance, extends device service life;Node
Number of addresses structure maps in bulk physical address the logical number of mixing internal memory;Multi-level hash table, improves ff effect
Rate.
Accompanying drawing explanation
By the detailed description non-limiting example made with reference to the following drawings of reading, the further feature of the present invention,
Purpose and advantage will become more apparent upon:
Fig. 1 is SCM internal memory logical partition hierarchical structure schematic diagram;
Fig. 2 is file system overall structure schematic diagram;
Fig. 3 is node address tree construction schematic diagram;
Fig. 4 is inode data block schematic diagram;
Fig. 5 is the multi-level hash table that file system directories uses.
Detailed description of the invention
Below in conjunction with specific embodiment, the present invention is described in detail.Following example will assist in those skilled in the art
Member is further appreciated by the present invention, but limits the present invention the most in any form.It should be pointed out that, the common skill to this area
For art personnel, without departing from the inventive concept of the premise, it is also possible to make some changes and improvements.These broadly fall into
Protection scope of the present invention.
According to the implementation method of the log type isomery mixing memory file system that the present invention provides, comprise the steps:
Mixing internal memory partiting step: to mixing internal memory divide by different level, be i.e. divided into zone, section, segment,
These four different logical blocks of page, and four logical blocks are carried out logical number;
Wherein, Page is the ultimate unit of Memory Allocation;Information is used to carry out statistics description in the space of each segment,
Each segment only stores data block or node block;The information such as the space utilization rate according to each segment are carried out
The distribution of memory space, recovery etc. operate;
Mixing internal memory logical block maps step: build the mapping to bottom SCM physical address of the topmost paper system logic block,
And realize the dynamically allocation and reclaiming of the page;
Specifically, quote a kind of virtual memory abstract mechanism, multiple SCM equipment can be regarded as an equipment, pass through
Unified mapping table realizes the access to multiple SCM equipment.Refer to an Address space mappinD mechanism, file unit
The logic unit nid of the simply data place block of data structure inode record, the physical address of data to be obtained needs to pass through
Nid is converted into physical address by NAT tree.This virtual technology is utilized to realize topmost paper system logic block to bottom SCM
The mapping of physical address, hides actual bottom physical address space, the operation of simplified-file object, and realizes the dynamic of the page
State distribution and removal process.By the Address Mapping of NAT, the change of physical address only need to revise mapping layer content,
It is not required to revise the logical address of topmost paper system, the amendment process of simplified-file system, improves the availability of file system.
Multi-level hash table construction step: use multiple hash table, each hash table to represent a level.Every level hash
The entry that table bucket comprises may be different, but the entry number that same hash table bucket comprises is identical.
The TOC level of file is mapped with the level of hash table, when n-th layer catalogue write one file time, attempt from
N-th hash table starts to map, and when the n-th hash table is full, reattempts and is mapped to (n+1)th hash table.Repeat this
Step, until inserting successfully.
Use multi-level hash table that file is made a look up, reduce time complexity, improve ff efficiency.
Specifically, as it is shown in figure 1, SCM internal memory is carried out logical partition, whole internal memory is divided into four layers, is followed successively by
Zone, section, segment, page tetra-layers.One memory block can comprise multiple zone;One zone comprises
2*n1 physical address continuous print section;One section comprises 2*n2 physical address continuous print segment;One
Individual segment comprises multiple physical address space continuous print page;Page is the ultimate unit of memory storage space distribution,
N1, n2 are positive integer.
Within save as the SCM of 1TB size as a example by, a file minimum gets the memory space of a page size, that
Maximum number of files on whole SCM is 1TB/4KB=228.Assume to describe the metadata required for a file
Size is 256 bytes, and the maximum memory space that the metadata of All Files needs is 256 bytes * 228=64GB.
File system needs often access these metadata, on the one hand higher to the response time requirement accessing metadata,
The most frequently read-write can reduce the service life of SCM.DRAM supports random read-write, and speed is faster than SCM
And the life-span is longer than SCM.When in carry file system, by metadata replication to DRAM;When system is run,
Data in direct read/write DRAM, and every some cycles with regard to the data syn-chronization of DRAM in SCM.
Such as Fig. 2, file system overall structure, wherein the metadata of file system includes: superblock (super block, SB),
Segment information table (segment information table, SIT), section Statistical Area (segment summary area, SSA),
Node address tree (Node Address Tree, NAT), can describe the overall of SCM internal memory by these information and use shape
State.
Superblock is positioned at the starting position of file system, contains all essential informations of file system, uses including system
Space, the total quantity of section, the total quantity of block, checkpoint list etc..
SIT comprises the fileinfo of each segment, such as the memory block number being currently in use, the record internal memory of segment
The bitmap whether block is currently being used.
SSA have recorded owner's information such as father node inode of block and it is in the side-play amount of father node.
Nid is converted to block physical address by node address tree (Node Address Tree, NAT).Node address tree is fixed
Use the tree construction of 4 layers, the block physical address that leaf node storage nid is corresponding.
Refer to an Address space mappinD mechanism, the logic unit nid of the simply data place block of inode record, obtain
The physical address obtaining data needs, by NAT tree, nid is converted into physical address.
As shown in Fig. 3 node address tree, the lookup mode of NAT is as follows:
Assume that nid is the integer of 36, with (nid > > 27) &0x1FF carry out the ground floor to NAT addressing, with (nid > >
18) &0x1FF carry out the addressing of the second layer to NAT, with (nid > > 9) &0x1FF carry out the third layer to NAT
Addressing, with (nid) &0x1FF carries out the to NAT the 4th layer of addressing and obtains leaf node, reads the interior of leaf node
Hold i.e. can get nid for physical address.
Utilize this virtual technology to realize the mapping to bottom SCM physical address of the topmost paper system logic block, hide real
Border bottom physical address space, the operation of simplified-file object, and realize the dynamically allocation and reclaiming process of the page.Pass through
The Address Mapping of NAT, the change of physical address only need to be revised mapping layer content, be not required to revise topmost paper system
Logical address, the operating process of simplified-file system, improves the availability of file system.
In SCM, store data message, owing to the service life of SCM is relevant with the number of times write, uniformly writes and can extend
The service life of SCM, so file system uses log-structured structure to SCM by COW (copy on wirte)
Mode sequential write.When a certain blocks of data revised by needs, do not modify in original place, but original block is read in
After DRAM completes amendment, amended piece is written in new SCM block.
Each segment only stores data block or node block.
Data block stores user data, and node block storage file system internal information, such as pointer gauge.
There are 6 kinds of node types in systems: inode, direct node, indirect node, nat node, checkpoint
node.The size of each type node block is all 4KB.Inode stores file section dot information, and direct node storage refers to
To the pointer of data block, the pointer of direct node, nat node storage nat tree node letter are pointed in indirect node storage
Breath.Checkpoint node storage checkpoint information.One node is by inode metadata, 929 data block pointer,
2 indirect block pointers, 2 secondary indirect block pointers, 1 three indirect block pointers composition.Such as Fig. 4 inode number
Shown in block schematic diagram.One directory entry occupies 11 bytes, including the hash value of file name, inode number, literary composition
Part name length, file type four partial information.One directory block comprises 214 directory entries and filename, and containing one
It is the most effective that individual bitmap is used for recording corresponding directory entry.In the realization of bibliographic structure, introduce multi-level hash
Table, as shown in the multi-level hash table that Fig. 5 file system directories uses.The calculating of the number of the block of each bucket is public
Formula is f (n):
The bucket number of each level is g (n, dirlevel):
In formula: n represents the hash number of plies, dirlevel represents the number of plies at catalogue place, and N represents total number of plies of hash.
When file system locating file name in catalogue, the hash value of first calculation document name, then file system
Begin look for comprising the directory entry of file name from the 0th layer.Without finding, system can search the hash of next layer
Table.In the hash table of each layer, only need to search for a bucket.Bufcket sequence number h (n) to be searched for is by following public affairs
Formula gives:
H (n)=Hash (filename) %g (n, dirlevel)
Hash (filename) is the hash value of filename, and g (n, dirlevel) is the bucket number of the n-th+dirlevel layer.
Above the specific embodiment of the present invention is described.It is to be appreciated that the invention is not limited in
Stating particular implementation, those skilled in the art can make a variety of changes within the scope of the claims or revise,
This has no effect on the flesh and blood of the present invention.In the case of not conflicting, in embodiments herein and embodiment
Feature can arbitrarily be mutually combined.
Claims (6)
1. the implementation method of a log type isomery mixing memory file system, it is characterised in that comprise the steps:
Mixing internal memory partiting step: the mixing internal memory of file system is divided by different level, be i.e. divided into zone, section,
These four different logical blocks of segment, page, and four logical blocks are carried out logical number;
Mixing internal memory logical block maps step: build the mapping to bottom SCM physical address of the file system logic block, and
Realize the dynamically allocation and reclaiming of the page;
Multi-level hash table construction step: build multi-level bibliographic structure based on hash table, be used for carrying out ff.
The implementation method of log type isomery mixing memory file system the most according to claim 1, it is characterised in that
The mixing internal memory of file system is divided into four layers by described mixing internal memory partiting step, specifically, a file system
Mixing internal memory comprises multiple zone;One zone comprises 2*n1 physical address continuous print section;One section
Comprise 2*n2 physical address continuous print segment;One segment comprises multiple physical address space continuous print page;
Page is the ultimate unit of memory storage space distribution, and n1, n2 are positive integer.
The implementation method of log type isomery mixing memory file system the most according to claim 1, it is characterised in that
Also include sequential write step: use log-structured structure and copy-on-write technique that isomery mixing internal memory is carried out sequential write
Operation;Specifically, when a certain blocks of data revised by needs, do not modify in original place, but original blocks of data is read
Enter after DRAM completes amendment, amended data block is written in new SCM block.
The implementation method of log type isomery mixing memory file system the most according to claim 1, it is characterised in that
In described mixing internal memory logical block maps step, use node address tree by the logical number of internal memory and bottom SCM physics
Address maps.
The implementation method of log type isomery mixing memory file system the most according to claim 1, it is characterised in that
The dynamically allocation and reclaiming of the page refers to: the mapping table of one logical number of predistribution to bottom SCM physical address, when
When file system needs to distribute storage unit stores data, the actual physical address being assigned to is inserted a table of mapping table
?;When wanting freeing of page, only need to be labeled as corresponding mapping item losing efficacy, manage bottom SCM by mapping table
The distribution of the page and recovery.
The implementation method of log type isomery mixing memory file system the most according to claim 1, it is characterised in that
Described multi-level hash table construction step includes: use multiple hash table, each hash table to represent a level, every level
There is difference in the entry that hash table bucket comprises, but the entry that same hash table bucket comprises is the same number of;
The TOC level of file is mapped with the level of hash table, when n-th layer catalogue write one file time, attempt from
N-th hash table starts to map, and when the n-th hash table is full, reattempts and is mapped to (n+1)th hash table, carry out n
From increasing, until map successfully.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610218095.3A CN105930356A (en) | 2016-04-08 | 2016-04-08 | Method for implementing log type heterogeneous hybrid memory file system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610218095.3A CN105930356A (en) | 2016-04-08 | 2016-04-08 | Method for implementing log type heterogeneous hybrid memory file system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105930356A true CN105930356A (en) | 2016-09-07 |
Family
ID=56840441
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610218095.3A Pending CN105930356A (en) | 2016-04-08 | 2016-04-08 | Method for implementing log type heterogeneous hybrid memory file system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105930356A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106649152A (en) * | 2016-12-29 | 2017-05-10 | 郑州云海信息技术有限公司 | System log storage method and device based on non-volatile memory |
CN106776361A (en) * | 2017-03-10 | 2017-05-31 | 安徽大学 | A kind of caching method and system towards extensive non-volatile memory medium |
CN107544756A (en) * | 2017-08-03 | 2018-01-05 | 上海交通大学 | Method is locally stored in Key Value log types based on SCM |
CN108920110A (en) * | 2018-07-25 | 2018-11-30 | 湖南机电职业技术学院 | A kind of parallel processing big data storage system and method calculating mode based on memory |
CN109144432A (en) * | 2018-10-08 | 2019-01-04 | 郑州云海信息技术有限公司 | A kind of Nonvolatile memory reservoir page distribution method and device |
CN110688345A (en) * | 2019-09-26 | 2020-01-14 | 重庆大学 | Multi-granularity structured space management mechanism of memory file system |
CN112306896A (en) * | 2019-07-31 | 2021-02-02 | 华为技术有限公司 | Garbage recovery method and device |
CN113051241A (en) * | 2019-12-27 | 2021-06-29 | 中国移动通信集团湖南有限公司 | Database persistence method, device and equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100332725A1 (en) * | 2009-06-24 | 2010-12-30 | Post Samual D | Pinning content in nonvolatile memory |
CN103186350A (en) * | 2011-12-31 | 2013-07-03 | 北京快网科技有限公司 | Hybrid storage system and hot spot data block migration method |
CN103942159A (en) * | 2014-03-19 | 2014-07-23 | 华中科技大学 | Data read-write method and device based on mixed storage device |
US20150205727A1 (en) * | 2014-01-17 | 2015-07-23 | Netapp, Inc. | Set-associative hash table organization for efficient storage and retrieval of data in a storage system |
-
2016
- 2016-04-08 CN CN201610218095.3A patent/CN105930356A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100332725A1 (en) * | 2009-06-24 | 2010-12-30 | Post Samual D | Pinning content in nonvolatile memory |
CN103186350A (en) * | 2011-12-31 | 2013-07-03 | 北京快网科技有限公司 | Hybrid storage system and hot spot data block migration method |
US20150205727A1 (en) * | 2014-01-17 | 2015-07-23 | Netapp, Inc. | Set-associative hash table organization for efficient storage and retrieval of data in a storage system |
CN103942159A (en) * | 2014-03-19 | 2014-07-23 | 华中科技大学 | Data read-write method and device based on mixed storage device |
Non-Patent Citations (9)
Title |
---|
CHANGMAN LEE ETAL.: "F2FS: A New File System for Flash Storage", 《THE PROCEEDINGS OF THE 13TH USENIX CONFERENCE ON FILE AND STORAGE TECHNOLOGIES》 * |
CSDN: "F2FS文件系统架构与原理分析(一)——设计背景与功能", 《HTTPS://BLOG.CSDN.NET/ROBERTSONG2004/ARTICLE/DETAILS/43347403》 * |
CSDN: "F2FS文件系统架构与原理分析(三) ——文件索引树", 《HTTPS://BLOG.CSDN.NET/ZWJ0403/ARTICLE/DETAILS/12970811》 * |
CSDN: "F2FS文件系统架构与原理分析(二)——磁盘布局", 《HTTPS://BLOG.CSDN.NET/FYBON/ARTICLE/DETAILS/36008605》 * |
CSDN: "F2FS文件系统架构与原理分析(五)——元数据组织及管理", 《HTTPS://BLOG.CSDN.NET/FYBON/ARTICLE/DETAILS/38752977》 * |
CSDN: "F2FS文件系统架构与原理分析(六)——块分配与空间管理", 《HTTPS://BLOG.CSDN.NET/ZWJ0403/ARTICLE/DETAILS/12970835》 * |
CSDN: "F2FS文件系统架构与原理分析(四)——F2FS的目录结构与目录哈希", 《HTTPS://BLOG.CSDN.NET/ZWJ0403/ARTICLE/DETAILS/12970819》 * |
SOOJUN IM,DONGKUN SHIN: "Differentiated Space Allocation for Wear Leveling on Phase-Change Memory-based Storage Device", 《IEEE》 * |
道客巴巴: "F2FS文件系统实现分析及其在移动设备上的优化", 《HTTP://WWW.DOC88.COM/P-1806941151295.HTML》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106649152A (en) * | 2016-12-29 | 2017-05-10 | 郑州云海信息技术有限公司 | System log storage method and device based on non-volatile memory |
CN106776361A (en) * | 2017-03-10 | 2017-05-31 | 安徽大学 | A kind of caching method and system towards extensive non-volatile memory medium |
CN106776361B (en) * | 2017-03-10 | 2020-07-10 | 安徽大学 | Caching method and system for large-scale nonvolatile storage medium |
CN107544756A (en) * | 2017-08-03 | 2018-01-05 | 上海交通大学 | Method is locally stored in Key Value log types based on SCM |
CN107544756B (en) * | 2017-08-03 | 2020-11-06 | 上海交通大学 | Key-Value log type local storage method based on SCM |
CN108920110A (en) * | 2018-07-25 | 2018-11-30 | 湖南机电职业技术学院 | A kind of parallel processing big data storage system and method calculating mode based on memory |
CN109144432A (en) * | 2018-10-08 | 2019-01-04 | 郑州云海信息技术有限公司 | A kind of Nonvolatile memory reservoir page distribution method and device |
CN112306896A (en) * | 2019-07-31 | 2021-02-02 | 华为技术有限公司 | Garbage recovery method and device |
CN112306896B (en) * | 2019-07-31 | 2022-12-27 | 华为技术有限公司 | Garbage recycling method and device |
CN110688345A (en) * | 2019-09-26 | 2020-01-14 | 重庆大学 | Multi-granularity structured space management mechanism of memory file system |
CN113051241A (en) * | 2019-12-27 | 2021-06-29 | 中国移动通信集团湖南有限公司 | Database persistence method, device and equipment |
CN113051241B (en) * | 2019-12-27 | 2023-08-15 | 中国移动通信集团湖南有限公司 | Database persistence method, device and equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105930356A (en) | Method for implementing log type heterogeneous hybrid memory file system | |
CN104395904B (en) | Efficient data object storage and retrieval | |
US9149054B2 (en) | Prefix-based leaf node storage for database system | |
Teng et al. | LSbM-tree: Re-enabling buffer caching in data management for mixed reads and writes | |
CN105468542B (en) | Address distribution method and device | |
CN105677826A (en) | Resource management method for massive unstructured data | |
CN104331453B (en) | A kind of construction method of distributed file system and distributed file system | |
CN106021031B (en) | A kind of the deletion data reconstruction method and device of BTRFS file system | |
CN103823865A (en) | Database primary memory indexing method | |
CN102411634B (en) | A kind of date storage method promoting embedded database real-time | |
CN101571869B (en) | File memory and read method of smart card and device thereof | |
CN101488153A (en) | Method for implementing high-capacity flash memory file system in embedded type Linux | |
Brinkhoff et al. | The impact of global clustering on spatial database systems | |
CN102567415B (en) | Control method and device of database | |
CN107544756A (en) | Method is locally stored in Key Value log types based on SCM | |
CN109407985B (en) | Data management method and related device | |
CN108052541A (en) | The realization of file system based on multi-level page-table bibliographic structure, access method, terminal | |
Min et al. | A system framework for map air update navigation service | |
CN111857582B (en) | Key value storage system | |
CN103942301B (en) | Distributed file system oriented to access and application of multiple data types | |
Amur et al. | Design of a write-optimized data store | |
CN113253926A (en) | Memory internal index construction method for improving query and memory performance of novel memory | |
CN106528756A (en) | Network map data organization method based on space-time relevance | |
CN102253985A (en) | File system data management method and system | |
CN107273443B (en) | Mixed indexing method based on metadata of big data model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160907 |