CN102915340A - Expanded B+ tree-based object file system - Google Patents
Expanded B+ tree-based object file system Download PDFInfo
- Publication number
- CN102915340A CN102915340A CN2012103515250A CN201210351525A CN102915340A CN 102915340 A CN102915340 A CN 102915340A CN 2012103515250 A CN2012103515250 A CN 2012103515250A CN 201210351525 A CN201210351525 A CN 201210351525A CN 102915340 A CN102915340 A CN 102915340A
- Authority
- CN
- China
- Prior art keywords
- extent
- tree
- space
- idle
- free
- 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
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
An expanded B+ tree-based object file system comprises a free space organization module, a free space distribution module and an object space distribution module, wherein the free space organization module is used for setting that extent refers to disk free space; the extent is represented by a binary array which uses a block initial address and a continuous block length as parameters; for a plurality of the free extents in each group, except for considering the size of each extent, the initial address of each extent is taken as a key word to be organized by a B+ tree; the free space distribution module is used for distributing according to the extent when the disk free space is distributed; when a free block is distributed, the size of the free block is considered, the initial address of the free block is also considered, and a proximal optimal matching principle is used; and the object space distribution module is used for distributing space during request writing at a time through a pre-distribution method, and for setting a fixed pre-distribution value. By the expanded B+ tree-based object file system, reliability and expandability are improved, and input and output speed and data throughput rate of a storage system are increased.
Description
Technical field
The present invention relates to computer memory technical, especially a kind of object-based file system.
Background technology
The develop rapidly of infotech is so that information storage application is day by day simple, and memory capacity increases severely, and the file system of a new generation need to support to have efficient expansion and efficient high-performance computing environment of managing.Traditional distributed file system generally comprises file management and two assemblies of storage administration.What document component management was realized is the functions such as layer management, access control and name; The storage administration assembly is that logic data structure is mapped on the bottom physical storage medium, realizes physical store.In the large-scale distributed storage system of a new generation, with each independently memory device add respectively the storage administration assembly, like this can so that memory device to have efficient expansion and a high-efficiency pipe rational.
Object-based file system also is to adopt so model, and the memory device of conservation object is called as OSD (object-based storage devices).OSD is a kind of smart machine, comprises the control assemblies such as internal memory, processor, storage medium, network interface.OSD provides external access interface and manages mapping relations between native object and object and memory device.
The major function of OSD is data storage and secure access.The OSD Utility object manages all data.It gets up to consist of object with magnetic track and the sector of deposit data at disk with some magnetic tracks and sector combination, and provides access to data to the external world by this object.Each object includes the access interfaces such as read operation, write operation, and each object may comprise some files, also may be the part of certain file, and each object is independent of operating system.Except concrete user data, OSD has also recorded each attribute information, and existing static attribute (such as time of setting up etc.) also has dynamic attribute (such as the last access time etc.).According to different feature application, the founder can arrange different contents and size for object properties.These information leave on the OSD, have greatly alleviated the burden of meta data server, have strengthened concurrent access performance and the extensibility of whole storage system.The OSD model as shown in Figure 1.
Since the rapid increase of memory capacity, the continuous increase of memory device quantity, and the storage administration of memory device efficient will have influence on the performance of whole system.OSD in the distributed system of a lot of Object-based Storages adopts disk file system to carry out storage administration.Yet these object-based file systems are in following problem:
At first object-based file system does not have the TOC level concept, and all object IDs have consisted of a huge flat namespace.This meeting so that in the local file system optimization based on TOC level information can not directly apply on the OSD.For example, local file system (ext2 or ext3 etc.) is usually take catalogue as clue to file data layout, buffer memory with look ahead and effectively optimize, and OSD difficult when upper was moved in these optimizations.In addition, huge flat namespace need to be carried out new optimization so that object is searched.
Secondly, object-based file system may be supported a lot of large files.Majority all is small documents in the local file system, thereby the organization and administration mode centering small files of file is effective especially.May just lose original high efficiency for large file.
Under specific upper strata distributed file system was disposed, the object-based file system load may have distincter characteristic distributions on Object-Based Storage Controller, at this moment needs to adopt special magnetic disk and allocation strategy to improve the performance of object-based file system at last.
Summary of the invention
Relatively poor for the reliability that overcomes existing object-based file system, extensibility is relatively poor, the input and output speed of storage system and the lower deficiency of data throughput, the invention provides a kind of input and output speed that improves reliability and extensibility, raising storage system and data throughput expansion based on B+ tree object-based file system.
The technical solution adopted for the present invention to solve the technical problems is:
A kind of expansion based on B+ tree object-based file system, describedly comprise based on B+ tree object-based file system: the free space molded tissue block, be used for to set extent and represent disk slack space, extent is that the binary array of parameter shows with piece start address and continuous blocks length; Be divided into a plurality of groups by the whole disk slack space of the large young pathbreaker of idle extent, each group usage data structure bucket represents, extent size in each group satisfies boundary limitation, set a group factor B, suppose that free space is divided into N group altogether, N 〉=2, for any idle extent in the k group, its big or small S satisfies the formula following formula so:
Wherein, P represents the figure place of storage system;
Only comprise a large idle extent during initialization on the disk, it belongs to N-1 bucket; For a plurality of idle extent in every group, except the size of considering each extent, also organize with the B+ tree take the start address of each extent as key word;
The free space distribution module, be used for when distributing disk slack space, distributing by extent, both considered when distributing free block that the free block size also considered the free block start address, adopt immediate Optimum Matching principle, at first determine that according to required free block size minimum comprises the bucket of required free block size, if the bucket that satisfies condition does not have the space that can distribute, then look for the bucket that comprises larger free block, find first address among this bucket near the idle extent of required free block start address according to required free block first address again; In case the extent that satisfies condition finds, remove from bucket near the part of required free block among the extent and distribute to this object, remaining idle component is inserted into start address by size and goes among the corresponding bucket among the extent;
The object space distribution module, give the object allocation space for the employing method for pre-distributing a write request, a fixing predistribution value is set, the end of supposing object also and does not arrive and need to be the outer storage space of object allocation, successively according to the idle extent of the ascending ordering in predistribution value address, until find the extent of this operation actual needs size; If find an idle extent who surpasses predistribution value size, divider will distribute this free time extent by predistribution value size.
Further, in the described free space molded tissue block, the B+ tree is carried out two kinds search, a kind of is to begin to carry out sequential search from minimum key word; Another is to begin to find leafy node from root node; The idle extent that utilizes B+ tree quick-searching to need, B+ leaf node points to an idle extent, and wherein key word represents the start address of each extent, and Pointer represents the length of each extent, and intermediate node is used for the index of B+ tree node piece.
Beneficial effect of the present invention is mainly manifested in: can the greatly reducing the extent number of single object inside and safeguard the continuity of single object internal data based on B+ tree object-based file system of expansion, thus look for expense in the time of can reducing object accesses.Because the disk space management method based on B+ tree object-based file system of expansion does one's utmost to keep the continuity of allocation space of disk, thereby so that whole object-based file system can have good performance in long-term use procedure.
Description of drawings
Fig. 1 is the synoptic diagram of OSD model.
Fig. 2 is the synoptic diagram of disk slack space tissue.
Fig. 3 is the synoptic diagram that disk slack space distributes.
Fig. 4 is the synoptic diagram that distributes the front disk spatiality.
Fig. 5 is the synoptic diagram of disk space state after distributing.
Embodiment
The invention will be further described below in conjunction with accompanying drawing.
With reference to Fig. 2 ~ Fig. 5, a kind of expansion based on B+ tree object-based file system, comprise three parts: 1. free space tissue; 2. free space distributes; 3. object space distributes.
Free space tissue: suppose that extent represents disk slack space, then extent can be that the binary array of parameter shows with piece start address and continuous blocks length, represents that with extent continuous blocks represent to have soon continuously higher efficient and terseness than general with piece.The metadata that need to write when extent can make large file operation is the minimizing of the order of magnitude, thereby can assist to improve the document order readwrite performance.In addition, because in most cases disk has a large amount of continuously free spaces, extent also can more effective hyperdisk free space.
In order to improve the efficiency of management of disk slack space, a kind of expansion be divided into a plurality of groups based on B+ tree object-based file system by the whole disk slack space of the large young pathbreaker of idle extent, each group usage data structure bucket(bucket) expression.Extent size in each group satisfies certain boundary limitation.A kind of expansion set a group factor B based on B+ tree object-based file system, suppose that free space is divided into individual group of N (N 〉=2) altogether, any idle extent in organizing for k so, its big or small S satisfies the formula following formula:
A kind of expansion based on the initialization of B+ tree object-based file system the time only comprise a large idle extent (the maximum free space of initial disk) on the disk, it belongs to N-1 bucket(bucket).Along with object space ceaselessly distributes and reclaims each bucket(bucket) in idle extent ceaselessly delete and add, idle extent is occuring to change on the disk, the fragment of disk slack space is on the increase.
For a plurality of idle extent in every group, except the size of considering each extent, also the organizing with the B+ tree take the start address of each extent as key word (Key) based on B+ tree object-based file system of a kind of expansion.The B+ tree is a kind of multichannel balance search tree, can carry out two kinds to the B+ tree and search, and a kind of is to begin to carry out sequential search from minimum key word, and this is conducive to the situation of sequential index; Another is to begin to find leafy node from root node, and this is conducive to the processing of stochastic searching.Because good performance (time complexity is O (logn)) is being searched, inserting, can provided when deleting to the B+ tree, the idle extent that utilizes B+ tree quick-searching to need.B+ leaf node points to an idle extent, and wherein Key represents the start address of each extent, and Pointer represents the length of each extent.Intermediate node is used for the index of B+ tree node piece.
Free space distributes: expansion distribute disk slack space based on B+ tree object-based file system the time distribute by extent, always attempt to distribute the continuous single extent of bulk for object at disk.For small object, only need little idle extents just can finish distribution.For large object, if independent large free block lacks, may need to distribute a plurality of idle extent.In this case, these extent should be as far as possible close on disk, thereby so that object-order when access tracking expense is minimum.Tracking expense when reducing object accesses, expansion distribute free block based on B+ tree object-based file system the time both considered that the free block size also considered the free block start address.That expands adopts immediate Optimum Matching principle based on B+ tree object-based file system, at first determines that according to required free block size minimum comprises the bucket(bucket of required free block size).If the bucket that satisfies condition does not have the space that can distribute, then look for the bucket(bucket that comprises larger free block).Expansion find this bucket(bucket according to required free block first address again based on B+ tree object-based file system) in first address near the idle extent of required free block start address.In case the extent that satisfies condition finds, among the extent near the part of required free block from the bucket(bucket) remove and distribute to this object, remaining idle component is inserted among the corresponding bucket with start address by size and goes among the extent.
Object space distributes: the object space distribution refers to object-based file system is that how many free block spaces are single object write operation requests distribute.That expands has much relations based on the performance of B+ tree object-based file system and the partition size of object space.For fragment and the raising system performance that reduces system, expansion be not that needs according to this write operation are that object distributes idle extent based on B+ tree object-based file system, but adopting method for pre-distributing in a write request, to distribute more space to object, the data that can make repeatedly like this write operation are more close and can reduce the expense that allocation of space is brought on disk.That expands arranges a fixing predistribution value based on B+ tree object-based file system, the end of supposing object also and does not arrive and need to be the outer storage space of object allocation, expansion will be by the immediate idle extent in predistribution value address based on B+ tree object-based file system divider.If this extent can not find, divider will find slightly little extent, finds till the extent of this operation actual needs size always.If find an idle extent who surpasses predistribution value size, divider will distribute this free time extent by predistribution value size.
Embodiment: a kind of expansion be divided into a plurality of groups based on B+ tree object-based file system by the whole disk slack space of the large young pathbreaker of idle extent, each group usage data structure bucket(bucket) expression.Extent size in each group satisfies certain boundary limitation.A kind of expansion set a group factor B based on B+ tree object-based file system, suppose that free space is divided into individual group of N (N 〉=2) altogether, any idle extent in organizing for k so, its big or small S satisfies the formula following formula:
For a plurality of idle extent in every group, except the size of considering each extent, also the organizing with the B+ tree take the start address of each extent as key word (Key) based on B+ tree object-based file system of a kind of expansion.The B+ tree is a kind of multichannel balance search tree, can carry out two kinds to the B+ tree and search, and a kind of is to begin to carry out sequential search from minimum key word, and this is conducive to the situation of sequential index; Another is to begin to find leafy node from root node, and this is conducive to the processing of stochastic searching.Because good performance (time complexity is O (logn)) is being searched, inserting, can provided when deleting to the B+ tree, the free space extent that utilizes B+ tree quick-searching to need.B+ leaf node points to an idle extent, and wherein Key represents the start address of each extent, and Pointer represents the length of each extent.Intermediate node is used for the index of B+ tree node piece.
An example based on the disk free block spatial organization of B+ tree object-based file system of expansion.Group factor B=4 among Fig. 2, the 3rd expression of bucket array be the free block collection of all sizes between 16 to 64 on front disk.These free blocks are organized by the B+ tree, and the 3rd of the bucket array comprises a pointer that points to this B+ tree root node.As can be seen from Figure 2, when front disk comprise three start addresses less than 95, the free block (extent) of size between 16 to 64, namely (16,18), (40,21), (65,24), these idle extent sizes are respectively 18,21 and 24, and they couple together with pointer by the order of address increment.
If give that an object allocated size is 17, start address is 41 free block space, then allocation algorithm at first finds bucket[3] in idle extent.If bucket[3] in do not have idle extent, so just at bucket[4] in search, in the bucket that comprises larger free block space, seek, until find the bucket that meets the demands or failure always.When finding the bucket[3 shown in Fig. 3] after, allocation algorithm is chosen the immediate free block (40 in address, 21), and with free block (41,17) distribute to this object, then with in the former free block because of distribute remaining new free block (40, l) and (58,3) join bucket[0] and bucket[1] in (according to their size and start address).
Expansion adopt method for pre-distributing in a write request, to distribute more space to object based on B+ tree object-based file system, the data that can make repeatedly like this write operation are more close and can reduce the expense that allocation of space is brought on disk.That expands arranges a fixing predistribution value based on B+ tree object-based file system, the end of supposing object also and does not arrive and need to be the outer storage space of object allocation, expansion will be by the immediate idle extent in predistribution value address based on B+ tree object-based file system divider.If this extent can not find, divider will find slightly little extent, finds till the extent of this operation actual needs size always.If find an idle extent who surpasses predistribution value size, divider will distribute this free time extent by predistribution value size, with reference to Fig. 4 and Fig. 5.
Claims (2)
- An expansion based on B+ tree object-based file system, it is characterized in that: describedly comprise based on B+ tree object-based file system:The free space molded tissue block be used for to be set extent and is represented disk slack space, and extent is that the binary array of parameter shows with piece start address and continuous blocks length; Be divided into a plurality of groups by the whole disk slack space of the large young pathbreaker of idle extent, each group usage data structure bucket represents, extent size in each group satisfies boundary limitation, set a group factor B, suppose that free space is divided into N group altogether, N 〉=2, for any idle extent in the k group, its big or small S satisfies the formula following formula so:Wherein, P represents the figure place of storage system;Only comprise a large idle extent during initialization on the disk, it belongs to N-1 bucket; For a plurality of idle extent in every group, except the size of considering each extent, also organize with the B+ tree take the start address of each extent as key word;The free space distribution module, be used for when distributing disk slack space, distributing by extent, both considered when distributing free block that the free block size also considered the free block start address, adopt immediate Optimum Matching principle, at first determine that according to required free block size minimum comprises the bucket of required free block size, if the bucket that satisfies condition does not have the space that can distribute, then look for the bucket that comprises larger free block, find first address among this bucket near the idle extent of required free block start address according to required free block first address again; In case the extent that satisfies condition finds, remove from bucket near the part of required free block among the extent and distribute to this object, remaining idle component is inserted into start address by size and goes among the corresponding bucket among the extent;The object space distribution module, give the object allocation space for the employing method for pre-distributing a write request, a fixing predistribution value is set, the end of supposing object also and does not arrive and need to be the outer storage space of object allocation, successively according to the idle extent of the ascending ordering in predistribution value address, until find the extent of this operation actual needs size; If find an idle extent who surpasses predistribution value size, divider will distribute this free time extent by predistribution value size.
- A kind of expansion as claimed in claim 1 based on B+ tree object-based file system, it is characterized in that: in the described free space molded tissue block, the B+ tree is carried out two kinds search, a kind of is to begin to carry out sequential search from minimum key word; Another is to begin to find leafy node from root node; The idle extent that utilizes B+ tree quick-searching to need, B+ leaf node points to an idle extent, and wherein key word represents the start address of each extent, and Pointer represents the length of each extent, and intermediate node is used for the index of B+ tree node piece.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012103515250A CN102915340A (en) | 2012-02-29 | 2012-09-20 | Expanded B+ tree-based object file system |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210050098.2 | 2012-02-29 | ||
| CN201210050098 | 2012-02-29 | ||
| CN2012103515250A CN102915340A (en) | 2012-02-29 | 2012-09-20 | Expanded B+ tree-based object file system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102915340A true CN102915340A (en) | 2013-02-06 |
Family
ID=47613707
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012103515250A Pending CN102915340A (en) | 2012-02-29 | 2012-09-20 | Expanded B+ tree-based object file system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102915340A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104731905A (en) * | 2015-03-24 | 2015-06-24 | 浪潮集团有限公司 | Volume reducing method for simplified storage pool |
| CN105243112A (en) * | 2015-09-25 | 2016-01-13 | 天脉聚源（北京）科技有限公司 | Object indexing control method and apparatus |
| WO2016192025A1 (en) * | 2015-06-01 | 2016-12-08 | SZ DJI Technology Co., Ltd. | Systems and methods for memory architecture |
| CN109086002A (en) * | 2018-06-28 | 2018-12-25 | 平安科技（深圳）有限公司 | Space management, device, computer installation and the storage medium of storage object |
| WO2020034729A1 (en) * | 2018-08-17 | 2020-02-20 | 华为技术有限公司 | Data processing method, related device, and computer storage medium |
| CN111221472A (en) * | 2019-12-26 | 2020-06-02 | 天津中科曙光存储科技有限公司 | Multi-block allocation strategy optimization method and system for disk space allocation |
| CN111581206A (en) * | 2019-03-15 | 2020-08-25 | 北京忆芯科技有限公司 | B + tree operation device and method |
| CN112667637A (en) * | 2020-12-31 | 2021-04-16 | 中移（杭州）信息技术有限公司 | Data management method, device and computer readable storage medium |
| CN113805816A (en) * | 2021-09-27 | 2021-12-17 | 济南浪潮数据技术有限公司 | Disk space management method, device, equipment and storage medium |
| CN114238343A (en) * | 2021-12-23 | 2022-03-25 | 南京华飞数据技术有限公司 | Implementation method of multidimensional variability automatic data model based on big data |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1845093A (en) * | 2006-04-30 | 2006-10-11 | 华中科技大学 | Attribute extensible object file system |
| US20110252067A1 (en) * | 2010-04-12 | 2011-10-13 | Symantec Corporation | Insert optimization for b+ tree data structure scalability |
-
2012
- 2012-09-20 CN CN2012103515250A patent/CN102915340A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1845093A (en) * | 2006-04-30 | 2006-10-11 | 华中科技大学 | Attribute extensible object file system |
| US20110252067A1 (en) * | 2010-04-12 | 2011-10-13 | Symantec Corporation | Insert optimization for b+ tree data structure scalability |
Non-Patent Citations (2)
| Title |
|---|
| 王继敏: "对象存储设备本地文件系统的研究", 《中国优秀硕士论文电子期刊网》 * |
| 雷伟: "OSD中对象存储管理的方法和实现技术", 《中国优秀硕士论文电子期刊网》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104731905A (en) * | 2015-03-24 | 2015-06-24 | 浪潮集团有限公司 | Volume reducing method for simplified storage pool |
| WO2016192025A1 (en) * | 2015-06-01 | 2016-12-08 | SZ DJI Technology Co., Ltd. | Systems and methods for memory architecture |
| US10635633B2 (en) | 2015-06-01 | 2020-04-28 | SZ DJI Technology Co., Ltd. | Systems and methods for memory architecture |
| CN105243112A (en) * | 2015-09-25 | 2016-01-13 | 天脉聚源（北京）科技有限公司 | Object indexing control method and apparatus |
| CN109086002A (en) * | 2018-06-28 | 2018-12-25 | 平安科技（深圳）有限公司 | Space management, device, computer installation and the storage medium of storage object |
| WO2020034729A1 (en) * | 2018-08-17 | 2020-02-20 | 华为技术有限公司 | Data processing method, related device, and computer storage medium |
| CN111581206A (en) * | 2019-03-15 | 2020-08-25 | 北京忆芯科技有限公司 | B + tree operation device and method |
| CN111221472A (en) * | 2019-12-26 | 2020-06-02 | 天津中科曙光存储科技有限公司 | Multi-block allocation strategy optimization method and system for disk space allocation |
| CN111221472B (en) * | 2019-12-26 | 2023-08-25 | 天津中科曙光存储科技有限公司 | Multi-block allocation strategy optimization method and system for disk space allocation |
| CN112667637A (en) * | 2020-12-31 | 2021-04-16 | 中移（杭州）信息技术有限公司 | Data management method, device and computer readable storage medium |
| CN112667637B (en) * | 2020-12-31 | 2023-09-19 | 中移（杭州）信息技术有限公司 | Data management method, device and computer readable storage medium |
| CN113805816A (en) * | 2021-09-27 | 2021-12-17 | 济南浪潮数据技术有限公司 | Disk space management method, device, equipment and storage medium |
| CN113805816B (en) * | 2021-09-27 | 2024-02-23 | 济南浪潮数据技术有限公司 | Disk space management method, device, equipment and storage medium |
| CN114238343A (en) * | 2021-12-23 | 2022-03-25 | 南京华飞数据技术有限公司 | Implementation method of multidimensional variability automatic data model based on big data |
| CN114238343B (en) * | 2021-12-23 | 2022-10-28 | 南京华飞数据技术有限公司 | Implementation method of multidimensional variability automatic data model based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102915340A (en) | Expanded B+ tree-based object file system | |
| CN110825748B (en) | High-performance and easily-expandable key value storage method by utilizing differentiated indexing mechanism | |
| CN102158546B (en) | Cluster file system and file service method thereof | |
| CN100498740C (en) | Data cache processing method, system and data cache device | |
| CN104850358B (en) | A kind of magneto-optic electricity mixing storage system and its data acquisition and storage method | |
| CN101997918B (en) | Method for allocating mass storage resources according to needs in heterogeneous SAN (Storage Area Network) environment | |
| CN102255962B (en) | Distributive storage method, device and system | |
| CN100424699C (en) | Attribute extensible object file system | |
| CN103838853B (en) | Mixed file system based on different storage media | |
| CN103106286B (en) | Method and device for managing metadata | |
| CN109766312A (en) | A kind of block chain storage method, system, device and computer readable storage medium | |
| CN101488153A (en) | Method for implementing high-capacity flash memory file system in embedded type Linux | |
| CN103186350A (en) | Hybrid storage system and hot spot data block migration method | |
| CN101571869B (en) | File memory and read method of smart card and device thereof | |
| CN103176754A (en) | Reading and storing method for massive amounts of small files | |
| CN103229164B (en) | Data access method and device | |
| CN102541985A (en) | Organization method of client directory cache in distributed file system | |
| CN102521419A (en) | Hierarchical storage realization method and system | |
| CN102521330A (en) | Mirror distributed storage method under desktop virtual environment | |
| WO2021208239A1 (en) | Low-latency file system address space management method and system, and medium | |
| CN102024019A (en) | Suffix tree based catalog organizing method in distributed file system | |
| CN103942301A (en) | Distributed file system oriented to access and application of multiple data types | |
| CN105516313A (en) | Distributed storage system used for big data | |
| CN108052541A (en) | The realization of file system based on multi-level page-table bibliographic structure, access method, terminal | |
| CN103473258A (en) | Cloud storage file system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130206 |