CN102073464A - Method for creating allocation-on-demand incremental volume - Google Patents
Method for creating allocation-on-demand incremental volume Download PDFInfo
- Publication number
- CN102073464A CN102073464A CN 201110007887 CN201110007887A CN102073464A CN 102073464 A CN102073464 A CN 102073464A CN 201110007887 CN201110007887 CN 201110007887 CN 201110007887 A CN201110007887 A CN 201110007887A CN 102073464 A CN102073464 A CN 102073464A
- Authority
- CN
- China
- Prior art keywords
- volume
- real
- space
- data
- user
- 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
Abstract
The invention belongs to the technical field of network storage, and discloses a method for creating an allocation-on-demand incremental volume. The volume has a two-layer volume structure, and realizes separation of a physical space and a logical space. The creating method comprises that: a user state interface participates in a new parameter (N) in a user state command lvcreate of a logical volume manger (LVM). The real volume of the bottom layer is a common logical volume, and the virtual volume of the upper layer is a data structure in a Linux kernel. The physical space of the real volume is allocated according to the actual requirement, so the utilization rate of a disk space is improved. Real-time monitoring of the physical space of the real volume and automatic capacity expansion of the actual space effectively avoid overflow of the spaces. A user can randomly designate capacity expansion increment magnitude as required to perform capacity expansion of the virtual volume. The volume processes the write request by adopting a mode of additional write. Moreover, the volume processes garbage fragments by adopting a quick and effective garbage reclamation algorithm.
Description
Technical field
The invention belongs to the Network storage technology field.
Background technology
Along with the development of infotech, growing, the capacity of disk is increasing to the demand of information storage capacity for enterprise or individual, how rationally to utilize these spaces also to become a hot issue in to the demand of storage space satisfying people.
Studies show that, in present storage system, the utilization factor of capacity is all very low, be stored as example with traditional magnetic disk, consider the continuous growth of user data, in fact the disk size of average 40%--50% never was utilized, and storage resources a large amount of, preciousness has been wasted, and does not bring into play due effect.For satisfying the requirement of aspects such as system performance and later upgrade expanding, the general purchase of client surpasses actual data capacity demand 3-4 disk doubly, is used for additional functions such as disk mirroring, thereby caused the significant wastage of storage space resource simultaneously.The main cause of this present situation is: the distribution method to storage resources is static at present, promptly according to user's estimation (very big surplus is often arranged), in advance storage resources being divided into subregion or volume, is that unit distributes to the user with storage resources with subregion or volume then.In the use afterwards, distribute to have in user's the storage resources and be in idle state greatly for a long time.These idle resources can't be used by other user worse, thereby have caused the low phenomenon of memory capacity utilization factor.
The demand assigned method can improve the utilization factor of disk space undoubtedly, and this method allows system's allocation space dynamically according to demand, and idle space reduces, thereby disk space has obtained reasonably utilizing the waste of having avoided the space.If can adopt dynamic method distribution according to need storage resources, undoubtedly, will greatly improve the utilization factor of storage resources, thereby not increase the capacity that increases storage system under the condition of cost.
Summary of the invention
The objective of the invention is to solve the low problem of current disk space utilization factor, a kind of creation method of distribution according to need increment volume of real space dynamic assignment be provided, and and then can be used as a platform, launch multiple application thereon.
The present invention proposes a kind of distribution according to need increment volume AoDI(Allocation-On-Demand Incremental Volume), this volume adopts the structure of two-layer volume, promptly empty volume and real volume, empty volume size is separated with real volume size, realize demand assigned mechanism, improved the space availability ratio of disk effectively.
The creation method of a kind of volume of distributing according to need provided by the invention comprises:
1st, user's attitude interface of creating of distribution according to need increment volume: at LVM LVM(Logical Volume Management) increased new parameter N in the order lvcreate of user's attitude, what be used to represent to set up is distribution according to need increment volume;
2nd, the process of distribution according to need increment volume establishment: at first create a common logical volume LV(Logical Volume), be real volume with this volume, on this General Logic volume, create an empty volume again, form the structure of two-layer volume.The void volume on upper strata is a data structure in the linux kernel, does not take the actual magnetic disk space.The size of empty volume is the size of user's appointment, and the initial size of real volume is rolled up less than void;
3rd, the automatic dilatation of real volume: in the process that the user uses, along with increasing of amount of user data, the occupancy of physical space improves gradually, when the physical space occupancy surpasses preset threshold, automatically be real volume allocation space, satisfy the more memory requirement of user;
4th, the dilatation of empty volume: in the process that the user uses, the situation that empty volume space size in the original plan can not practical requirement may occur, at this moment, can finish the dilatation of empty volume by the size of any fixed time of user and dilatation;
5th, the magnetic disk of the real volume of distribution according to need increment volume: comprise meta-data region and data field, meta-data region comprises superblock, initialization bitmap, fragment bitmap and Chunkmap;
6th, distribution according to need increment volume data write method: employing is appended the mode of writing and is finished;
7th, rubbish recovering method:
Utilize a variable to serve as " counter ", the value of this variable increases with the increase of gibberish piece, when useless data block reaches default quantity (such as 1000), starts garbage reclamation automatically, the fragment in the elimination system.
The zone of desiring to clear up is provided with two pointers, from the beginning one scan to tail, one scans from the Caudad head, head pointer is responsible for finding first invalid piece, and tail pointer is responsible for finding first effective piece, in case find, the block copy that tail pointer is pointed to is the position of pointed to the end, so repetitive operation when pointer meets end to end, shows that then garbage-cleaning finishes.
The automatic dilatation of described real volume of the 3rd step is specially:
3.1st, each real volume all being provided with a finger daemon monitors its disk space utilization rate;
3.2nd, the disk space utilization rate utilizes the proc interface to obtain, specific practice: create a sub-directory snapproc under/proc catalogue, create file f ull, record space utilization rate then under this sub-directory;
3.3rd, after finishing write operation, the position of active data area tail pointer is the position of freed data blocks head, and the ratio according to this position and space, data field just can calculate the space utilization rate, therefore, behind each write operation, calculate the space utilization rate and be updated to file f ull;
3.4th, when the space utilization rate reached preset threshold value (such as 80%), finger daemon can start the dilatation process, promptly utilized LVM user's attitude order lvextend.
The dilatation of described empty volume of the 4th step is specially:
Adopt the mode of array of indexes, original empty volume and newly-increased part are organized together effectively; Each empty volume space dilatation only needs to increase newly part metadata corresponding district first address and is kept in this array of indexes and gets final product, and can manage the meta-data region of dispersion easily with this.
In the magnetic disk of the real volume described in the 5th step, superblock is responsible for preserving the initial address message (IAM) of device number, snapshot number of times and each snapshot; The initialization bitmap is used for the accelerating initialization process; The fragment bitmap is used for carrying out defragmentation, reclaims the rubbish fragment; Chunkmap preserves the mapping table of station address to amount of physical memory; The data field part is used to store user's data.
The 6th described employing of step is appended the mode of writing and is finished the concrete grammar that data write and be:
The current sector number that is not used of real time record, when carrying out write operation, the base address adds the sector number of record at every turn, is exactly real side-play amount, realizes appending thus writing.
Advantage of the present invention and beneficial effect:
The present invention has realized a kind of distribution according to need increment volume of real space dynamic assignment, this volume has adopted the structure of two-layer volume, comprise empty volume and real volume, realized that empty volume size and real volume size are separated, visible to the user is the void volume on upper strata, and the actual size of the real volume of bottom is a demand assigned, has improved the disk space utilization factor.
User's attitude interface that the present invention creates is to have participated in a new parameter (N) in user's attitude order lvcreate, changes very for a short time, is easy to user's acceptance.Distribution according to need increment volume can be finished the automatic dilatation of monitoring in real time of real volume and real space, and this automatic dilatation technology has effectively been avoided overflowing of space than manual dilatation.The user can carry out the dilatation of empty volume as required, and the size of dilatation increment can be specified by the user, has good dirigibility, is very easy to user's use.
The present invention adopts to append to write and replaces traditional covering writing technology processing write requests to improve the performance of system with this, by postponing the decline of the system performance of metadata replacement policies such as renewal when slowing down the disk metadata access, guaranteed that data can not lose and reduce seek time.Utilize the initialization bitmap to quicken the metadata initialization, and adopted a kind of algorithm of garbage reclamation fast to handle fragment, the removal process of rubbish is finished automatically, can put the fragment in the volume fast in order, clear up rubbish, guaranteed the consistance of data effectively.
Description of drawings
Fig. 1 is that the distribution according to need increment is rolled up two-layer volume structure synoptic diagram;
Fig. 2 is the automatic expansion method synoptic diagram of the real volume of distribution according to need increment volume;
Fig. 3 is the empty volume of a distribution according to need increment volume dilatation synoptic diagram;
Fig. 4 is a distribution according to need increment volume magnetic disk;
Fig. 5 is that distribution according to need increment volume memory-mapped reaches and the mutual synoptic diagram of hard disk;
Fig. 6 is that distribution according to need increment volume appends the write method synoptic diagram;
Fig. 7 is the process flow diagram that distribution according to need increment volume is handled read request;
Fig. 8 is the process flow diagram of distribution according to need increment volume processing write requests.
Embodiment
For ease of understanding above-mentioned purpose of the present invention, feature and advantage, the present invention is described in further detail below in conjunction with the drawings and specific embodiments.
Consulting Fig. 1, is the two-layer volume structure synoptic diagram of distribution according to need increment volume.Specific as follows:
The process in distribution according to need increment scrolling attitude distribute data space is transparent to the user, real volume space take with empty volume space size be not wait.For this reason, a kind of structure of two-layer volume has been proposed, promptly empty volume and real volume.Visible to the user is empty volume, and it is a data structure in the linux kernel, does not take the actual magnetic disk space, is responsible for station address is converted to real volume address.The bottom volume that real volume then is actual preservation metadata and data.
At the initial phase that distribution according to need increment volume is created, the scale of the physical space that real volume distributes is less than the size of void volume.In the process that the user uses, along with increasing of user's data amount, the occupancy of physical space improves gradually.Use too much when physical space, when surpassing a threshold value, should continue as its allocation space, satisfy the more memory requirement of user.The physical space of historical facts or anecdotes volume is variable, is meta-data region and the big or small sum of the data field of having distributed.This process is transparent to the user, and the user needn't be concerned about the actual physical space what are arranged fully, only needs to understand in the aspect of logic the size of empty volume.
Consulting Fig. 2, is the synoptic diagram of the automatic dilatation of real volume.Specific as follows:
Reasonably the distribution incremental process of real volume should be that system finishes automatically on the backstage, just realizes the automatic dilatation of real volume.
For each real volume, have a finger daemon in the system and be responsible for its space operating position is monitored.System can judge in real time whether work as the front space utilization rate reaches threshold value and whether need to start the dilatation process.For reaching user's attitude program and kernel state program high efficiency interactive, the space utilization rate is obtained by/proc interface.Specific practice: set up a proc interface under/proc catalogue, responsible record is the time space utilization rate, and upgrades the space utilization rate behind each write operation.Be convenient management, under/proc catalogue, create a sub-directory snapproc, under this sub-directory, create file f ull, record space utilization rate then.
The value of renewal _ full behind each write operation.Different with conventional block equipment, distribution according to need increment volume appends WriteMode owing to having adopted, and can grasp memory allocation (use) situation naturally.After system finished write operation, the position of active data area tail pointer was the position of freed data blocks head.Ratio according to this position and space, data field just can calculate the space utilization rate.Therefore, behind each write operation, calculate the space utilization rate, its assignment is given _ full, carry out Data Update.During numerical value in reading the full file, call proc_entry and read method, return _ value of full.By this proc interface, finger daemon just can read in real time/ numerical value in the proc/snapproc/full file, grasp is the time space utilization rate.
When the space utilization rate reached preset threshold value (for example 80%), finger daemon can start the dilatation program.The process of System Expansion is this volume of first temporary suspension and starts automatic dilatation process, distributes required newly-increased space for the dilatation volume in the LVM physical space.If the current free space of LVM can satisfy the dilatation demand, LVM volume metadata is then upgraded in the allocation of space success, and the new data space of notice kernel distributes.Automatically after dilatation finishes,, continue executable operations with roll recovery.The read-write of this mode is not to read and write disk veritably, and is the read-write operation in the internal memory, so almost without any overhead.System can read the numerical value of this document in real time, judges whether to start automatic dilatation with this.
Consulting Fig. 3, is the synoptic diagram of empty volume dilatation.Specific as follows:
In the process that the user uses, the situation that space size in the original plan can not satisfy actual demand may appear, need the size of the empty volume space of change.The size of meta-data region is linear with void volume size.When void involved in capable dilatation, the metadata in newly-increased space needed the corresponding real space of system assignment to preserve.
As Fig. 3, empty volume L is former distribution, and meta-data region M is the meta-data region in L corresponding data district, and data field D is the data field of having distributed before the dilatation of empty volume.Empty volume L ' is the newly-increased part of empty volume space, and its size is specified by the user.Introduction by the front knows, the size of newly-increased meta-data region should be the product of the ratio of the spatial value of L ' and the shared gross space of meta-data region, so just obtains the size of meta-data region M '.When the dilatation of void volume, the dilatation increment of the real volume of bottom should be the spatial value of M '.After the dilatation of empty volume, the meta-data region of real volume just is made up of meta-data region M and M '.
After the dilatation of empty volume, be the meta-data region of these dispersions of unified management, the address of all meta-data region will be kept in the index of metadata array.The index of metadata array is kept in the superblock of meta-data region.When the I/O to real volume operates, just can determine the position of real volume by this array of indexes.This method realization is very simple and system overhead is very little.
Consulting Fig. 4, is the synoptic diagram of the magnetic disk of distribution according to need increment volume.Specific as follows:
The magnetic disk of the real volume of the bottom of distribution according to need increment volume totally can be divided into superblock, meta data block and data field.Meta-data region comprises initialization bitmap, fragment bitmap and Chunkmap.Superblock is responsible for preserving the global information of volumes such as the start address of device number, snapshot number of times and each snapshot.Chunkmap preserves the mapping table of station address to amount of physical memory.The initialization bitmap is used for the accelerating initialization process.The fragment bitmap is used for carrying out defragmentation exactly, recovery rubbish fragment.The data field part is used to store user's data.A piece (Chunk) is the continuous hard drive space of a slice, and size is 4KB.System with piece as storage unit.
Superblock (Super Chunk) is responsible for the global information of record distribution according to need increment volume, comprises the device number, array of indexes of volume etc.
Meta data block (Metadata Chunk) comprises initialization bitmap, fragment bitmap and Chunkmap.
When creating volume, Chunkmap all need be set to illegal physical address, all be initialized as-1, represent that all station addresss all do not write as yet.If the volume logic size is very huge, then the scale of mapping table also can be very big, and initialization procedure will be very consuming time so.For avoiding excessive system overhead, used the initialization bitmap.The continuous zone of a slice among the corresponding Chunkmap of in the bitmap each.Disk Chunkmap is not filled in during initialization, but be 0 to represent Chunkmap respective regions no initializtion still with corresponding position in the bitmap.Like this, when visit during this part metadata, just needn't reading disk, need think directly that it is illegal address (data of respective user address does not write as yet).When this part metadata is updated to disk, be 1 with correspondence position in the initialization bitmap.Obviously, the initialized method of this postponement has improved initialization speed effectively.
Write but not cover WriteMode owing to adopt to append, then may there be a plurality of versions in station address data in system, thereby cause rubbish internal memory and fragment problems.The fragment bitmap is used for carrying out defragmentation exactly, recovery rubbish fragment.Whether a data block of each the corresponding data district part in the bitmap, identifying this data block is the rubbish internal memory.
The mapping table of piece is stored among the Chunkmap.
Consulting Fig. 5, is distribution according to need increment volume memory-mapped and and the mutual synoptic diagram of hard disk.Specific as follows:
Tissue in the internal memory has used salted hash Salted, and each Hash bucket is safeguarded the chained list of a metadata page or leaf, and adopts the replacement policy of first in first out.When Cache does not hit, read the item that comprises requested page in the mapping table.The metadata of these pages is dirty, and these pages or leaves are replaced.The page or leaf accumulation that these are dirty writes in the bigger write request.Be redirected to the block device layer after handling bio request from the common apparatus layer.At last, check the initialization bitmap, then corresponding metadata on dirty metadata in the exchange memory and the disk.
Consulting Fig. 6, is that distribution according to need increment volume appends the write method synoptic diagram, specific as follows:
For seek time can not lost and reduce to the assurance data, the employing of distribution according to need increment volume is appended writing technology and has been replaced traditional covering to write.Specific algorithm is the current sector number that is not used of real time record, and when carrying out write operation, the base address adds the sector number of record at every turn, is exactly real side-play amount.Can realize appending thus writing.
Consult Fig. 7, be the treatment scheme synoptic diagram that distribution according to need increment volume is read method, specific as follows:
The specific implementation method of reading algorithm is according to the command lookup metadata, replaces if metadata not in internal memory, is then carried out metadata.If in internal memory, then go reading corresponding data according to metadata.
Consult Fig. 8, be the treatment scheme synoptic diagram of distribution according to need increment volume write method, specific as follows:
The specific implementation method of writing algorithm is at first to search metadata according to the request that issues, and replaces if metadata not in internal memory, is then carried out metadata.After metadata is replaced and finished, check whether position corresponding in the initialization bitmap is-1.If not-1, illustrate that correspondence position had write operation before this, so Dui Ying data then show it has been legacy version, and it is masked as fragment, are 1 with correspondence position in the fragment bitmap just.Append write operation at last.
Claims (7)
1. the creation method of increment volume of distributing according to need is characterized in that described distribution according to need increment volume adopts the structure of two-layer volume, promptly empty volume and real volume, and empty volume size is separated with real volume size, has realized demand assigned mechanism, and concrete creation method comprises:
1st, user's attitude interface of creating of distribution according to need increment volume: increase new parameter N in the order lvcreate of LVM LVM user attitude, what be used to represent to set up is distribution according to need increment volume;
2nd, the process of distribution according to need increment volume establishment: at first create a common logical volume LV, be real volume, on this General Logic volume, create an empty volume again, form the structure of two-layer volume; The void volume on upper strata is a data structure in the linux kernel, does not take the actual magnetic disk space; The size of empty volume is the size of user's appointment, and the initial size of real volume is rolled up less than void;
3rd, the automatic dilatation of real volume: in the process that the user uses, along with increasing of amount of user data, the occupancy of physical space improves gradually, when the physical space occupancy surpasses preset threshold, automatically be real volume allocation space, satisfy the more memory requirement of user;
4th, the dilatation of empty volume: in the process that the user uses, the situation that empty volume space size in the original plan can not practical requirement may occur, at this moment, can finish the dilatation of empty volume by the size of any fixed time of user and dilatation;
5th, the magnetic disk of the real volume of distribution according to need increment volume: comprise meta-data region and data field, meta-data region comprises superblock, initialization bitmap, fragment bitmap and Chunkmap;
6th, distribution according to need increment volume data write method: employing is appended the mode of writing and is finished.
2. method according to claim 1 is characterized in that the automatic dilatation of described real volume of the 3rd step is specially:
3.1st, each real volume all being provided with a finger daemon monitors its disk space utilization rate;
3.2nd, the disk space utilization rate utilizes the proc interface to obtain, specific practice: create a sub-directory snapproc under/proc catalogue, create file f ull, record space utilization rate then under this sub-directory;
3.3rd, after finishing write operation, the position of active data area tail pointer is the position of freed data blocks head, and the ratio according to this position and space, data field just can calculate the space utilization rate, therefore, behind each write operation, calculate the space utilization rate and be updated to file f ull;
3.4th, when the space utilization rate reaches preset threshold value, finger daemon can start the dilatation process, promptly utilizes LVM user's attitude order lvextend.
3. method according to claim 1 is characterized in that the dilatation of described empty volume of the 4th step is specially:
Adopt the mode of array of indexes, original empty volume and newly-increased part are organized together effectively; Each empty volume space dilatation only needs to increase newly part metadata corresponding district first address and is kept in this array of indexes and gets final product, and can manage the meta-data region of dispersion easily with this.
4. method according to claim 1 is characterized in that superblock is responsible for preserving the initial address message (IAM) of device number, snapshot number of times and each snapshot in the magnetic disk of the real volume described in the 5th step; The initialization bitmap is used for the accelerating initialization process; The fragment bitmap is used for carrying out defragmentation, reclaims the rubbish fragment; Chunkmap preserves the mapping table of station address to amount of physical memory; The data field part is used to store user's data.
5. method according to claim 1 is characterized in that the 6th described employing of step appends the mode of writing and finish the concrete grammar that data write and be:
The current sector number that is not used of real time record, when carrying out write operation, the base address adds the sector number of record at every turn, is exactly real side-play amount, realizes appending thus writing.
6. method according to claim 1 is characterized in that and can also involve in capable garbage reclamation to the distribution according to need increment, and detailed process is:
Utilize a variable to serve as " counter ", the value of this variable increases with the increase of gibberish piece, when useless data block reaches default quantity, starts garbage reclamation automatically, the fragment in the elimination system.
7. method according to claim 6 is characterized in that the concrete grammar of garbage reclamation is:
The zone of desiring to clear up is provided with two pointers, from the beginning a scanning backward, one scans before Caudad, head pointer is responsible for finding first invalid piece, and tail pointer is responsible for finding first effective piece, in case find, the block copy that tail pointer is pointed to is the position of pointed to the end, so repetitive operation when pointer meets end to end, shows that then garbage-cleaning finishes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110007887 CN102073464A (en) | 2011-01-14 | 2011-01-14 | Method for creating allocation-on-demand incremental volume |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110007887 CN102073464A (en) | 2011-01-14 | 2011-01-14 | Method for creating allocation-on-demand incremental volume |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102073464A true CN102073464A (en) | 2011-05-25 |
Family
ID=44032017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110007887 Pending CN102073464A (en) | 2011-01-14 | 2011-01-14 | Method for creating allocation-on-demand incremental volume |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102073464A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102955793A (en) * | 2011-08-23 | 2013-03-06 | 腾讯科技（深圳）有限公司 | Method and device for processing file carrier space |
| CN103778064A (en) * | 2012-10-23 | 2014-05-07 | 阿里巴巴集团控股有限公司 | Data management method and device |
| WO2015058493A1 (en) * | 2013-10-23 | 2015-04-30 | 华为技术有限公司 | Storage data processing method and device |
| CN105262680A (en) * | 2015-10-21 | 2016-01-20 | 浪潮(北京)电子信息产业有限公司 | Multi-threaded NAS Gateway applied to cloud storage system |
| CN105468989A (en) * | 2014-09-04 | 2016-04-06 | 中国石油化工股份有限公司 | Cloud storage quota management method based on Linux kernel monitoring |
| CN105760553A (en) * | 2016-03-25 | 2016-07-13 | 北京奇虎科技有限公司 | Data management method and device |
| CN106168885A (en) * | 2016-07-18 | 2016-11-30 | 浪潮(北京)电子信息产业有限公司 | A kind of method and system of logical volume dynamic capacity-expanding based on LVM |
| CN106227472A (en) * | 2016-08-19 | 2016-12-14 | 浪潮(北京)电子信息产业有限公司 | A kind of from simplification volume expansion method and system |
| CN106528003A (en) * | 2016-12-09 | 2017-03-22 | 曙光信息产业(北京)有限公司 | Method and device for rebuilding disk array |
| CN107239412A (en) * | 2017-06-19 | 2017-10-10 | 杭州宏杉科技股份有限公司 | Memory space collocation method, method for writing data and storage device based on Thin LUN |
| CN107544864A (en) * | 2016-06-24 | 2018-01-05 | 深圳市深信服电子科技有限公司 | A kind of virtual-machine data copy method and virtual-machine data copy system |
| CN109739688A (en) * | 2018-12-18 | 2019-05-10 | 杭州宏杉科技股份有限公司 | Snapshot Resources space management, device, electronic equipment |
| CN109960463A (en) * | 2017-12-26 | 2019-07-02 | 中兴通讯股份有限公司 | A kind of disk management method, device and storage medium |
| CN112181309A (en) * | 2020-10-14 | 2021-01-05 | 上海德拓信息技术股份有限公司 | Online capacity expansion method for mass object storage |
| CN112463077A (en) * | 2020-12-16 | 2021-03-09 | 北京云宽志业网络技术有限公司 | Data block processing method, device, equipment and storage medium |
| CN115033378A (en) * | 2022-06-06 | 2022-09-09 | 北京熵核科技有限公司 | Management method for volatile memory |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1766845A (en) * | 2005-11-30 | 2006-05-03 | 吴晓栋 | Method for realizing high security and recoverable file system |
| CN101320351A (en) * | 2008-06-27 | 2008-12-10 | 华中科技大学 | Internal memory distribution, cleaning and releasing method, and internal memory management apparatus |
-
2011
- 2011-01-14 CN CN 201110007887 patent/CN102073464A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1766845A (en) * | 2005-11-30 | 2006-05-03 | 吴晓栋 | Method for realizing high security and recoverable file system |
| CN101320351A (en) * | 2008-06-27 | 2008-12-10 | 华中科技大学 | Internal memory distribution, cleaning and releasing method, and internal memory management apparatus |
Non-Patent Citations (2)
| Title |
|---|
| 《华南金融电脑》 20081110 覃振杰 逻辑卷管理 , 第11期 2 * |
| 《计算机工程》 20080331 韦理，周悦芝，夏楠 用于网络存储系统的存储空间动态分配方法 第34卷, 第5期 2 * |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102955793A (en) * | 2011-08-23 | 2013-03-06 | 腾讯科技（深圳）有限公司 | Method and device for processing file carrier space |
| CN102955793B (en) * | 2011-08-23 | 2016-08-03 | 腾讯科技（深圳）有限公司 | The method and apparatus processing file carrier space |
| CN103778064A (en) * | 2012-10-23 | 2014-05-07 | 阿里巴巴集团控股有限公司 | Data management method and device |
| CN103778064B (en) * | 2012-10-23 | 2017-10-27 | 阿里巴巴集团控股有限公司 | A kind of data managing method and device |
| WO2015058493A1 (en) * | 2013-10-23 | 2015-04-30 | 华为技术有限公司 | Storage data processing method and device |
| CN105468989B (en) * | 2014-09-04 | 2018-10-12 | 中国石油化工股份有限公司 | Cloud storage quota management method based on linux kernel monitoring |
| CN105468989A (en) * | 2014-09-04 | 2016-04-06 | 中国石油化工股份有限公司 | Cloud storage quota management method based on Linux kernel monitoring |
| CN105262680A (en) * | 2015-10-21 | 2016-01-20 | 浪潮(北京)电子信息产业有限公司 | Multi-threaded NAS Gateway applied to cloud storage system |
| CN105760553A (en) * | 2016-03-25 | 2016-07-13 | 北京奇虎科技有限公司 | Data management method and device |
| CN107544864A (en) * | 2016-06-24 | 2018-01-05 | 深圳市深信服电子科技有限公司 | A kind of virtual-machine data copy method and virtual-machine data copy system |
| CN107544864B (en) * | 2016-06-24 | 2021-08-13 | 深信服科技股份有限公司 | Virtual machine data copying method and virtual machine data copying system |
| CN106168885A (en) * | 2016-07-18 | 2016-11-30 | 浪潮(北京)电子信息产业有限公司 | A kind of method and system of logical volume dynamic capacity-expanding based on LVM |
| CN106168885B (en) * | 2016-07-18 | 2019-09-24 | 浪潮(北京)电子信息产业有限公司 | A kind of method and system of the logical volume dynamic capacity-expanding based on LVM |
| CN106227472A (en) * | 2016-08-19 | 2016-12-14 | 浪潮(北京)电子信息产业有限公司 | A kind of from simplification volume expansion method and system |
| CN106528003A (en) * | 2016-12-09 | 2017-03-22 | 曙光信息产业(北京)有限公司 | Method and device for rebuilding disk array |
| CN107239412A (en) * | 2017-06-19 | 2017-10-10 | 杭州宏杉科技股份有限公司 | Memory space collocation method, method for writing data and storage device based on Thin LUN |
| CN109960463A (en) * | 2017-12-26 | 2019-07-02 | 中兴通讯股份有限公司 | A kind of disk management method, device and storage medium |
| CN109739688B (en) * | 2018-12-18 | 2021-01-26 | 杭州宏杉科技股份有限公司 | Snapshot resource space management method and device and electronic equipment |
| CN109739688A (en) * | 2018-12-18 | 2019-05-10 | 杭州宏杉科技股份有限公司 | Snapshot Resources space management, device, electronic equipment |
| CN112181309A (en) * | 2020-10-14 | 2021-01-05 | 上海德拓信息技术股份有限公司 | Online capacity expansion method for mass object storage |
| CN112463077A (en) * | 2020-12-16 | 2021-03-09 | 北京云宽志业网络技术有限公司 | Data block processing method, device, equipment and storage medium |
| CN115033378A (en) * | 2022-06-06 | 2022-09-09 | 北京熵核科技有限公司 | Management method for volatile memory |
| CN115033378B (en) * | 2022-06-06 | 2023-01-06 | 北京熵核科技有限公司 | Management method for volatile memory |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102073464A (en) | Method for creating allocation-on-demand incremental volume | |
| TWI709073B (en) | Distributed storage system, distributed storage method and distributed facility | |
| JP6304406B2 (en) | Storage apparatus, program, and information processing method | |
| CN101556557B (en) | Object file organization method based on object storage device | |
| US20180276233A1 (en) | Techniques for implementing hybrid flash/hdd-based virtual disk files | |
| US10359952B1 (en) | Method and system for implementing writable snapshots in a virtualized storage environment | |
| CN104317742B (en) | Automatic thin-provisioning method for optimizing space management | |
| US8639880B2 (en) | Hierarchical storage management for database systems | |
| US10170151B2 (en) | Method and system for handling random access write requests for a shingled magnetic recording hard disk drive | |
| CN102667704B (en) | Mixing storage subsystem and management method thereof | |
| US20170212891A1 (en) | Scheduling technique to smooth metadata processing latency | |
| CN103761053B (en) | A kind of data processing method and device | |
| CN109344090B (en) | Virtual hard disk system of KVM virtual machine in data center and data center | |
| US9965381B1 (en) | Indentifying data for placement in a storage system | |
| US20230046216A1 (en) | Data management system and method of controlling | |
| CN101488153A (en) | Method for implementing high-capacity flash memory file system in embedded type Linux | |
| CN103530387A (en) | Improved method aimed at small files of HDFS | |
| CN103544045A (en) | HDFS-based virtual machine image storage system and construction method thereof | |
| US20150006787A1 (en) | Techniques for dynamically relocating virtual disk file blocks between flash storage and hdd-based storage | |
| JP2006293981A (en) | Database storing method, and database storing system | |
| KR101017067B1 (en) | Locality-Aware Garbage Collection Technique for NAND Flash Memory-Based Storage Systems | |
| US20100057755A1 (en) | File system with flexible inode structures | |
| CN103020077A (en) | Method for managing memory of real-time database of power system | |
| US7424574B1 (en) | Method and apparatus for dynamic striping | |
| US9697216B2 (en) | Method and apparatus for differentiated data placement |
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: 20110525 |