CN101820445B - Distribution method for two-dimensional tiles in object-based storage system - Google Patents
Distribution method for two-dimensional tiles in object-based storage system Download PDFInfo
- Publication number
- CN101820445B CN101820445B CN2010101323740A CN201010132374A CN101820445B CN 101820445 B CN101820445 B CN 101820445B CN 2010101323740 A CN2010101323740 A CN 2010101323740A CN 201010132374 A CN201010132374 A CN 201010132374A CN 101820445 B CN101820445 B CN 101820445B
- Authority
- CN
- China
- Prior art keywords
- dimensional tiles
- data
- storage system
- tiles
- virtual 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.)
- Expired - Fee Related
Links
Images
Abstract
The invention discloses a distribution method for two-dimensional tiles in an object-based storage system. The method is characterized by comprising the following steps of: firstly, constructing a group of virtual storage devices V= {V0, V1,...,VM-1} with totally same performance and unlimited capacities; secondly, disturbing the two-dimensional tiles at the corresponding positions in a query window to M virtual storage devices; and finally, mapping the two-dimensional tiles in real storage device intervals onto the corresponding real storage devices. The method has the advantages of optimizing the distribution balancing performance of the two-dimensional tiles in the object-based storage system, reducing the average response time of the tiles, laying a firm topographic data storage foundation for popularization and application of a large-scale topographic real-time roaming system, and facilitating applying topographic data in each aspect of human life.
Description
Technical field
The present invention relates to the location mode in a kind of storage system, relate in particular to a kind of two-dimensional tiles based on the location mode in the object storage system.
Background technology
(like the digital earth, extensive Military Simulation etc.) have obtained extensive use in a plurality of fields in large-scale terrain real time roaming system, and the network user of terrain data sharply increases; Along with improving constantly of terrain data resolution, the data volume of terrain data also is explosive increase.Current Network storage technology mainly contains three kinds of direct attached storage, network attached storage and storage area networks, the storage solution that they provide for the massive terrains data kind to satisfy simultaneously high performance stores service and cross-platform safety to share.Based on object storage technology is follow-on distributed network memory technology, has high-performance, enhanced scalability, high concurrency, ease of manageability and safe and reliable characteristics.Terrain data is divided into several tiles by planar grid, each tile with the stored in form of object in based on object storage system, the requirement of can the storage administration of satisfying magnanimity terrain data high-performance stores service and cross-platform safety being shared.The landform roaming system generally need extract the terrain data of a plurality of tiles adjacent in the query window; It is different from the object storage equipment to take certain location mode that tile in the query window is distributed to as much as possible; Making full use of the high concurrency based on object storage system, is to reduce the key that the tile average response time improves systematic function.
Based on object storage is a kind of distributed network memory technology, has the concurrent access characteristics of height.Data mainly contain two types in a plurality of methods that distribute on based on object storage equipment: one type of method that is based on mapping table, the another kind of method that is based on mapping function.The positional information of coming record data through the metadata catalog mapping table based on the method for mapping table is like the mapping of order burst or the Hash mode of common employing.The lack of uniformity that these class methods cause data to distribute on the one hand, and the difference of performance between the consideration equipment seldom, it is unbalanced more to have aggravated usage ratio of equipment; On the other hand, along with the expansion of system, adopt the required time and the space cost in position of mapping table record data object distribution to make metadata search the bottleneck that performance is called systematic function.Method based on mapping function is directly to calculate the positional information of data object at storage customer end adopted mapping function.Though the general more complicated of mapping function can solve based on the data skewness weighing apparatus property and the metadata of the method for mapping table and search problems such as performance bottleneck, is just becoming the focus of present large memory system data distribution method research.
In the landform roaming system, all tiles all meet with a response and just can obtain the integral terrain data in the query window, and its response time is depended on the longest tile of response time in the query window.This just requires all the tile response times in the query window identical as far as possible.For this reason, each tile in the query window should be distributed to identical mutually different from the object storage equipment of performance, existingly can not reach this requirement fully based on object storage object distribution method.
Summary of the invention
The object of the present invention is to provide a kind of two-dimensional tiles based on the location mode in the object storage system, this method reduces the average response time of tile in query window, improves landform roaming system terrain data access performance.
The present invention realizes like this, it is characterized in that method step is: at first, make up one group of identical virtual memory facilities V={V of performance
0, V
1..., V
M-1, and capacity is not limit; Terrain data is divided into several two-dimensional tiles, the two-dimensional tiles on the correspondence position in the query window is distributed in M the virtual memory facilities, the data in M virtual memory facilities evenly be mapped to successively [0,1) in M equally spaced interval; Then, according to N actual stored equipment R={R
0, R
1..., R
N-1The data volume that is distributed accounts for the percentage Ω={ ω of total amount of data
0, ω
1..., ω
N-1Successively between dividing regions [0,1); At last, the two-dimensional tiles that drops between each actual stored battery limits is mapped on the corresponding actual stored equipment.
Described terrain data is divided into N by level and vertical gridlines
1* N
2Individual tile, be designated as (tx, ty), 0≤tx≤N wherein
1-1,0≤ty≤N
2-1, N
1And N
2Be integer.
Described M is a positive integer, is the number of two-dimensional tiles in the query window.
Described N is a positive integer, is the number of actual stored equipment.
DSMC is adopted in described even mapping.
The data volume that said actual stored equipment is distributed accounts for the percentage Ω={ ω of total amount of data
0, ω
1..., ω
N1Be to go out the data volume A={a that each actual stored equipment is distributed according to its actual Performance Calculation
0, a
1..., a
N-1, then according to formula
Calculate.
Said virtual memory facilities just carries out existing in the process at algorithm, and after two-layer mapping finished, virtual memory facilities had not just existed.
Advantage of the present invention is: optimize two-dimensional tiles based on the distributing equilibrium performance in the object storage system, reduced the average response time of tile; To establish solid terrain data storage basis for the popularization and application of large-scale terrain real time roaming system, and use terrain data for human lives's aspect fully and bring huge convenience.
Description of drawings
Fig. 1 is mapped to the sketch map of actual stored equipment from virtual memory facilities for the two-dimensional tiles in the query window of the present invention.
Embodiment
Referring to 1 being query window among accompanying drawing 1, the figure, 2 is virtual memory facilities, and 3 is mapping relations, and 4 is actual stored equipment, and two-dimensional tiles provided by the invention based on the location mode implementation method in the object storage system is being:
(1) mapping from the query window to the virtual memory facilities is designated as f
1
Suppose the pairing distributed area of virtual memory facilities i for [i/M, (i+1)/M), 0≤i≤M-1 wherein.Earlier with the interval pseudo random number of all tiles (tx, ty) be mapped to [0,1), and then through conversion, with the two-dimensional tiles vector base to the element L in the distribution template
iAll corresponding tiles are mapped to the pairing interval of virtual memory facilities i respectively.Note f
0(tx is the function of pseudo random number that tile (tx, ty) produce and to be distributed in [0,1) is interval ty), then shines upon f
1Can be expressed as
f
1:
if
then
(tx,ty)→V
i
(2) mapping from virtual memory facilities to actual stored equipment is designated as f
2
According to mapping f
1Building method, mapping f
2Can be expressed as
f
2:
if
then
(tx,ty)→R
j
Wherein, 0≤i≤M-1,0≤j≤N-1, ω
1=0.
Referring to Fig. 1, the basic step of location mode provided by the invention is:
(a) terrain data is divided into N according to planar grid
1* N
2Individual tile, be designated as (tx, ty), 0≤tx≤N wherein
1-1,0≤ty≤N
2-1, N
1And N
2Be integer.
(b) with all two-dimensional tiles, through function f
0(tx, ty) be mapped to that consistent pseudorandom produces be evenly distributed to the unit interval [0,1) on set of random numbers;
(c) M that structural behavior is identical, capacity is not limit the set V={V that virtual memory facilities is formed
0, V
1..., V
M-1, corresponding successively unit interval [0,1) on M the distributed area of equidistantly dividing [i/M, (i+1)/M) (0≤i≤M-1);
(d) all two-dimensional tiles on the correspondence position in the query window are mapped to the pairing interval of virtual memory facilities i respectively;
(e) to the set R={R of the composition of N actual stored equipment
0, R
1..., R
N-1, go out the data volume A={a that each actual stored equipment is distributed by its actual Performance Calculation
0, a
1..., a
N-1, and by formula
Calculate the data that each actual stored equipment is distributed
Amount accounts for the distribution percentage Ω={ ω of total amount of data
0, ω
1..., ω
N-1;
(f) according to distribution percentage successively with the unit interval [0,1) be divided into N distributed area
) (0≤j≤N-1, ω
1=0);
Claims (5)
1. a two-dimensional tiles is characterized in that method step is: at first, make up one group of identical virtual memory facilities V={V of performance based on the location mode in the object storage system
0, V
1..., V
M-1, and capacity is not limit; Terrain data is divided into several two-dimensional tiles, the two-dimensional tiles on the correspondence position in the query window is distributed in M the virtual memory facilities, the data in M virtual memory facilities evenly be mapped to successively [0,1) in M equally spaced interval; Then, according to N actual stored equipment R={R
0, R
1..., R
N-1The data volume that is distributed accounts for the percentage Ω={ ω of total amount of data
0, ω
1..., ω
N-1Successively between dividing regions [0,1); At last, the two-dimensional tiles that drops between each actual stored battery limits is mapped on the corresponding actual stored equipment; Described M is a positive integer, is the number of two-dimensional tiles in the query window; Described N is a positive integer, is the number of actual stored equipment.
2. a kind of two-dimensional tiles according to claim 1 is characterized in that based on the location mode in the object storage system described terrain data is divided into N by level and vertical gridlines
1* N
2Individual tile, be designated as (tx, ty), 0≤tx≤N wherein
1-1,0≤ty≤N
2-1, N
1And N
2Be integer.
3. a kind of two-dimensional tiles according to claim 1 is characterized in that described even mapping employing DSMC based on the location mode in the object storage system.
4. a kind of two-dimensional tiles according to claim 1 is based on the location mode in the object storage system, it is characterized in that data volume that said actual stored equipment is distributed accounts for the percentage Ω={ ω of total amount of data
0, ω
1..., ω
N-1Be to go out the data volume that each actual stored equipment is distributed according to its actual Performance Calculation
5. a kind of two-dimensional tiles according to claim 1 is characterized in that based on the location mode in the object storage system said virtual memory facilities just carries out existing in the process at algorithm, and after two-layer mapping finished, virtual memory facilities had not just existed.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010101323740A CN101820445B (en) | 2010-03-25 | 2010-03-25 | Distribution method for two-dimensional tiles in object-based storage system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010101323740A CN101820445B (en) | 2010-03-25 | 2010-03-25 | Distribution method for two-dimensional tiles in object-based storage system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101820445A CN101820445A (en) | 2010-09-01 |
CN101820445B true CN101820445B (en) | 2012-09-05 |
Family
ID=42655389
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2010101323740A Expired - Fee Related CN101820445B (en) | 2010-03-25 | 2010-03-25 | Distribution method for two-dimensional tiles in object-based storage system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101820445B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107016924B (en) | 2016-12-20 | 2020-04-07 | 阿里巴巴集团控股有限公司 | Tile map generation method, updating method and updating device in virtual map |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1971562A (en) * | 2006-11-29 | 2007-05-30 | 华中科技大学 | Distributing method of object faced to object storage system |
CN101040282A (en) * | 2004-08-13 | 2007-09-19 | 潘纳萨斯公司 | Distributed object-based storage system that stores virtualization maps in object attributes |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8285757B2 (en) * | 2007-01-31 | 2012-10-09 | Agency For Science, Technology And Research | File system for a storage device, methods of allocating storage, searching data and optimising performance of a storage device file system |
-
2010
- 2010-03-25 CN CN2010101323740A patent/CN101820445B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101040282A (en) * | 2004-08-13 | 2007-09-19 | 潘纳萨斯公司 | Distributed object-based storage system that stores virtualization maps in object attributes |
CN1971562A (en) * | 2006-11-29 | 2007-05-30 | 华中科技大学 | Distributing method of object faced to object storage system |
Also Published As
Publication number | Publication date |
---|---|
CN101820445A (en) | 2010-09-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104252489B (en) | A method of position character description information is quickly obtained according to longitude and latitude data | |
CN101963996B (en) | Global raster data storing method | |
CN103955511B (en) | Cloud platform data organization and retrieval method for 3D (three-dimensional) urban building data | |
CN102214228B (en) | Large-grid distributed real-time database system and data management method thereof | |
CN110213304B (en) | Block chain network interconnection method and system | |
CN101976468B (en) | Method and system for visualizing multiresolution dynamic landform | |
US20150227553A1 (en) | Method for generating a dataset structure for location-based services and method and system for providing location-based services to a mobile device | |
CN103714145A (en) | Relational and Key-Value type database spatial data index method | |
CN105913347A (en) | House coding method orienting house information acquisition | |
CN102999585A (en) | Geographical-position related hash virtual geocoding method and system | |
CN103838770A (en) | Logic data partition method and system | |
CN105427190B (en) | Calculation method for ground three-dimensional power frequency electric field below UHVAC power transmission line in complex terrain | |
CN104935462A (en) | Deploying method and device for terminal servers | |
CN102495888B (en) | Parallel digital terrain analysis oriented data splitting and distributing method | |
CN108268614A (en) | A kind of distribution management method of forest reserves spatial data | |
CN103559209B (en) | A kind of efficient spatial K-NN search method that Voronoi Diagram is combined with virtual grid | |
CN101964157B (en) | Sphere rhombic grid recursive subdivision method | |
CN105741340B (en) | A kind of transmission line of electricity three-dimensional scenic emulation mode and system for web page display | |
CN102223709A (en) | Method for acquiring specific reference Gaussian plane rectangular coordinate based on continuous operational reference system (CORS) in real time | |
CN101820445B (en) | Distribution method for two-dimensional tiles in object-based storage system | |
CN101014044A (en) | Network GIS system and data transmitting method thereof | |
CN102103600B (en) | Map building method and map system | |
CN107257356A (en) | A kind of social user data optimization laying method based on hypergraph partitioning | |
CN102054009A (en) | Distributed heterogeneous raster spatial data seamless integration method | |
CN103514017B (en) | A kind of three-dimensional data loading method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
C17 | Cessation of patent right | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120905 Termination date: 20140325 |