RESEARCH ON SPATIAL DATABASE DESIGN AND TUNING BASED
ON ORACLE AND ARCSDE
Li Yingcheng, Li Ling
Chinese Academy of Surveying and Mapping, Beitaiping Road 16, Beijing, 100039, casmrsec@sina.com
Commission VI, WG VI/5
KEY WORDS: Database, Block, Configuration, GIS, Image
ABSTRACT:
With the development of the Database technology and Geographic Information System technology, the spatial data managing mode
go through three stages, entirely File system management, File system and Relation Database System mixed management and
entirely Database management. At the same time database model also develop from Relation Model, Object-Relation Model, to
Object-oriented Model. At the present time there are two main spatial database management approaches using commercial database.
One is using Spatial Data Engine (Such as ArcSDE) in the Relation Database Management System. The other is using
Object-Relation Database. Whichever spatial database management approach is used, in practical project we need a set of technics
flow about the data preparing, database design, database configuration and tuning. Especially in commercial database, the rational
database design, configuration and tuning are more important.
In this article the experiment construct a fundamental spatial database system for Tibet province using Geodatabase model of the
ArcSDE and the Oracle database management System. Concretely, the main fields are researched as following: (1) Analyse the
character of fundamental spatial data including the Digital Orthophoto Map data, the Digital Elevation Modal data, the Digital Raster
Graphic data and the Digital line Graphic data. And scheme out a data processing plan for spatial database construction. (2)
Rationally Design Database, this process includes the design for Geodatabase model and the design for Oracle database. (3)
Configure and tune Oracle database, this process includes the configuration of the table and index of the spatial object and the
configuration and tuning of the database initial parameters.
1. INTRODUCTION
With the development of the RS (Remote Sensing), DP (Digital
Photogrammetry) and GPS (Global Positioning System), the
National Spatial Data Infrastructure constructing, the Digital
Geo-Spatial Data Framework (spatial fundamental data, such as
Digital Orthophoto Map, Digital Elevation Modal, Traffic data,
Water System data and Administration Boundary data etc.)
brought forward, the amount of available spatial data, especially
the most fundamental and public datasets for researching and
observing the earth is expanding continuously and the kinds of
available spatial data are increasing frequently. Such a large
amount of spatial data and their great demands make it is no
time to delay to store and manage the spatial data efficiently,
reasonably and safely. At present, the popular data management
approach based on database system is applying to spatial data
iianarciient whoever, the. Geographic Information. software
produ dix Remote Sensing software producer and the
Database software producer, provides his own storage and
manage spatial data approach.
Database model develop from Relation Model, Object-Relation
Model, to Object-oriented Model. And the main large
commercial database manage systems, such as Oracle, DB2,
Informix, SQL Server, Sybase etc., extend their capabilities for
spatial data base on a certain database model. At present there
are two main spatial database management approaches using
commercial database. One is creating a new Spatial Data Engine,
Such as ArcSDE, based on the Relation Database Manager
System. It uses the internal large object data type, such as Long
Raw or Blob of oracle database, to storage the spatial object.
The other is directly using the object data type of the
Object-Relation Database, such as the SDO GEOMETRY
object of oracle. Whichever spatial database management
370
approach is used, in practical project we need a set of technics
flow about the data preparing, database design, database
configuration and tuning. Especially in commercial database,
the rational database design, configuration and tuning are more
important.
2. SPATIAL DATABASE MANAGEMENT APPROACHS
BASE ON ARCSDE AND ORACLE
Oracle is one of large commercial database manage systems.
When the research and evolution of Database technology
entering the Object-oriented technology phase, Oracle Database
becomes an Object-Relation Database. The great function and
reliability of Relation Database and the flexibility and modeling
function of Object Database make it easy to design, update and
develop for expert application. The support and application of
Oracle embody the character of Relation Database and
Object-Relation Database both.
ArcSDE is a Spatial Database Engine product of ESRI. It use
the middleware technology to store and manage the spatial data
in the commercial database, such as Oracle, DB2, Informix,
SQL Server, Sybase etc. Although there arc differences between
different Spatial Database Engine products, there are also
similarities in spatial data management approaches. So in this
article we make an example base on the ArcSDE and Oracle.
In Geodatabase, the object-oriented data model, of ArcSDE the
real world is described as a set of data objects. ArcSDE models
data objects in three basic ways: as a collection of discrete
features in vector format, as a grid of cells with spectral or
attribute data, or as a set of triangulated points modeling a
Internat
pe ron,
surface.
2.1 Arc:
The Arc
mation :
2.2 Ras
ArcSDE
user-de!
nalysis
With a :
can sav
defined
need of
raster b.
Spendir
ArcSDI
are RA?
Raster |
Block 1
2.2.1 F
an Arc!
the dat:
name tl
to the b
22.2 B
It regar
this ob
foreign
2.2.3 R
set acc
the pri
identifi
2.24 F
may be
band ta
size, th
of the |
table's :
2.3.5 R
stores
image
mosaic
referen
2.2.6 R
pixel d
blocks
block t
is a for
À cor
rasterb:
ArcSD
lation c
tegrity.
raster c
proced
