[PLE
hin a
in this
intenance
re coming up
reasing use is
| metadata
c utility data
g the
tegy. The
> solutions;
streamlining
amline data
nterprise GIS
iscussed in
icable to the
asis for
discuss the
'eamlimed
The
ware solutions
ective 1s to
any of the
h using
s will in turn
investments in
ition
MENTS
ich should start
to those
te data
thers:
nterprise
ince strategies
es and
_ transactional,
ronment for
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
4. Developing adequate metadata on enterprise-wide
database to assist with making data currency and
accuracy assessment analysis
A cursory examination of these four major factors will show
that they are related and are best addressed by a combination of
prudent technology (software) choices, hardware architecture
and functional/effective organizational IT practices. The major
components of this approach are discussed in the remaining
sections of this paper. The technology component uses ESRI's
ArcGIS 8x/ArcSDE 8x technology and Microsoft SQL 2000
RDBMS and Microsoft .NET suite of development
environments. The hardware architecture component is flexible,
and can be addressed on a case-by-case basis, while the overall
IT best-practices are very critical to the success of the whole
approach.
3. TECHNOLOGY SOLUTIONS FOR MAINTAINING
SPATIAL DATABASE INFRASTRUCTURE
It is reasonable to expect that any adopted technology for
developing an enterprise spatial database infrastructure should
natively integrate functionalities for database maintenance.
While this is often the case, the generic nature of most solutions
makes it necessary for specific utilities to design and utilize
protocols and procedures that best resolve their specific data
maintenance requirements.
One of the lessons learnt in the empirical implementation
discussed in this paper is the fact that buying into a technology
solution is probably the easiest aspect of adopting an enterprise
GIS; making the solution work for existing workflow processes
that have evolved and have been standardized over time is the
most difficult aspect of systems integration.
This paper is based on an empirical implementation of ESRI's
geodatabase data model for developing an electric utility
database and data maintenance infrastructure that addresses in
varying degrees, the four factors earlier identified.
The geodatabase model supports the physical storage of spatial
data inside a DBMS (e.g. SQL 2000) while also supporting
transactional views of data (versioning), objects with attributes
and behavior. These characteristics highlight the concept of
intelligent GIS data, simplifying data management tasks, and
allowing for more meaningful use of data.
With this architecture, multiple users can access, share and
edit GIS data through the use of versioning and long
transactions subject to DBMS permissions and GIS
administration tools. Differences between versions of the
database can be reconciled and the master version update with
the reconciled version. Support for disconnected editing (data
check-out and check-in) provides a platform for integrating
field-based data collection and reconciliation with production
data.
Key attributes of edited database objects are also automatically
maintained e.g. shape area and shape length. Moreover, this
model offers support for intelligent features, rules and
relationships. allowing users to define topological and
associative relationships/rules that determine how database
objects interact, how they are edited and how referential
integrity is preserved.
Data security and protection are also features of the software
platforms being discussed. although security is best managed
Un
LI
UJ
jointly with the hardware infrastructure. Nevertheless, at the
software level, additional data protection is available by
specifying validation rules, network connectivity rules.
relationship rules and custom rules. Specifying these rules
further enables the GIS systems administrator to maintain
database integrity.
We will now discuss the features of the geodatabase model and
how data maintenance tasks are streamlined using this model.
3.1 Multiple Data Access, Sharing and Updating
One of the most useful functionality of an enterprise
geodatabase is the ability to support multiple users, sharing and
editing the same data objects at the same time. Data sharing
could involve a simple task as simply displaying the same data
layer on multiple users’ desktops, or more complex task like
querying the same data object and retrieving records by
multiple users.
Similarly, it is possible for multiple users to be editing the same
data object (layer). These edits have to register in the database
without the possibility of corrupting the database or making it
unstable for other users who may not be necessarily carrying
out edits to the database. This is where versioning and long
transaction strategies come into play. Some practical scenarios
are depicted in Figure | and Figure 2
3.2 Versioning and Long Transaction
One of the ways that a database can be effectively maintained is
to grant specific editing privileges to certain users by the
DBMS/ArcSDE administrator. This means that only users with
appropriate read-write access can edit the database. Versioning
is used to track multi-user editing, post edit conflicts and
reconcile database versions. Un-registering the enterprise
geodatabase as a versioned prevents all users from editing it
through the clients application.
Initial Editors Post-Edit Reconciled
Versions Versions Versions
Default
£
Figure 1. A Typical Versioning Scenario
Locking is also used to enforce data integrity during editing,
and three types of locaks are available for this purpose. These
are schema lock, share lock and exclusive lock. Schema lock
ensures that the geodatabase structure cannot be modified when
users are currently connected to the database. An exclusive
lock is required to alter the structure (schema) of a geodatabase
and once acquired, no share lock can be applied. Only the
geodatabase owner can alter schema. Share edit locks are
ce
acquired when users are querying or editing a feature class or
table.
