Entity schema for data acquisition
FLOW
SOURCE_TARGET
SELECTION
TERM_CRITERION
DESTINATION PRE_PROCESSING
04 Adzic, Fiore, & Sisto
Copyright ?© 2007, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of
Idea Group Inc. is prohibited.
two methods to programmatically interact with a server: Pro*C/Fortran/Cobol and
so forth, which are a set of precompilers in order to embed SQL into the specific
language and OCI which are an API callable from C/C++. Pro*C is quite easy to use
but has several limitations (e.g., massive loading) while the OCI is more difficult (or
better speaking, prolix) to use, but constitutes a flexible and complete interface.
Loading data in Oracle, also in massive parallel mode, can be simply done with the
utility SQLldr which takes a file as input (or many files in parallel mode) and loads
them into specified tables. Another more elegant way to do the same thing is to use
the external tables; this method is substantially analogous to SQLldr with the difference
managed in SQL. Using these simple techniques in ETL has the drawback
to add the cost to write a file and to read and parse it by SQLldr. When data sum
to dozens of gigabytes, it becomes convenient to use Direct path OCI4 to load data
directly into the database avoiding useless write and read to file.
As the whole OCI, these sets of functions for direct path are too complex/prolix to
be utilized directly in the code.
Pages:
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217