Copying or distributing in print or electronic forms without written permission of
Idea Group Inc. is prohibited.
The abstract processing plan comprising of the phases is illustrated in Figure 5(a)
and can be used to answer the star queries that belong to the query template of Figure
3. This plan is abstract in the sense that it does not determine specific algorithms
for each processing step; it just defines the processing that needs to be done. That is
why it is expressed in terms of abstract operators (or logical operators), which in
turn can be mapped to a number of alternative physical operators that correspond
to specific implementations.
An example abstract processing plan is shown in Figure 5(b) and it corresponds to
the query of Figure 4.
Having described the framework for query processing of OLAP queries, we move
next to discuss how this can be materialized on a hierarchical clustering-preserving
data structure, namely the CUBE File.
Figure 5. (a) The abstract processing plan; (b) the abstract processing plan for
the example query
MD_Range_Access
DATE
Residual_Jo n
(Day)
LOCATION
Group_Select
(area, month)
Main execution phase
DATE PRODUCT LOCATION
Create_Range
(Year= )
Create_Range
(Category=???a r
cond t on???)
Create_Range
(Populat on >
000000)
SALES_FACT
Residual_Jo n
(Store_id)
h-surrogate processing
MD_Range_Access
Residual_Jo n
Group_Select
Main execution phase
D Dj
Create_Range
FT
h-surrogate processing
Create_Range
D
Dj
Order_By
Residual_Jo n
.
Pages:
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294