Some sample exam 2 questions:
1. Define the following terms:
Cursor
Entity, Entity set (in an ER model)
Relationship, Relationship set (in an ER model)
Key constraint (in an ER model)
Participation constraint (ER)
Weak entity (ER)
ISA hierarchy (ER)
3NF (Third Normal Form)
BCNF (Boyce-Codd Normal Form)
Functional dependency
Lossless join decomposition
Dependency preserving decomposition
2. A SQL query may have five different clauses, a SELECT, FROM, WHERE,
GROUP BY and HAVING clause. Explain how each of these clauses works
and how they are evaluated in an SQL query. Give an example using
all five clauses.
3. For the following relations (keys are shown with _s over the field name):
Student Class_Grade Class
___ ___ ___ ___
sname sid age gpa sid cid grade cid cdept cnum
-------------------- ------------- --------------
Ann 1 21 3.5 1 1 A- 1 CS 1621
Bob 2 19 3.4 1 2 A 2 BIOL 1011
Cal 3 20 2.6 2 1 B 3 ECE 1315
Dee 4 22 4.0 2 3 C
3 1 A
3 2 C
3 3 F
4 2 A
How would you design the following queries in SQL:
- The id#s of students who are 19 or 20.
- The student names (sname) of those students who have taken CS 1621.
- Students (names) who have taken a biology course or ECE course.
- Students (names) who have taken every course in Class.
- Students (names) who have received an A.
For added practice, produce queries for exercises 8.3, 9.1, and 9.3.
4. Answer the question above using QBE and skeletons as shown in class.
5. How is a view defined in SQL? How are queries on views resolved? What
are some of the difficulties introduced with views?
6. If you have a selection operator
SELECT equality condition ( relation )
how would you perform the operation under the following conditions, and
what would be the estimated cost of your approach in disk I/Os.
- There is no index, data is unsorted
- There is a clustered tree index
- There is a hash index
How would your answer change if the condition were not an equality
condition?
If there is more than one index available, how do you choose which
mechanism(s) to use?
7. How is the projection operator implemented with sorting? How about hashing?
What are the expected costs for these operations in terms of disk I/Os?
8. Give pseudo-code for the following join operations (you may assume the join
is performed on a single attribute with an equality condition). Also give
the expected cost for each operation.
Simple Nested Loops Join
Index Nested Loops Join
Block Nested Loops Join
Sort-Merge Join
Hash-Join
9. What is meant by the term reduction factor? How does it apply in query
evaluation? Give at least two examples of reduction factors and how they
might affect a particular query.
10. SQL queries can always be mapped to a simple query plan. Explain how this
simple query plan can be generated from any SQL query and why we would not
want to use such query plans in general.
11. System R only considers left deep plans. Why is this? Give an example of
a plan that is left deep and one that is not.
12. Give an example of an optimization that might be performed on a query
plan. What is the advantage of your optimization?
13. What are Armstrong's Axioms for reasoning with a set of functional
dependencies? What is the closure of a set of FDs? What is the
attribute closure of a set of attributes with respect to a set of FDs?
14. What do functional dependencies relate to the concept of redundancy in a
database? What is our main mechanism for combatting redundancy? Give
an example.