Consider a relation s(A,B) with the following characteristics: • s has 700,000 tuples with 70 tuples per disk block • s has an index on attribute A with index height x=2 • The values that the attribute A takes in relation s are integers which are uniformly distributed in the range 1 ... 200. [Uniformly distributed ➔ there are roughly 3500 tuples for each A value.] (1.a) Assuming that the index on A is a secondary index, estimate the number of disk block accesses needed to compute the query σA=18(s). (1.b) What would be the estimated cost if the index were a clustering index?
Q1. Consider a relation s(A,B) with the following characteristics:
• s has 700,000 tuples with 70 tuples per disk block
• s has an index on attribute A with index height x=2
• The values that the attribute A takes in relation s are integers which are uniformly distributed in the range 1 ... 200.
[Uniformly distributed ➔ there are roughly 3500 tuples for each A value.]
(1.a) Assuming that the index on A is a secondary index, estimate the number of disk block accesses needed to
compute the query σA=18(s).
(1.b) What would be the estimated cost if the index were a clustering index?
Question processing is the translation of high-level questions into low-level expressions. This is a stepwise process, which is used at the physical level of the file system, to actually run the query to get the query optimization and results. It requires the basics of relational algebra and file structure.
Trending now
This is a popular solution!
Step by step
Solved in 2 steps
