Problem 1: Indexing on Sequential Files
Solution
1. 100 / 10 = 10 blocks. Explanation: A block can hold 10
keypointer pairs, and we need a pair for every tuple. 100
tuples / 10 keypoints pairs per block = 10 blocks.
2. (100 / 2) / 10 = 5 blocks. Explanation: A block can hold 10
keypointer pairs, and we need a pair for every tuple that is
the first in its block. Every block can hold 2 tuples. Therefore
there are 100 / 2 = 50 tuples that need keypointer pairs.
Now, it's as if we're building a dense index on these 50
tuples, and 50 / 10 = 5 blocks.
3. Clustered index reorders the data, unclustered does not.
So with unclustered index, can't have ordered tuples within
blocks (because nothing is sorted). Therefore sparse index
is impossible.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Problem 2: B+ Tree
Solution
1.
First valid answer (recommended): The following changes
will be done to the original Btree:
Step 1. The first node in the leaf level split to two nodes: (1,2)
and (4,5,6).
Step 2. The first node in the second level split to two nodes
(4,8) and (32,40).
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '07
 Winslett
 ﬁrst node, ORIGINAL Btree

Click to edit the document details