Tree-Structured Indexes

Computer Science Department Columbia University

CS4112, A. Biliris

Introduction
Tree-structured indexing techniques support both range searches and equality searches. ISAM: static structure; B+ tree: dynamic, adjusts gracefully under inserts and deletes. As for any index, 3 alternatives for data entries k*:

Data record with key value k <k, rid of data record with search key value k> <k, list of rids of data records with search key k>

Choice is orthogonal to the indexing technique used to locate data entries k*.
2

CS4112, A. Biliris

Index File

Example: Find all students with gpa = 3.0

Assume sorted data, do binary search to find first such student, then scan to find others. Cost of binary search can be quite high.

Simple idea: Create an index file and do binary search on (smaller) index file!
k1 k2 kN

Index File

Page 1
CS4112, A. Biliris

Page 2

Page 3

Page N

Data File
3

ISAM
P 0

index entry
K P K 2 P K m Pm

Index file may still be quite large. But we can apply the idea repeatedly! Rule:

Ki Ks pointed by Pi < Ki+1 Ks pointed by P0 < K1

Non-leaf Pages

Leaf Pages Overflow page Primary pages

Leaf pages contain data entries.

CS4112, A. Biliris 4

Comments on ISAM

Data Pages

File creation: Leaf pages allocated sequentially, sorted by search key; then index pages allocated, then space for overflow pages. Overflow pages Index entries: <search key value, page id>; they `direct search for data entries, which are in leaf pages. Search: Start at root; use key comparisons to go to leaf. Cost is log F N ; F = # entries/index pg, N = # leaf pgs Insert: Find leaf data entry belongs to, and put it there. Delete: Find and remove from leaf; if empty overflow page, de-allocate.
5

Index Pages

Static tree structure: inserts/deletes affect only leaf pages.

CS4112, A. Biliris

Example ISAM Tree

Rule: Ki Ks pointed by Pi < Ki+1 Ks pointed by P0 < K1

Each node can hold 2 entries; no need for `next-leaf-page pointers. (Why?)
Root
40

20

33

51

63

10*

15*

20*

27*

33*

37*

40*

46*

51*

55*

63*

97*

CS4112, A. Biliris

Insert 23*, 48*, 41*, 42* ...

Root Index Pages
40

Rule: Ki Ks pointed by Pi < Ki+1 Ks pointed by P0 < K1

20

33

51

63

Primary Leaf Pages

10* 15* 20* 27* 33* 37* 40* 46* 51* 55* 63* 97*

Overflow Pages

23*

48*

41*

42*

CS4112, A. Biliris

... Then Delete 42*, 51*, 97*, 55*

Root Index Pages
40

This leaf page is not deallocated

20

33

51

63

Primary Leaf Pages

10* 15* 20* 27* 33* 37* 40* 46* 51* 55* 63* 97*

Overflow Pages

23*

48*

41*

42*

Note that 51* appears in index levels, but not in leaf!

CS4112, A. Biliris 8

B+ Tree: Most Widely Used Index

Keeps tree height-balanced. (F = fanout, N = # leaf pages) Minimum 50% occupancy (except for root). Each node contains d <= m <= 2d entries. The parameter d is called the order of the tree. Supports equality and range-searches efficiently.

Index Entries (Direct search)

Data Entries ("Sequence set")

CS4112, A. Biliris 9

Example B+ Tree

Rule: Ki Ks pointed by Pi < Ki+1 Ks pointed by P0 < K1

Search begins at root, and key comparisons direct it to a leaf (as in ISAM). Search for 5*, 15*, all data entries >= 24* ...
Root
13 17 24 30

2*

3*

5*

7*

14* 16*

19* 20* 22*

24* 27* 29*

33* 34* 38* 39*

Based on the search for 15*, we know it is not in the tree!

CS4112, A. Biliris 10

Inserting a Data Entry into a B+ Tree

Find correct leaf L. Put data entry onto L.

If L has enough space, done! Else, must split L (into L and a new node L2)
Redistribute entries evenly, copy up middle key. Insert index entry pointing to L2 into parent of L.

This can happen recursively

To split index node, redistribute entries evenly, but push up middle key. (Contrast with leaf splits.) Tree growth: gets wider or one level taller at top.
11

Splits grow tree; root split increases height.

CS4112, A. Biliris

Example: Insert 8*
Root
17

Rule: Ki Ks pointed by Pi < Ki+1 Ks pointed by P0 < K1

17? 13 5 17 13 24 30

Push-up 17
24 30

2*

3*

5*

7*

14* 16*

19* 20* 22*

24* 27* 29*

33* 34* 38* 39*

Copy-up 5

5*

7*

8*

CS4112, A. Biliris

12

Example B+ Tree After Inserting 8*

Root
17

13

24

30

2*

3*

5*

7* 8*

14* 16*

19* 20* 22*

24* 27* 29*

33* 34* 38* 39*

Notice that root was split, leading to increase in height. In this example, we can avoid split by re-distributing entries; however, this is usually not done in practice.
CS4112, A. Biliris 13

Deleting a Data Entry from a B+ Tree

Start at root, find leaf L where entry belongs. Remove the entry.

If L is at least half-full, done! If L has only d-1 entries, Try to re-distribute, borrowing from sibling (adjacent node with same parent as L). If re-distribution fails, merge L and sibling.

If merge occurred, must delete entry (pointing to L or sibling) from parent of L. Merge could propagate to root, decreasing height.

CS4112, A. Biliris 14

Example: (After Inserting 8*, Then) Delete 19* and 20* ...
Root
17

13

Copy-up 27

24 27

30

2*

3*

5*

7* 8*

14* 16*

19* 20* 22* 24* 22*

24* 27* 29* 27* 29*

33* 34* 38* 39*

Deleting 19* is easy. Deleting 20* is done with re-distribution.

CS4112, A. Biliris 15

Example: and then Delete 24*

Root
17 Delete this index entry 5 13 27 30

2* a

3*

5* b

7* 8*

14* 16* c

22* 27* 29* 24* d Pull down key & copy pointer

27* 29* e merge

33* 34* 38* 39* f

Root
17

5 a
CS4112, A. Biliris

13 b c

17 d

30 f d

30 f
16

merge

Tree (after Deleting 24*)

Root
5 13 17 30

2*

3*

5*

7*

8*

14* 16*

22* 27* 29*

33* 34* 38* 39*

CS4112, A. Biliris

17

Example of Non-leaf Re-distribution

An example tree shown below after a deletion in a leaf that results in the roots right child having 2 pointers. In contrast to previous example, can re-distribute entry from left child of root to right child.

Root
22

13

17

20

30

2* 3*
CS4112, A. Biliris

5* 7* 8*

14* 16*

17* 18*

20* 21*

22* 27* 29*

33* 34* 38* 39*

18

After Re-distribution
Intuitively, entries are re-distributed by `pushing through the splitting entry in the parent node. It suffices to re-distribute index entry with key 20; weve re-distributed 17 as well for illustration.

Root
17

13

20

22

30

2* 3*
CS4112, A. Biliris

5* 7* 8*

14* 16*

17* 18*

20* 21*

22* 27* 29*

33* 34* 38* 39*

19

Prefix Key Compression

Important to increase fan-out. (Why?) Key values in index entries only `direct traffic; can often compress them.

E.g., If we have adjacent index entries with search key values Dannon Yogurt, David Smith and Devarakonda Murthy, we can abbreviate David Smith to Dav. (The other keys can be compressed too ...)
Is this correct? Not quite! What if there is a data entry Davey Jones? (Can only compress David Smith to Davi) In general, while compressing, must leave each index entry greater than every key value (in any subtree) to its left.

Insert/delete must be suitably modified.

20

CS4112, A. Biliris

B+ Trees in Practice; Performance

Typical order: 100. Typical fill-factor: 67%.

Average fanout = 133 Performance (# of I/Os) of search/insert/delete: logFN Performance of scans next/previous operations: O(1) Height 3: 1333 = 2,352,637 records Height 4: 1334 = 312,900,700 records Level 1 = 1 page = 8 Kbytes Level 2 = 133 pages = 1 Mbyte Level 3 = 17,689 pages = 133 MBytes

Lets get practical typical capacities:

Can often hold top levels in buffer pool:

CS4112, A. Biliris

So, safe to assume search cost is 2 4 I/Os.

21

A Note on `Order

Order (d) concept replaced by physical space criterion in practice (`at least half-full (in bytes)).

Index pages can typically hold many more entries than leaf pages. Variable sized records and search keys mean different nodes will contain different numbers of entries. Even with fixed length fields, multiple records with the same search key value (duplicates) can lead to variable-sized data entries (if we use Alternative (3)).

CS4112, A. Biliris

22

Bulk Loading of a B+ Tree

If we have a large collection of records, and we want to create a B+ tree on some field, doing so by repeatedly inserting records is very slow. Bulk Loading can be done much more efficiently. Initialization: Sort all data entries, insert pointer to first (leaf) page in a new (root) page.

Root Sorted pages of data entries; not yet in B+ tree

3* 4*
CS4112, A. Biliris

6* 9*

10* 11*

12* 13*

20* 22* 23* 31* 35* 36*

38* 41* 44*

23

Bulk Loading (Contd.)

Root
10 20

Index entries for leaf pages always entered into rightmost index page just 3* above leaf level. When this fills up, it splits. (Split may go up right-most path to the root.) Much faster than repeated inserts, especially when one considers locking!

Data entry pages

6 12 23 35

not yet in B+ tree

4*

6* 9*

10* 11* 12* 13* 20*22* 23* 31* 35*36* 38*41* 44*

Root

20

10

35

Data entry pages not yet in B+ tree

38

12

23

CS4112, A. Biliris

3* 4*

6* 9*

10* 11* 12* 13* 20*22* 23* 31* 35*36* 38*41* 44*

24

Summary
Tree-structured indexes are ideal for rangesearches, also good for equality searches. ISAM is a static structure.

Only leaf pages modified; overflow pages needed. Overflow chains can degrade performance unless size of data set and data distribution stay constant. Inserts/deletes leave tree height-balanced; log F N cost. High fanout (F) means depth rarely more than 3 or 4. Almost always better than maintaining a sorted file.
25

B+ tree is a dynamic structure.

CS4112, A. Biliris

Summary (Contd.)
Typically, 67% occupancy on average. Preferable to ISAM; adjusts to growth gracefully. If data entries are data records, splits can change rids! Key compression increases fanout, reduces height. Bulk loading can be much faster than repeated inserts for creating a B+ tree on a large data set. Can easily control fill factor on pages. Most widely used index in database management systems because of its versatility. One of the most optimized components of a DBMS.

CS4112, A. Biliris

26