F28DM : Database Management Systems Indexes in Oracle

1
F28DM Database Management Systems Indexes in
Oracle

• Monica Farrow
• monica_at_macs.hw.ac.uk
• Room EMG30, Ext 4160
• Material on Vision my web page
• Content taken from HW lecturers,
• books by Rob Coronel, and by Connolly Begg
• For more info on everything, see Oracle9i
  Database Concepts (on the web)

2
Basic Use of Indexes

• Primary, Secondary, Composite
• When to use

3
Indexes

• An index helps speed up retrieval.
• A column that is frequently referred to in the
  WHERE clause is a potential candidate for
  indexing.
• Why use indexing?
• It's easy and provides immediate value
• It can be used to speed up existing applications
  without changing any of their logic or queries
• One of the main reasons for the success of
  relational databases is that they have efficient
  indexing systems.

4
Index recap

• An index typically consists of the search key
  (one or more attributes) and the row ID(s) of
  record(s) which match this key
• An index on a file speeds up selections on the
  search key fields specified for that index
• Any subset of the fields of a relation can be the
  search key for an index on the relation
• A primary index allows efficient searching on the
  primary key of a relation
• A secondary index allows efficient searching on
  other attributes which are often used in queries

5
Example the Sailors and Boats Domain

• Information needs to be stored about sailors,
  boats and reservations that sailors are making
  for boats
• There are three relations with the following
  schemas
• Sailor (idint, namestring, ratingint, ageint,
  basestring)
• Boat (idint, namestring, colourstring,
  basestring)
• Reservation (sidint, bidint, daydate)

6
Automatically Created Indexes

• On every relation, Oracle creates a non-clustered
  Btree index for the primary key
• A B-tree is shown on the next slide
• A B-tree stays balanced when items are added and
  removed
• Indexes will be created on
• Sailor table, id column
• Boat table, id column
• Reservations table, a composite index on sid, bid
  and day
• Additionally Oracle will also create an index for
  every UNIQUE constraint

7
B-tree index
Diagram just to give you an idea, you dont have
to memorise this
8
Composite index keys

• If the key consists of more than one attribute,
  the index is a composite or concatenated index
• i.e. an index for several attributes
• A composite index is sorted lexicographically
  according to the order of the attributes
• E.g for a composite key (name,age), we
  have(Kelly, 22) lt (Kelly, 63) lt (Smith, 18) lt
  (Smith,36)
• This is different to a composite key (age, name)

9
Secondary indexes

• Create secondary indexes for other attributes in
  your tables which are frequently searched by
• In the SQL WHERE clause, not the SQL SELECT line
• In this case, you must explicitly create the
  index, as shown on the next slide

10
Explicitly Created Indexes

• An index
• Has a name. Choose an informative one.
• Is created for a sequence of attributes over a
  table which must already exist
• Can be DROPped
• Examples
• CREATE INDEX sailor_name_idx ON Sailor(name)
• DROP INDEX sailor_name_idx
• CREATE INDEX sailor_name_and_age_idx ON
  Sailor(name, age)

11
Using CREATE INDEX (1) 

• The CREATE INDEX statement creates a sorted index
  on the specified columns of the named table.
• Indexes are automatically used to improve the
  performance of queries issued to the database,
  and to sort queries with an ORDER BY clause.
• Once an index is created, it is never referenced
  in a SQL statement again except to validate it
  (VALIDATE INDEX) or delete it (DROP INDEX).
• Indexes cannot be created for views.
• Add an index to the underlying base table

12
Using CREATE INDEX (2)

• Indexes are always owned by the owner of the
  table. The index name must be unique for each
  table and owner.
• Exclusive table use.   
• You cant create an index if the table is in use.
• CREATE INDEX can be time consuming and the server
  will not process requests referencing the same
  table while the statement is being processed.
• As with all other CREATE commands, store the
  command in a text file
• To remind you what the index is like
• So it can be rerun if you decide to re-create
  your DB

13
Overheads v performance

• There is an overhead involved in the maintenance
  and use of secondary indexes.
• The index must be updated whenever the table is
  updated
• Disk space is needed for the index
• Balance the overhead against the performance
  benefits
• Faster data retrieval

14
How to choose secondary indexes (1)

• Add a secondary index to a foreign key if it is
  frequently accessed
• E.g. Boat id in Reservations, if we frequently
  want to know the name of the boat in a
  reservation
• Add a secondary index to any attribute that is
  frequently used as a search key
• E.g. day in Reservations (what reservations do we
  have for today or tomorrow?)
• E.g. name in Sailor (what is the rating for the
  sailor called Lubber?). People often search by
  name rather than id.

15
How to choose secondary indexes (2)

• Add a secondary index to any attributes
  frequently used in order by, group by, min, max,
  avg
• E.g. age in Sailor if frequently want list in age
  order
• Add a composite secondary index to a small set of
  attributes if the index would provide the full
  details for a frequently used query without
  having to search the base table
• E.g. rating and age in Sailor if a frequent query
  isSELECT rating, AVG(age) FROM SailorGROUP BY
  rating

16
When not to use secondary indexes

• If the relation is small not many rows
• If the relation or attribute is frequently
  updated
• If periodic large updates, drop the index, update
  the data, re-create the index
• If the attribute is always used in queries that
  retrieve a significant proportion of the rows in
  the relation
• E.g. If the attribute has low sparsity (i.e. the
  number of different values is small, such as
  gender F or M).

17
More Advanced Indexing Topics

• Index-organised tables
• Bitmap indexes
• Clusters

18
Index organised tables

• Index organised tables
• Entries of index are table records rather than
  row IDs
• Useful if frequent access to complete table via
  the index
• In Oracle, the default table organisation is heap
• To store the table in primary key order, add the
  words ORGANIZATION INDEX to the end of the CREATE
  TABLE command
• In the B-tree, the leaf nodes are replaced by all
  the non-primary key attributes in the row

19
.Bitmap indexes

• Bitmap indexes
• An example on the next slide
• For each value of a domain, there is a bitmap
  identifying the row Ids of satisfying tuples
• Usually applied to attributes with a sparse
  domain
• In Oracle, lt100 distinct values
• E.g. bitmaps for all tuples with sex male and
  for sexfemale
• Updating a bitmap takes a lot of time, so use for
  tables with hardly any updates, inserts, deletes
• Ideal for data warehousing
• See later lecture, if time

20
Bitmap indexes example

• The first table is the Sailors table
• The second table shows a bitmap index for rating,
  assuming values are only from 1-3

21
Clusters in Oracle

• Oracle supports 2 types of clusters
• Indexed clusters
• Hashed clusters

22
Clusters

• Clusters are groups of 1 or more tables
  physically stored together because they share
  common columns and are often used together
  (Connolly Begg)
• This improves disk access
• The related columns of the tables are called the
  cluster key

23
Saiilor/Reservation cluster example

• Sailor and Reservation relations could be stored
  together, with the cluster key Sailor id
• Each relation can be retrieved independently
• Whenever the joined relations are required,
  access is faster

24
Should we cluster Sailors and Reservations?

• A cluster for Sailors and Reservations could be
  created if many queries in the application joined
  these tables together.
• A cluster for Sailors and Reservations would be
  inefficient if reservations were only a minor
  part of the application, and many more queries
  involved Sailors joined with other tables

25
Indexed clusters

• Use when
• Queries retrieve records over a range of cluster
  key values
• Clustered tables may grow unpredictably
• With an indexed table or index cluster, Oracle
  locates the rows in a table using key values that
  Oracle stores in a separate index.
• E.g.
• No idea how many reservations there are going to
  be
• Queries like SELECT FROM Sailors S,
  Reservations RWHERE S.id R.sidAND S.id gt 50

26
Defining a indexed cluster in SQL(1)

• Create a cluster first
• CREATE CLUSTER sailor_id_cluster (id int)
• A cluster can be made for a fixed number of typed
  attributes (here just one)
• Define an index on the cluster
• CREATE INDEX sailor_id_idx ON CLUSTER
  sailor_id_cluster

27
Defining an indexed cluster in SQL (2)

• Create table(s) for the cluster
• CREATE TABLE Sailor (id INT NOT NULL,
• name VARCHAR(20),
• 
  ranking INT,
• age
  INT)CLUSTER sailor_id_cluster(id)
• CREATE TABLE Reservation (sid INT, bid
  INT.........as before....)CLUSTER
  sailor_id_cluster(sd)

28
Hash Clusters

• The key of a hash cluster, like the key of an
  index cluster, can be a single column or
  composite key (multiple column key).
• A hash function is applied to the cluster key to
  obtain a hash value.
• The hash value determines the location where the
  rows of a table are physically stored. So all
  records with the same cluster key are stored
  together

29
Creating a hash cluster

• Create a hash cluster
• CREATE CLUSTER sailor_id_hashcluster (id
  int)HASH IS id HASHKEYS 10000
• Add table(s) as before
• CREATE TABLE Sailor (id INT NOT NULL,
• name VARCHAR(20),
• 
  ranking INT,
• age
  INT)CLUSTER sailor_id_hashcluster(id)

30
Use hashing when (1)

• Most queries are equality queries on the cluster
  key
• SELECT ... WHERE cluster_key ...
• the cluster key in the equality condition is
  hashed, and the corresponding hash key is usually
  found with a single read.
• In comparison, for an indexed table the key value
  must first be found in the index (usually several
  reads), and then the row is read from the table
  (another read).
• AND WHEN . . .

31
Use hashing when (2)

• AND
• The tables in the hash cluster are primarily
  static in size so that you can determine the
  number of rows and amount of space required for
  the tables in the cluster.
• If tables in a hash cluster require more space
  than the initial allocation for the cluster,
  performance degradation can be substantial
  because overflow blocks are required.

32
Hashing v Indexing

• To find or store a row in an indexed table or
  cluster, a minimum of two (there are usually
  more) I/Os must be performed
• One or more I/Os to find or store the key value
  in the index
• Another I/O to read or write the row in the table
  or cluster
• In contrast, if a hash function is used to locate
  a row in a hash cluster no I/O is required.
• As a result, a minimum of one I/O operation is
  necessary to read or write a row in a hash
  cluster.