Transcript Course Review

Course Review
We’ve made it to Lecture 10
This is the last lecture in the Course CO24
© 2003 Monash IT Pty Ltd
1
Lecture 10 - Review of Course Material
This is the final lecture for Semester 1, 2003
I will introduce the major components of a Database
Management System as a means of reviewing much of the
material
Then we will move onto a slightly wider course review
© 2003 Monash IT Pty Ltd
2
© 2003 Monash IT Pty Ltd
3
Some final thoughts on SQL
IBM’s DB2 is overtaking Oracle’s DBMS 8.0.n and 8i as the
DBMS of the period.
An SQL analyst who also is a member of X3H2, a committee
set up to determine amendments and extensions to SQL
and which then become ANSI standard (if they are approved
etc.), was asked why this migration and customer
redirection was occurring. His name is Joe Celko.
© 2003 Monash IT Pty Ltd
4
Some final thoughts on SQL
His comments ?
DB2 products are now more uniform across their platforms.
This means that the same database occurs at all levels in a
company
Oracle runs on many platforms but is retrofitted and not
designed from the start for that niche.
DB2 has a better Optimiser and other internals (read indexing
and join functions)
© 2003 Monash IT Pty Ltd
5
Course Review
We will look briefly at the material covered in the preceding
lectures
Introduce some more material on SQL
and have a few thoughts about the examination
© 2003 Monash IT Pty Ltd
6
Introduction to SQL, Database Design
Data, Information, Need for Information
Database, Database Management Systems
Commercial Models - Hierarchical, Network, Relational
Business Activities, Functions, Processes
Need for Information to Manage Resources
© 2003 Monash IT Pty Ltd
7
Introduction to SQL, Database Design
Need for Information to be
Accurate, Representative, Timely, Adequately Based,
Protected against Loss, Corruption, Unauthorised Access
Communication capabilities
© 2003 Monash IT Pty Ltd
8
Introduction to SQL, Database Design
Designing a database
User requirements - current and future
Information outlines
Sources of data
Processes to be applied
Restrictions on data occurrences, either singly or
when associated with other data
© 2003 Monash IT Pty Ltd
9
Introduction to SQL, Database Design
Development of LOGICAL model
External Schema - User views of data
Conceptual Schema - Composite of all user views and
requirements
Development of Entity Relationship Diagram
Shows the ‘natural grouping’ of data
Shows the relationships between the ‘natural groups’
Shows the modality and cardinality (1:M, optional,
mandatory, number of sets which are related)
© 2003 Monash IT Pty Ltd
10
Introduction to SQL, Database Design
Designing a database (cont’d)
User Rules for existence and operations on data
Dependencies of data
Life of data
Value or worth of data
© 2003 Monash IT Pty Ltd
11
Introduction to SQL, Database Design
Determining what ‘data sets’ exist (entities)
Determining the relationships between these sets
Optional
Mandatory
1 to 1 (1:1) , 1 to Many (1:M)
or Many to Many (M:N)
Unary, binary or ternary relationships
Special conditions for existence of data
© 2003 Monash IT Pty Ltd
12
Introduction to SQL, Database Design
Expanding the Entities to Attributes
Conditions of existence for attributes
Dependencies of attributes on a determinant
Primary Key (and its characteristics)
Primary - Foreign Key occurrences (and referred to the
Entity Relationships)
Performance of the relationships on Insert, Update,
Modification of the Primary and Foreign keys
© 2003 Monash IT Pty Ltd
13
Introduction to SQL, Database Design
What are the attributes of each set ?
Data Structure Diagram
Primary Key(s), Foreign key(s)
Nullable, non nullable attributes
Suspicion of the effect of update, insert, delete operations
© 2003 Monash IT Pty Ltd
14
Introduction to SQL, Database Design
Normalisation - dependencies of attributes on the Primary
Key
Special conditions which could restrict some data
occurrences
Self checking features ?
Long term storage requirements
© 2003 Monash IT Pty Ltd
15
Introduction to SQL, Database Design
Access considerations - File Organisation and
Access Methods
Indexing - Changing Structure - Secondary Indexing
Performance Requirements
Logical Path Access evaluation - Data Base and Equipment
Loading
Growth, Volatility of Database - Scalability Aspects
© 2003 Monash IT Pty Ltd
16
Introduction to SQL, Database Design
Transaction Processing - Transaction Manager, Scheduler,
Recovery Manager
Stages of access to the physical database
Effect of power, software and hardware interruptions
Checkpoint, Logging facilities
Restart / Recovery / Backup outline
© 2003 Monash IT Pty Ltd
17
Introduction to SQL, Database Design
Concurrency Aspects - multi user (transaction) control
Time stamping, Locks
Adverse effects of no or poor concurrency (optimistic and
pessimistic)
Integrity Considerations - Entity, Referential, Domain,
Security of
© 2003 Monash IT Pty Ltd
- hardware, communications, data
access limitations, authorisations
18
Introduction to SQL, Database Design
Use of Data Dictionary information about data
Optimisation
- query execution plan
Data Base Administration
Distributed database
Data Warehousing
Data Mining
Features of a database management system
© 2003 Monash IT Pty Ltd
19
Introduction to SQL, Database Design
SQL - what is it ? What is its purpose ?
The commands : Create (object), delete, drop,
insert, update,
Querying with SQL
Select (list of attributes)
From (named tables)
Where (conditions for selection, including nested queries)
Group By (selection of aggregations
Having (condition for Grouping)
Order by (viewing sequence)
© 2003 Monash IT Pty Ltd
20
Introduction to SQL, Database Design
Operators :
> < = !=
in, between
exists, not exists
any, not any
AND OR
NOT
Avg, max, min, std, sum - set functions
Correlation - multiple table query
© 2003 Monash IT Pty Ltd
21
An Introduction to SQL - Some Terms
In the next few overheads, there will be some terms and
explanations which should help you to make the transition
from the methods of data storage and file processing to that
of the relational database style of storage and processing of
data.
© 2003 Monash IT Pty Ltd
22
More on Tables
Attributes which make up a row in a table are always scalar
or atomic data types
They are complete in themselves and cannot be broken into
smaller parts
Attributes can have constraints placed on them when
the table is created
or when the table is altered
These constraints are part of the database
Unlike spreadsheet cells, an attribute set of values can only
hold data.
© 2003 Monash IT Pty Ltd
23
Schemas and Tables
A Schema : is the data, the operators and the rules of the
defined database
Tables can be permanent (base tables) or virtual (views)
Any operation in SQL will return a result table
© 2003 Monash IT Pty Ltd
24
More Terms
Are you ready for a few more terms ?
Try these :
Scalar Values : This is a value which is not an array - one
dimension. 1 is a scalar value, 1,4 is not -it’s an array.
Literals : This is a constant which is referred to by its value.
Literal 3 is ‘the integer data type having the value 3’
Column Expression : This is a calculation which will return a
value for any data item in that attribute - also known as a
‘derived value’.
© 2003 Monash IT Pty Ltd
25
A Transaction
A transaction is a sequence of SQL statements which
Oracle treats as a single unit
The set of changes is made permanent with the Commit
statement
Part or all of a transaction can be undone with the
Rollback statement
A transaction starts with the execution of the first SQL
statement in the transaction and ends with either the Commit or
Rollback statement
© 2003 Monash IT Pty Ltd
26
A Transaction
Oracle guarantees that a transaction has statement-level
read consistency (the data stays the same while Oracle is
gathering and returning it)
If a transaction has multiple queries, then each query is
consistent, but not with each other
Transaction-level read consistency can be achieved with the
Set Transaction Read Only - (queries only)
© 2003 Monash IT Pty Ltd
27
Some Guidelines
1. The name allocated to a table, known as the table name,
must be unique in the schema.
2. The attribute names used in a table must be unique in that
table
3. It is good practice NOT to have a table name repeated as an
attribute name in the same table (e.g. a table named
Personnel could have an attribute ‘employee’ - not an
attribute named ‘personnel’.)
4. It is good practice to use the same name for an attribute
which occurs in other tables.
5. It is good practice to have related attributes in natural order e.g. Person’s name, title, address, salary rather than
address, person’s name, title, salary.
© 2003 Monash IT Pty Ltd
28
Some Thoughts about Keys
A KEY is an attribute (or a set of attributes) in a table which
uniquely identifies each row in that table
There are several types of keys :
1. A candidate key - this is one of the ‘possible’ keys in a
table. A an example, a motor car has a Registered Number,
and also a Vehicle identification Number (VIN).
2. A surrogate key is a key with no meaning in itself. They
are normally generated by the database system using some
form of number generator. Users rarely see these values except in Microsoft Access
© 2003 Monash IT Pty Ltd
29
Some Thoughts about Keys
Intelligent Keys : have a meaning. Longitude and Latitude of
an object (say a building) uniquely identifies a building and
also nominates where it is.
Super Key : This is normally a key with too many attributes
in it. Some of the columns could be removed and a key
would still exist. A good example is a key which contains ‘old
value/new value’ such as ‘old telephone number/new
telephone number’ when numbers are upgrades or altered.
A Primary Key : The unique, not null expression of the
access to each row in a database table - normally
constrained by a domain set.
© 2003 Monash IT Pty Ltd
30
Embedded SQL
In commercial processing, SQL normally is between the
database and a host programming language, passing data
in either a forward or backward direction ( accepting new
data, or returning results of queries.)
The host program communicates with the ‘outside world’
There are also PC tools which are designed to use SQL PowerBuilder, Designer2000, Developer2000, Delphi are
examples.
© 2003 Monash IT Pty Ltd
31
Embedded SQL
There are 3 problems
1. SQL has a rich or extended set of datatypes which might not
match those of the host program
2. The ‘Null’ may not exist in the host program language
3. SQL uses a set model of data - not a file model.
A file is processed (read, updated, deleted) one record at a
time
A set in SQL is read, updated, deleted ‘all at once’
Embedded SQL handles these problems.
© 2003 Monash IT Pty Ltd
32
Embedded SQL
Embedded SQL provides for SQL statements to be recognised
by the keywords ‘exec sql’ preceding the SQL commands.
A precompiler converts the SQL statements into host language
procedure calls.
The database calls are part of an Application Program
Interface library
© 2003 Monash IT Pty Ltd
33
New SQL Commands and Features
The OLAP functions proposed for SQL-99 are
ceiling
percentile_cont
regr_slope
corr
percent_rabk
regr_sxx
covar_pop power
regr_sxy
covar_samp range
regr_syy
cume_dist rank
row_number
dense_rank regr_avg
sqrt
exp
regr-avgx
stddev_pop
floor
regr_agvy
stddev_samp
ln
regr_count
car_pop
moving_avg regr_intercept
var_samp
moving_sum regr_r2
© 2003 Monash IT Pty Ltd
34
New SQL Commands and Features
The principal statistical functions are ;
correlation
cumulative distribution
percentile
standard deviation
© 2003 Monash IT Pty Ltd
covariance
ranking
linear regression
variance
35
Business Intelligence
Analysis is a fundamental to Business Intelligence
For example :Is there a correlation between the frequency of Web site visits
(hits) and customer profitability
Who would be the ‘interested parties’ to the results of this
analysis ?
© 2003 Monash IT Pty Ltd
36
Introduction to SQL, Database Design
Some other features:
Not all DBMS’s offer the same facilities
Not all SQL software supplier versions are the same
Some interesting differences : data types, (money, currency,
number - with or without ‘decimal’ provision)
Nulls and their processing capabilities
Aggregate functions especially in the handling of nulls
© 2003 Monash IT Pty Ltd
37
Some Commands You Have Met
Create : table, view, index, sequence (also create as select ..)
Drop : table, view, index, sequence
Insert : Adds a single row to an existing table
Delete : Deletes a row or rows or all contents from an
existing table
Update : Updates a row or rows
SqlLoad : Loads bulk data from an external file to the current
database table (sqlldr80 for Personal Oracle)
© 2003 Monash IT Pty Ltd
38
Some Commands You Have Met
Alter : With care, provides alteration to some of the table
attributes and properties
Copy : Copies data from 1 database to a named database
Export : Copies data from a named table in a database to the
Operating system level
Import : Copies data from the Operating system level to a
named database
Union : Combines 2 tables into 1
© 2003 Monash IT Pty Ltd
39
The Query Form
The form of a Query :
Select
(list of attributes)
From
(table, or list of table names)
Where
(conditions of selection)
Group By (used for aggregate functions)
Having (conditions of selection of the Group By)
Order By (determines the order of content displayed from
the Result table - Ascend, Descend on a named
attribute)
© 2003 Monash IT Pty Ltd
40
Operators and Predicates
Arithmetic :
+ - / * **
Logical
And
:
Or
Comparison Predicates :
Not
= > < <=
>= <> != ^=
Like Predicate : Pattern matching or Not matching
Between Predicate : Numeric valued constraints
In Predicate : String or numeric lists
© 2003 Monash IT Pty Ltd
41
More Predicates
Any
All
Exists
© 2003 Monash IT Pty Ltd
42
Constraints
Null (and Not null)
Unique
Check (attribute property limit - In, Between )
Aggregate Functions
Count, Sum, Average, Minimum, Maximum
and don’t forget the ‘additional’ functions in SQL99
© 2003 Monash IT Pty Ltd
43
Data Types
Varchar2(number) e.g. varchar2(15)
Number(p,s) e.g. number (10,2) 8 whole number values, 2
decimal values)
Date
Blob - Binary large object
Clob - Compressed binary large object
Other datatype supported by Microsoft - currency, long and
short integer, text, various temporal options
© 2003 Monash IT Pty Ltd
44
The Exam Paper
A few words about the examination requirements :
There have been 3 ‘unit tests’
And these, plus the course project, are the components of
the course result.
However, if you really want a formal examination, just let me
know
© 2003 Monash IT Pty Ltd
45
Introduction to SQL, Database Design
© 2003 Monash IT Pty Ltd
46