Database Management Concepts Ch 17 syllabus 14'3

1
Database Management Concepts Ch 17 syllabus 14.3

• We have already covered normalisation at the end
  of LVI
• Please recap these notes

2
The purpose of a Database Management System (DBMS)

• Storage of data
• In separate tables with linked fields
• Within the structure of database or data
  dictionary.
• Provides interface between user and data
• And allows data to be queried
• And allows construction of reports.
• It also controls access to the data/security

3
How a DBMS can help consolidate data from several
different places

• Provides a buffer between the user and the
  underlying structure of the database so the user
  is not concerned with the structure of the
  database, just with making relevant use of the
  data
• Can be used to create a consistent database with
  a consistent interface so if users move from one
  location to another they already know how to use
  the system
• RDBMS will control access to data so all places
  will see the most up to date data
• In event of disaster, RDBMS should help with
  disaster recovery e.g. provide ability to
  consolidate data and roll back to a known correct
  state
• RDBMS may provide improved security to the system
  so that there is less chance of inexperienced
  users inadvertently changing data or malicious
  attempts to corrupt data succeeding

4
The role of a Database Administrator

• Designs the database by setting up appropriate
  tables, relationships, fields, naming conventions
  etc.
• Allocation of data access to users including
  access rights, providing user names
• Maintains the data dictionary i.e. the data
  structure
• Provides training to users as appropriate so new
  users learn how to use the system, and existing
  users get updates as necessary
• Monitors the performance of the RDBMS and makes
  appropriate adjustments to deal with problems as
  they arise e.g. system taking too long to produce
  a report
• In charge of the backup procedure and ensures the
  database is available to users at the times that
  they need it
• Notify users of any changes so that they can make
  changes to the way they use the database
  appropriately e.g. availability of new reports/
  queries

DAMPMIN
5
Data Consistency

• Data is stored as one value, and not stored again
  as another value.
• EG if Gayle Robinsons name is stored in one
  table, then stored as Gail Robinson in another
  table this data would be inconsistent.
• Consistency comes with reduced redundancy where
  data is only stored once, and this is the sole
  source of that data.

6
Data Integrity
Not corrupted

• Is the correctness of the data or how trustworthy
  the data is.
• It should be free from corruption and validated
  whenever it is transferred from one medium to
  another.
• EG Validation of data such as range check on an
  amount of currency as it is typed from a data
  capture form to a RDBMS
• EG Parity check of data as it is sent over a
  telecommunications link.

7
Data Redundancy

• Data is stored as one value in many places
• EG Lauren Carpenters name and contact details
  are stored in the personnel table, and her name
  and contact details are also stored in the
  payroll table.
• Redundancy means data is more likely to be
  inconsistent as any changes to one data item must
  be made manually to all its occurrences.
• A RDBMS reduces redundancy by creating a unique
  key field EG PersonID to replace duplicate
  details in other tables.

8
Data Independence

• Data tables are separate from the
  programs/functions that use them
• EG the same table may have two or more different
  views the delivery function can see customer
  name and address only however the credit control
  function can see the customer income, credit
  history, and contact details
• Changes in the structure of the data tables only
  affects those programs/ functions that are
  reliant on that part of the structure
• EG the introduction of a new field to a table
  only needs changes made to the functions that
  will use the field and not to all the functions
  that use the table.

9
Data Normalisation

• Process of breaking down complex data structures
  into simpler forms
• (1NF) Removal of repeating fields
• (2NF) Removal of composite keys/partial key
  dependencies
• (3NF) Removal of non-key dependencies
• Normalisation is the process for making the
  structure of a relational database more efficient
  by defining tables, fields, and relationships to
  minimise duplication of data and increase data
  consistency. Data integrity is assured by data
  validation specified in the data dictionary.
  Complex queries are also achieved using efficient
  relationships between tables.

10
DBMS Jargon

• Entity (Table/Record) Person / place / thing
  about which data is stored e.g. patient
• Attribute (Field) Property associated with the
  entity e.g. patients blood group
• Relationship Link between entities e.g. patient
  can have one ward but one ward can have many
  patients. This is a one to many relationship.

11
Entity Relationship Diagrams

• Provide a diagrammatic representation of the
  structure of the data and shows the types of
  relationships within the database. The
  relationships include
• one to one (not common)
• one to many (common)
• many to many (not physically possible) better to
  have

12
The Nature of a Client/Server Database System
(P210)

• Client Computer A local workstation that just
  provides the user interface to request data
  queries/reports from the server
• Server Computer A remote computer where the
  processing is carried out to provide the query
  results/reports to the client computer. The
  server computer is more powerful than the client
  computer because it has to contain the whole
  RDBMS.
• The advantages of a client/server database system
• cost-effective clients are cheaper than servers
  so this expensive resource is made available to a
  large user base
• data consistency is maintained as one copy of
  data is held on the server, rather than a copy
  held on each workstation
• communication between client server is
  minimal as only requests/results are
  communicated, not entire databases
• Department specific report formats or queries
  can be held on workstations meaning less room is
  taken up on the server and they are less likely
  to be accessed by the wrong people.

13
The End
Now do the exam questions! Do the exam questions!