Transcript Nonstandard - Sheffield Hallam University

What is this describing?
“ The columns in the tables cover a year at
ten day intervals. The decans are placed in
the order in which they arise and in the next
column, the second decan becomes the first
and so on….
Tables to help make these computations have
inside of coffin lids
been found on the _____________”
Faculty of ACES
© Sheffield Hallam University
So what IS a database?
•
•
•
•
Some sort of permanency
Allows data to be extracted (easily)
Can allow data to be appended
Can allow data to be inserted, deleted or modified
– more difficult pre-computer age
• Contains data!
© Sheffield Hallam University
Lets name some databases….
• I will go first:
– ORACLE !
© Sheffield Hallam University
Ways of managing modern databases
• Flavours of Database Management System
(DBMS) include
–
–
–
–
–
Relational
Object-Relational
Hierarchical Databases
Network Databases
Object Oriented
• But whatever the dbms, they are all about mapping logical constructs
to physical ons and offs recorded permanently, normally on a HDD.
© Sheffield Hallam University
Common building blocks
• BIT
• 0 or 1, true or false….
• BYTE = 8 bits
• 1010 1010 = ascii J
• FIELD or Column
• SURNAME (eg Jones)
• Record or Row
• Several Fields (Deptno, Desc, Loc)
• File or Table
• Several Records
• Database
• Several Files
© Sheffield Hallam University
Issues of Access
• Sequential
– with n records, n/2 probes required to find data
– when n is large, very inefficient
• So willingly spend time storing records with
retrieval in mind:
– Order the data itself
– Create indexes
© Sheffield Hallam University
From access method to DBMS
• Whatever the dbms, core physical decisions are
being taken
• It is all about mapping from the logical to the
physical disk storage in the most efficient way.
• This will differ, depending upon what you are
trying to do:
–
–
–
–
–
query intensive
data quantities
data complexity
user volumes
architecture available
© Sheffield Hallam University
DBMS - why the differences?
• Driven by both technology and different
needs:
– Hierarchical developed initially for product
assemblies
– Network DB driven by need for inventory
control (one supplier-several products AND
products supplied by several suppliers)
– RDBMS the need to keep records
– OODBMS: Complex business objects, CAD
© Sheffield Hallam University
RDBMS - some weaknesses
• Translation needed from real world to relational
– one argument for the move to OO paradigm
• Assumes all rows will contain the same attributes
– can lead to masses of virtual empty spaces which slows recovery
• Constraint checking is a hefty overhead
– should some of this be handled by the UI?
– Recent impact of middle-tier as a place for rules
• SQL is non-procedural and was specifically designed to be
approachable to non-programmers
– hence PL/SQL as a work-around
•
Is all data SET-orientated?
© Sheffield Hallam University
Hierarchical Databases
•
•
•
•
Common in the mainframe environment
One of the oldest DBMS
IMS still performing strongly
Organised like the branches of a tree:
– single parent can have many children, but can’t
be related to other parents
• record searches and updates can be rapid, as
the structure is pre-defined. They always
start from the top.
© Sheffield Hallam University
Hierarchical cont...
Salesman
Buyer A
Product 1
Buyer B
Product 4
Buyer C
Product 5
© Sheffield Hallam University
Object DBMS
• Industry Standards set by ODMG
– Object Database management group
– www.odmg.org
• ODBMS is defined (by them) as:
– ..a DBMS that integrates database capabilities
with object-oriented programming capabilities
• Major components include:
–
–
–
–
Core Object Model
ODL - for specifying objeccts
OQL - for querying (“very close to” SQL92)
Binding to C++,Smalltalk or Java
© Sheffield Hallam University
Common aspects of ODBMS
• Object (always has an OID)
• Objects are of a type
• Types describe a common set of properties (states)
and of behaviour (methods)
• State of Object defined by the values of properties
or of relationships with other objects
• Database has a Object Manager layer between
objects and permanent storage. Schema defined in
ODL.
• Much more info at the OODBMD manifesto site:
–
http://www-2.cs.cmu.edu/People/clamen/OODBMS/Manifesto/htManifesto/Manifesto.html
© Sheffield Hallam University
ODBMS is ideal for:
• Complex modelling environments because
objects allow high levels of abstraction
• Non record-focused applications:
– CAD
– Multimedia
• Reusability focused, modular systems
© Sheffield Hallam University
ODBMS - where are they?
• Performance, in terms of storing and retrieving
data, can be poor
• RDBMS have a stranglehold on the commercial
database marketplace
• The provision of a “halfway house” ObjectRelational model
• Whilst the Assembler programmer predicts
exactly, the architect of complex components is
too far removed from the disk access to be able to
design efficiently 
© Sheffield Hallam University
• http://www.dbazine.com/pascal10.html
ODBMS - where are they?
• Some examples which prove OO is very
much still alive:
• OBJECTIVITY http://www.objectivity.com/
• Versant http://www.versant.com/
• ObjectDB http://www.objectdb.com/
XML
• "XML is a human-readable, machine-understandable,
general syntax for describing hierarchical data,
applicable to a wide range of applications,
databases, e-commerce, Java, web development,
searching, and so on. "
– Oracle OTN website, my italics
• Building blocks are: Tags, Elements and attributes
• Separates Data from validity checking and from
presentation
– by using DTD or Schemas to validate, and stylesheets to present
© Sheffield Hallam University
Native XML
• A native XML database...
– Defines a (logical) model for an XML document -- as opposed to
the data in that document -- and stores and retrieves documents
according to that model.
– Has an XML document as its fundamental unit of (logical) storage,
just as a relational database has a row in a table as its fundamental
unit of (logical) storage.
– Is not required to have any particular underlying physical storage
model. For example, it can be built on a relational, hierarchical, or
object-oriented database, or use a proprietary storage format such
as indexed, compressed files.
© Sheffield Hallam University
Issues with XML in RDBMS
• Paradigm conflict:
– Hierarchical V Relational
• XML is not designed to be queriable.
– Xpath not as powerful as SQL?
– Whole new syntax to learn
• XML is not normalised
• ISO SQL does not lend itself to generating XML
output
© Sheffield Hallam University
How to store XML in RDBMS
• Assuming XML inbound, do you:
– break it up and store it in relational tables
– store the document as a CLOB (Character Large Object datatype)
– shred and store in a structured way, applying DTD or Schema rules?
• Depends how it will be used:
– part of an existing rdbmd-based system which allows users to do ad hoc
SQL queries
– just to forward on, as XML, to other users but seldom queried
– Held for use by XPath conversant users
– heavily CRUDed
– As permanence for Web-based Distributed Authoring and Versioning
(DAV). Extensions to HTTP which allows users to collaboratively edit
and manage files on remote web servers.
School of Computing and Management Sciences
© Sheffield Hallam University
The future?
• Oracle Vs DB2 and SQL Server?
• In memory databases
– http://www.timesten.com/products/dataserver.html
• Post relational
– http://www.intersystems.com/cache/
• Shared processing models (grid)