PostgreSQL - Object Relational Database

Presenter: Mubashar Iqbal
Senior Software Engineer
Object Relational Database System

Introduction
The world's most advanced open source object-relational database system. The open
source Oracle. PostgreSQL has a large distributed developer and user community.
Community-owned with many companies involved.
Supported operating systems
● Linux
● Unix
● Mac OS X
● Solaris
● Windows
Native programming interfaces for:
C/C++, Java, .Net, Perl, Python, Ruby, PHP

Development Priorities
● Designed by/for Database Administrators
● Data integrity
● Security
● Reliability
● Standards
● DB Features
● Performance
● Ease-of-use
● Programmer Features

Most Common Uses
● ERP
● Data Warehouse
● Geographic
● OEM applications
● Network tools
● CRM

Prominent users
● Yahoo! for web user behavioral analysis, storing two petabytes and claimed to be
the largest data warehouse using a heavily modified version of PostgreSQL
● Sony Online multiplayer online games.
● Reddit social news website.
● Skype VoIP application, central business databases.
● Sun xVM, Sun's virtualization and datacenter automation suite.
● MusicBrainz, open online music encyclopedia.
● MyYearbook social networking site.
● Instagram, a popular mobile photo sharing service
● Disqus, an online discussion and commenting service

Features
● PostgreSQL often described as an open-source version of Oracle.
● BSD/MIT type license
● Reliability is PostgreSQL's top priority.
● Well-engineered, capable of supporting high-transaction and mission-critical
applications.
● Comprehensive documentation and manuals available for free online.
● Commercial support is available from independent vendors.
● PostgreSQL is fully ACID compliant.
● PostgreSQL is considered the solemn, full-featured, workhorse for transactional
enterprise applications, with strong ACID compliance.

Features (contd..)
● PostgreSQL supports one storage engine.
● SSL encryption
● Online backup
● Point-in-time recovery: Restore to any time in the past.
● Regular expression

Tools
● Psql: Command line front-end
● pgAdmin: GUI front-end
● phpPgadmin: Web based front-end
● MS ODBC
● MS Office + Postgres
● NaviCat: $$
● DeZign: $$
● EMS SQL Manager for PostgreSQL: $$

Data Types
● Numeric Types
● Character Types
● Hierarchical Types
● Binary Data Types
● Geometric Types
● Network Address Types
● Text Search Types
● UUID Type
● XML Type
● JSON Type
● Arrays
● Composite Types

Indexes
B-tree: B-trees can handle equality and range queries on data that can be sorted into
some ordering (<, <=,=,>=,>)
Hash: Hash indexes can only handle simple equality comparisons
GIN: GIN indexes are inverted indexes which can handle values that contain more than
one key, arrays for example, GIN operator classes for one-dimensional arrays
(<@,@>,=,&&)
GiST: Generalized Search Tree, it is a tree-structured access method and also known as
two-dimensional geometric data types (<@,@>,=,&&,>>,<<,&<,>&,~=)

Functions
A stored procedure and user-defined function is a set of SQL and procedural statements
(declarations, assignments, loops, flow-of-control) that stored on the database server and
can be invoked using the SQL interface.
CREATE FUNCTION function_name(p1 type, p2 type)
RETURNS type AS
BEGIN
-- logic
END;
LANGUAGE language_name;

Triggers
On DML (Data Manipulation Language) SELECT, INSERT, UPDATE, DELETE
CREATE TRIGGER
name { BEFORE | AFTER } { event [ OR ... ] }
ON
table [ FOR [ EACH ] { ROW | STATEMENT } ]
EXECUTE PROCEDURE
funcname ( arguments )

Cursors
● Used instead of FOR.
● Avoid memory overrun.
● Large data set.
DECLARE curs1 refcursor;
curs2 CURSOR FOR SELECT * FROM tenk1;
OPEN curs1 FOR SELECT * FROM foo WHERE key = mykey;
FETCH curs2 INTO foo, bar, baz;
CLOSE curs1;

View
View consists of a stored query accessible as a virtual table in a relational database or a
set of documents in a document-oriented database composed of the result set of a query.
Views are a great way to simplify your data model.
CREATE VIEW table_information AS
SELECT * FROM table WHERE id = 123;
Now you can simply query your new table directly:
SELECT * FROM table_information;

User-defined objects
New types of almost all objects inside the database can be created, including:
● Casts
● Conversions
● Data types
● Domains
● Functions, including aggregate functions and window functions
● Indexes including custom indexes for custom types
● Operators (existing ones can be overloaded)
● Procedural languages

Replication Methods
1. Master/Slave
● Asynchronous
● Synchronous
2. Multi-Master
● Asynchronous
● Synchronous
3. Proxy
4. Standby system

Master/Slave Replication
Asynchronous Synchronous
High availability High availibility
Read performance Better read performance
Offline peers Worse write performance
async
M S M S
sync

Multi-Master Replication
Asynchronous Synchronous
Read performance High availiability
Faster access across WANs Read performance
Manage offline peers Difficult to get good write performance
M M M M
async sync

Comparison of scaling behaviour

Hierarchical Database
Data is organized into a tree like structure.
Representing information using parent/child relationships.
Each parent can have many children, but each child has only one parent also known as a 1-
to-many relationship.
Different ways store data like this are
• Enumeration path (ltree)
• Adjacency List
• Nested Sets

LTree – Label Tree
● Ltree is a PostgreSQL module.
● It is implements a data type ltree for representing labels of data stored in a
hierarchical tree-like structure.
● Labels must be less than 256 bytes long. ltree stores a label path.
● A label path is a sequence of zero or more labels separated by dots.
● ltree supports several types of indexes that can speed up the indicated operators.
● Ltree performance is much better when you need to do ad-hoc queries over the tree
● Faster than recursive function that constantly needs to recalculate the branching.
● Some other databases have similar types. SQL Server 2008 has a datatype called
HierarchyID which serves the same purpose as ltree but with different syntax.

Example
Technique Adjacency Ltree
Query WITH RECURSIVE d AS (
SELECT id
FROM sponsorship WHERE id = 799
UNION ALL
SELECT s.id
FROM d JOIN sponsorship s ON
s.parent_fk = d.id
)
SELECT * FROM d ORDER BY id
LIMIT 100;
WITH p AS (
SELECT path FROM
sponsorship
WHERE id=799
)
SELECT s.id
FROM sponsorship s, p
WHERE s.path <@ p.path
ORDER BY s.id LIMIT 100;
Total Runtime 1946.48 ms 28.00 ms

More Details
1. Value Expression
http://www.postgresql.org/docs/9.2/static/sql-expressions.html
2. String Functions and Operators
http://www.postgresql.org/docs/9.2/static/functions-string.html
3. Mathematical Functions and Operators
http://www.postgresql.org/docs/9.2/static/functions-math.html
4. MySQL vs PostgreSQL
http://get.enterprisedb.com/whitepapers/Postgres_Plus_8.4_vs_MySQL_5.5.pdf
5. Scaling Behaviour
http://tweakers.net/reviews/657/1/database-test-dual-intel-xeon-5160-introduction.html

References
http://www.postgresql.org/
http://tweakers.net/reviews/657/5/database-test-dual-intel-xeon-5160-
comparison-of-scaling-behaviour.html
http://www.slideshare.net/petereisentraut/replication-solutions-for-postgresql
http://gbif.blogspot.com/2012/06/taxonomic-trees-in-postgresql.html
http://www.slideshare.net/vuhung16plus/postgre-sqlintroduction20100506

PostgreSQL - Object Relational Database

More Related Content

What's hot (20)

Viewers also liked (13)

Similar to PostgreSQL - Object Relational Database (20)

Recently uploaded (20)

PostgreSQL - Object Relational Database