Basic Oracle Architecture

Presented to HIOUG
21-Jun-2005
By Byron Bush
Workshop Objectives
Identify high-level architectural components of an
Oracle Database
Describe the function of each of these components
Identify Oracle Database internal data structures
Define the mechanisms used when processing
Insert/Update/Delete statements
Most people call it a “Database”
Technically an Oracle Database is broken into two high-
level components

1. INSTANCE – Non-persistent, memory-based

processes and structures
2. DATABASE – Persistent, disk-based data and
control files
So Why Differentiate?
Most Oracle installations consist of only a single
“Instance” and single “Database”…

Instance

Database
Oracle RAC is Different
Oracle “Real Application Clusters” allow multiple
“Instances” to interact with a single “Database” to
provide high availability…

Instance 1 Instance 2 Instance 3

Database
“Instance” a Little Closer
An Oracle Instance…
Is a means to access an Oracle Database
Always opens one and only one Database
Consists of memory and background process structures
 Picture of an “Instance”
Instance
SGA
Shared Pool

Library Database Buffer Redo Log

Cache Cache Buffer
Memory
Data Structures
Dictionary Java Pool Large Pool
Cache

Back-
PMON SMON DBWR LGWR CKPT Others Ground
Processes
“Database” a Little Closer
An Oracle Database…
Is a collection of data that is treated as a unit
Consists of three file types
 Picture of a “Database”

Oracle Database

Data Files Control Files Redo Log Files

Parameter Archived
File Log Files

Password
File
 Picture of an “Instance”
Instance
SGA
Shared Pool

Library Database Buffer Redo Log

Cache Cache Buffer
Memory
Data Structures
Dictionary Java Pool Large Pool
Cache

Back-
PMON SMON DBWR LGWR CKPT Others Ground
Processes
Memory Structure
Oracle’s memory structure consists of two memory areas
known as:
1. System Global Area (SGA): Allocated at instance
start up, and is a fundamental component of an
Oracle Instance
2. Program Global Area (PGA): Allocated when the
server process is started
System Global Area
The SGA consists of several memory structures:
Shared Pool
Database Buffer Cache
Redo Log Buffer
Other structures
There are two additional memory structures that
can be configured within the SGA:
Large Pool
Java Pool
System Global Area
The size of the SGA is determined by the parameters
that set the sizes of the various pools; these parameters
are dynamic
The SGA_MAX_SIZE parameter sets the maximum
size of the SGA (so you can limit it) and is not a
dynamic parameter

Thanks Ned!
 Picture of an “Instance”
Instance
SGA
Shared Pool

Library Database Buffer Redo Log

Cache Cache Buffer
Memory
Data Structures
Dictionary Java Pool Large Pool
Cache

Back-
PMON SMON DBWR LGWR CKPT Others Ground
Processes
Shared Pool
Used to store:
Most recently executed SQL statements
Most recently used data definitions
It consists of two key performance-related memory
structures:
Library Cache
Data Dictionary Cache
Sized by the parameter SHARED_POOL_SIZE
Library Cache
Stores information about the most recently used
SQL and PL/SQL statements
Enables the sharing of commonly used statements
Is managed by a least recently used (LRU)
algorithm
Consists of two structures
Shared SQL area
Shared PL/SQL area
Size is determined by the Shared Pool sizing
 Picture of an “Instance”
Instance
SGA
Shared Pool

Library Database Buffer Redo Log

Cache Cache Buffer
Memory
Data Structures
Dictionary Java Pool Large Pool
Cache

Back-
PMON SMON DBWR LGWR CKPT Others Ground
Processes
Data Dictionary Cache
A collection of the most recently used definitions in
the database
Includes information about database files, tables,
indexes, columns, users, privileges, and other database
objects
During the parse phase, the server process looks at the
data dictionary for information to resolve object names
and validate access
Database Buffer Cache
Stores copies of data blocks that have been retrieved
from the data files
Enables great performance gains when you obtain and
update data
Managed through an LRU algorithm
DB_BLOCK_SIZE determines primary block size
 Picture of an “Instance”
Instance
SGA
Shared Pool

Library Database Buffer Redo Log

Cache Cache Buffer
Memory
Data Structures
Dictionary Java Pool Large Pool
Cache

Back-
PMON SMON DBWR LGWR CKPT Others Ground
Processes
Redo Log Buffer
Records all changes made to the database data blocks
Primary purpose is recovery
Changes recorded within are called redo entries
Redo entries contain information to reconstruct or redo
changes
Size defined by LOG_BUFFER
Large Pool
An optional area of memory in the SGA
Relieves the burden place on the Shared Pool
Used for:
Session memory for the Shared Server
I/O server processes
Backup and restore operations for RMAN
Does not use an LRU list
Sized by LARGE_POOL_SIZE
Can be dynamically resized
Java Pool
Services parsing requirements for Java commands
Required if installing and using Java
Sized by JAVA_POOL_SIZE parameter
Oracle “Process” Structure
Oracle takes advantage of various types of Processes:
User Process: Started at the time a database user requests
connection to the Oracle Server
Server Process: Connects to the Oracle instance and is
started when a user establishes a session
Background Processes: Started when an Oracle instance
is started
 User Process
A program that requests interaction with the Oracle
server
Must first establish a connection
Does not interact directly with the Oracle server

User Server
Process Process
Connection
Established
 Server Process
A program that directly interacts with the Oracle server
Fulfills calls generated and returns results
Can be dedicated or shared server
 Picture of an “Instance”
Instance
SGA
Shared Pool

Library Database Buffer Redo Log

Cache Cache Buffer
Memory
Data Structures
Dictionary Java Pool Large Pool
Cache

Back-
PMON SMON DBWR LGWR CKPT Others Ground
Processes
Background Processes
Maintains and enforces relationships between physical
and memory structures
Mandatory background processes:
DBWn PMON CKPT
LGWR SMON
Optional background processes:
ARCn LMDn QMNn
CJQ0 LMON RECO
Dnnn LMS Snnn
LCKn Pnnn
Database Writer (DBWn)
Instance

Shared DBWn writes when:

Shared Pool
Pool SGA
 Checkpoint occurs
Library Database Buffer Redo Log
Cache Cache Buffer  Dirty buffers reach
Data threshold
Dictionary Java Pool Large Pool  There are no free buffers
Cache
 Timeout occurs

PMON SMON DBWR LGWR CKPT Others  Tablespace OFFLINE

 Tablespace READ ONLY
 Table DROP or

Oracle TRUNCATE
Database
Control Files Redo Log Files  Tablespace BEGIN
Data Files
BACKUP
Log Writer (LGWR)
Instance

Shared Pool SGA

LGWR writes:
 At commit
Library Database Buffer Redo Log
Cache Cache Buffer  When one-third full
Data  When there is 1 MB of
Dictionary Java Pool Large Pool redo
Cache
 Every three seconds

PMON SMON DBWR LGWR CKPT Others  Before DBWn writes

Oracle
Database
Control Files Redo Log Files
Data Files
System Monitor (SMON)
Instance

Shared Responsibilities:
Shared Pool
Pool SGA
 Instance recovery
Library Database Buffer Redo Log
Cache Buffer  Rolls forward changes in
Cache
online redo log files
Data
Dictionary  Opens database for user
Java Pool Large Pool
Cache access
 Rolls back uncommitted

PMON SMON DBWR LGWR CKPT Others transactions

 Coalesces free space
 Deallocates temporary
segments
Oracle
Database
Control Files Redo Log Files
Data Files
Process Monitor (PMON)
Instance

Shared Cleans up after failed

Shared Pool
Pool SGA
processes by:
Library Database Buffer Redo Log
Cache Cache Buffer  Rolling back the

Data transaction
Dictionary Java Pool Large Pool  Releasing locks
Cache
 Releasing other resources
 Restarting dead
PMON SMON DBWR LGWR CKPT Others
dispactchers

Oracle
Database
Control Files Redo Log Files
Data Files
Checkpoint (CKPT)
Instance

Shared Responsible for:

Shared Pool
Pool SGA
 Signaling DBWn at
Library Database Buffer Redo Log
Cache Cache Buffer checkpoints
Data  Updating datafile headers
Dictionary Java Pool Large Pool with checkpoint
Cache information
 Updating control files
PMON SMON DBWR LGWR CKPT Others
with checkpoint
information

Oracle
Database
Control Files Redo Log Files
Data Files
 Archiver (ARCn)
Optional background process
Automatically archives online redo log files when
ARCHIVELOG mode is set
Preserves the record of all changes made to the
database
Oracle
Database
Control Files Redo Log Files Archived Logs
Data Files
ARCn
Oracle Data Storage
Oracle keeps all system and user data in two basic
storage containers:
1. Tablespace… a logical entity known only to Oracle
2. Data Files… physical files that may (typically) be
seen from the operating system
Tablespaces
Can belong to only one database at a time
Consist of one or more data files
Are further divided into logical units of store
 Data Files
Can belong to only one tablespace and one database
Are a repository for schema object data

Database

Tablespace

Data Data
File File
2 Types of Tablespace
1. Tablespaces required by Oracle for normal
operations
2. Tablespaces that contain data and indexes that
support your applications
Required Tablespaces
System – Contains the Oracle Data Dictionary
Sysaux - New to 10g; supports historic monitoring /
tuning
Temporary – Used for disk-based sorting of data
(select...from…order by)
Undo – Used for transaction consistency during
Insert / Update / Delete statements
 SQL Statements (IUDS)
Insert
What happens when
Update
someone connects to
Delete
the database and
Select
issues one of these
statements?
 Connect to Oracle – What Happens
User Server
Process Process
Connect User/Pwd

Are “User” definition and privileges in SGA “Dictionary

Cache”?
If not, try to fetch from the “System” tablespace
If there, validate Pwd and privileges to login
Allow connection if all is right; otherwise decline
connection
 “Select” – What Happens
Select *
First?
From MyTab
Order by 1;

Is this statement in the “Library Cache”?

If statement cached then it has been recently been “parsed”
and “executed”
If not cached then “parse” the statement
If cached then skip parsing stage and execute statement
 “Select” – ParsingSelect
the* Statement
From MyTab
Order by 1;

Does “MyTab” exist?

Does user have Privs to select from “MyTab”
What columns are in “MyTab”
What is the first column in “MyTab”
 “Select” – Executing the
Select *
Statement
From MyTab
Order by 1;

Is “MyTab” data in the SGA “Buffer Cache”?

If not, fetch data into Buffer Cache
Sort data in “MyTab” by the first column
If can sort records in memory then do so
If cannot sort in memory then use “Temporary” tablespace
as disk-based staging area
Return records to client process
 “I/U/D” – What Happens First?
Insert (a,b,c)
Into MyTab;

Is this statement in the “Library Cache”?

If statement cached then it has been recently been “parsed”
and “executed”
If not cached then “parse” the statement
If cached then skip parsing stage and execute statement
 “I/U/D” – Executing the Statement
Insert (a,b,c)
Into MyTab;

Validate values to be inserted

Execute the statement
Keep DB version of the record in “Undo” tablespace until
Commit or Rollback
Record changes in SGA “Redo Log Buffer”
Change records in SGA “Buffer Cache”
DBWn writes changed records to data file(s) as part of
buffer cache management
 “I/U/D” – Executing the Statement
Insert (a,b,c)
Into MyTab;
Commit;
or
Rollback;

If “Commit” then…

LGWn writes “Redo Log Buffer” entries to “Redo Logs”
Undo entries are invalidated
If “Rollback” then…
Migrate DB version of record in Undo back to
tablespace/data file
 Archivelog Mode
If Instance in “Archivelog” mode Redo Log Files are
“Archived” to Archive Logs; which may be used to
recover your database in the case of disaster

Oracle
Database
Control Files Redo Log Files Archived Logs
Data Files
ARCn
 Summary
What we called a “Database” is really an “Instance”
and a “Database.”
An Oracle “Instance” is not persistent. It consists of
the System Global Area (SGA) and a collection of
processes.
An Oracle “Database” is persistent. It consists of a
collection of “Required” and “User” tablespaces (with
corresponding data files) along with other supporting
files.
Nearly every component of the “Instance” and
“Database” are mobilized to execute SQL statements.