oracle 8 database administration volume 1 instruction guide phần 2 potx

Some of these are: • Parameter file: used to define the characteristics of an Oracle instance • Archived redo log files: offline copies of the redo log files that may be necessary to rec

Trang 1

Connecting to a Database

Apart from the database files, an Oracle server also uses other files Some of

these are:

• Parameter file: used to define the characteristics of an Oracle instance

• Archived redo log files: offline copies of the redo log files that may be

necessary to recover from media failures

The following sections of this lesson explain how a server process uses some

of the components of the Oracle instance and the database to service

requests made by a user process

Database

Other Key Physical Structures

Archived log files

Password

file

Parameter

file

Trang 2

Running a Query

The following are the main stages in the processing of a query:

1 Parse: In this stage, the user process sends the query to the server process

with a request to parse or compile the query The server process checks

the validity of the command and uses the area in the SGA known as the

shared pool to compile the statement At the end of this phase, the server

process returns the status—that is, success or failure of the parse

phase—to the user process

2 Execute: During this phase in the processing of a query, the server

process prepares to retrieve the data

3 Fetch: The rows that are retrieved by the query are returned by the server

to the user during this phase Depending on the amount of memory used

for the transfer, one or more fetches are required to transfer the results of

a query to the user

Instructor Note

This slide has a build There are three stages

User process

Server process

Processing a Query

Parse

Statement

OK Handle

ExecuteOK

Handle

FetchResults

SELECT *

FROM emp

ORDER BY ename;

Trang 3

Running a Query

The shared pool is a part of the SGA that is used during the parse phase The

size of the shared pool is specified by the initialization parameter,

SHARED_POOL_SIZE in the parameter file The components of the shared

pool described below are used to parse an SQL statement

Library Cache

The library cache stores the following information about the most recently

used SQL statements:

statement as determined by the optimizer

Since the library cache stores this information, if a query is reexecuted

before its execution plan is aged out by other statements, the server process

does not need to parse the statement Thus the library cache helps improve

the performance of applications

The Shared Pool

• Size defined by SHARED_POOL_SIZE

• Library cache contains statement text,

parsed code, and an execution plan

• Data dictionary cache contains table

and column definitions and privileges

Shared pool

Library cache

Data dictionary cache

Trang 4

Data Dictionary Cache

The data dictionary cache, also known as the dictionary cache or row cache,

is a part of the shared pool that stores the most recently used data dictionary

information, such as table and column definitions, usernames, passwords,

and privileges

During the parse phase, the server process looks for the information in the

dictionary cache to resolve the object names specified in the SQL statement

and to validate the access privileges If necessary, the server process initiates

the loading of this information from the data files

Trang 5

Running a Query

When a query is processed, the server process looks for the blocks needed in

the database buffer cache If the block is not found in the database buffer

cache, the server process reads the block from the data file and places a copy

in the buffer cache Subsequent requests for the same block are likely to find

the block in memory, and may not require physical reads

Function of the Database Buffer Cache

The database buffer cache, or simply the buffer cache, is an area in the SGA

that is used to store the most recently used data blocks The size of each

buffer in the buffer cache is equal to the size of a data block, and it is

specified by the DB_BLOCK_SIZE parameter The number of buffers is

equal to the value of the DB_BLOCK_BUFFERS parameter The Oracle

server uses a least recently used (LRU) algorithm to age out buffers that

have not been accessed recently to make way for new blocks to be

accommodated in the buffer cache

Database Buffer Cache

• Number of buffers defined by

DB_BLOCK_BUFFERS

• Size of a buffer based on DB_BLOCK_SIZE

• Stores the most recently used blocks

Trang 6

The Program Global Area (PGA) is a memory region that contains data and

controls information for a single server process or a single background

process In contrast to the SGA, which is shared and written to by several

processes, the PGA, which is also called the process global area, is an area

that is used by only one process When using the dedicated server

configuration, the PGA contains:

are processed or returned to the user

• Cursor state: which indicates the stage in the processing of various

cursors that are currently used by the session

The PGA is allocated when a process is created and deallocated when the

process is terminated

Note

Some of these structures are stored in the SGA when using an MTS

Program Global Area (PGA)

• Not shared and not writable

Trang 7

What follows are the steps used in executing the update statement:

1 The server process reads the data and rollback blocks from the data files,

if they are not already in the buffer cache

2 Copies of the blocks that are read are placed in the buffer cache

Instance SGA Shared pool

Library cache Redo log

buffer

Database buffer cache

Data dictionary cache

Control files

Redo log files Data files

Database

2

3 4 5

Trang 8

4 The server process records the changes to be made to the rollback

(before-image) and to the data (new value) in the redo log buffer

updates the data block, both in the database buffer cache Both the

changed blocks in the buffer cache are marked as dirty buffers—that is,

buffers that are not the same as the corresponding blocks on the disk

Note

The processing of a DELETE or INSERT command uses similar steps The

before-image for a delete contains the column values in the deleted row,

while inserts just require the row location information to be stored in the

rollback

Instructor Note

• The slide explaining the processing of a DML command has builds

There are five stages

• You may want to ask the students what the DML commands are In the

current context we are referring to only the INSERT, UPDATE, and

DELETE commands

Trang 9

Running a DML Statement

Before making a change, the server process saves the old value into a

rollback segment This image is used to:

the DML statement (read consistency)

• Recover the database to a consistent state in case of failures

Rollback segments, like tables and indexes, exist in data files and parts of

them are brought into the database buffer cache when required

Rollback Segment

DML statement

Old image

New image Rollback

segment

Table

Trang 10

The server process records changes made by an instance in the redo log

buffer, which is a part of the SGA The redo log buffer has the following

characteristics:

• It stores redo records, which record changes—that is, the block that is

changed, the location of the change, and the new value A redo entry

registers a change, but it makes no distinction between the type of block

that is changed So it cannot distinguish, for example, a change to a data

block from a change to an index or a rollback block

transaction may be interleaved with changes made by other transactions

• It is a circular buffer that is reused after it is filled up, but only after all

the old redo entries are recorded in the redo log files

Note

The redo log files are covered in more detail in the lesson “Maintaining

Redo Log Files.”

Redo Log Buffer

• Size defined by LOG_BUFFER

• Records changes made through the

instance

• Used sequentially

• Circular buffer

Trang 11

Running a DML Statement

The server process records changes to rollback and data blocks in the buffer

cache The database writer (DBWR) writes the dirty buffers from the

database buffer cache to the data files It ensures that sufficient number of

free buffers—buffers that can be overwritten when server processes need to

read in blocks from the data files—are available in the database buffer cache

Database performance is improved because server processes only make

changes in the buffer cache, and the DBWR defers writing to the

data files until one of the following events occur:

buffers and cannot find any

of the database, (Checkpoint is a means of synchronizing the database

buffer cache with the data file.)

Note

Checkpoints are covered in detail in the lesson “Managing Redo Log Files.”

DBWR

Database buffer cacheDatabase Writer (DBWR)

Control files

Trang 12

The Log Writer (LGWR) is a background process that writes entries from the

redo log buffer into the redo log files The LGWR performs sequential

writes to the redo log file under the following situations:

data files

Instructor Note

You may want to inform the students that although there is a timeout, on a

busy OLTP system, this may never occur because other events may occur

more frequently

Redo log buffer

LGWRLog Writer (LGWR)

Control files Redo log files Data files

Trang 13

COMMIT Processing

COMMIT Processing

Oracle uses a fast commit mechanism that guarantees that the committed

changes can be recovered in case of failures

System Change Number

Whenever a transaction commits, the Oracle server assigns a commit System

Change Number (SCN) to the transaction, which is monotonically

incremented and is unique within the database The SCN is used by the

Oracle server as an internal “time stamp” to synchronize data and to provide

read consistency when data is retrieved from the data files Using the SCN

allows the Oracle server to perform consistency checks without depending

on the date and time of the operating system

Steps in Processing COMMITs

When a COMMIT is issued, the following steps occur:

redo log buffer

to and including the commit record to the redo log files After this point,

the Oracle server can guarantee that the changes will not be lost even in

case of failures

Shared pool

Redo log buffer

LGWR

Control files

SGA Instance

Trang 14

4 The server process records information to indicate that the transaction is

complete and that resource locks can be released

Flushing of the dirty buffers to the datafile is performed independently by

DBWR, and can take place either before or after the commit

Note

Rolling back a transaction does not trigger LGWR to write to disk The

Oracle server always rolls back uncommitted changes when recovering from

failures If there is a failure after a rollback, before the rollback entries are

recorded on disk, the absence of a commit record is sufficient to ensure that

the changes made by the transaction are rolled back

Advantages of the Fast COMMIT

The use of these steps for processing COMMITs has the following

advantages:

• Sequential writes to the log files are faster than writing to different

blocks in the data file

to the log files, whereas writing to the data files would require whole

blocks of data to be written

transactions requesting to commit at the same time into a single write

• Unless the redo log buffer is particularly full, only one synchronous

write is required per transaction (can be less than one synchronous write

per transaction)

an actual COMMIT operation

Instructor Note

• Even though LGWR may write through the operating system buffer

cache, synchronous writes are used to ensure that confirmation of a

commit is only made after data is written to disk

• The COMMIT processing slide has builds There are four stages

Trang 15

Summary

Instructor Note

• Lessons have a summary showing the following:

- A summary slide highlighting the most important objectives for the

lesson

- A Quick Reference listing the initialization parameters, commands,

and packages discussed in the lesson

• You may want to ask the students how users specify the database they

want to connect to, and emphasize the terms instance and database.

Database

Control files

Data files Redo log

files

User

process

Server process PGA

Summary

Password file

Archived log files Parameter

file

Trang 16

Quick Reference

Server process Background processes:

- DBWR

- LGWR

- Shared pool

- Database buffer cache

- Redo log buffer PGA

Physical structures (files) Database files:

Trang 17

2

Using Administration

Tools

Trang 19

Objectives

Instructor Note

The purpose of this lesson is to ensure that participants get acquainted with

the use of the following tools:

• Server Manager Line Mode

• Oracle Enterprise Manager

It is not intended to provide details of all the components of OEM

Objectives

• Using the Server Manager Line Mode

• Identifying administration applications

supplied with the Oracle Enterprise

Manager

• Using Oracle Enterprise Manager

components

Trang 20

The table lists the common database administration tools and tasks that a

database administrator can perform using the tools

These tools will be covered in this course, but they are only a subset of the

utilities supplied by Oracle The commands for starting up the different tools

depend on the platform

This lesson will introduce the use of the following tools that assist a DBA in

performing administrative tasks:

Note

In addition to these tools, NT-specific utilities like ORADIM80 or Oracle

Database Assistant for creation of an Oracle database will be covered in the

lessons “Managing an Oracle Instance” and “Creating a Database.”

Export or Import utility

Password File utility

Description

A line mode utility used for administrative tasks like starting up, shutting down or recovering a database

Graphical user interface to administer, monitor, and tune multiple databases.

Utility for loading data from external files into Oracle tables Utility for exporting/importing data in Oracle format

Utility for creating database password file

Database Administration

Tools: Examples

Trang 21

Using the Server Manager

Using the Server Manager

Server Manager line mode is useful for performing unattended operations,

such as running nightly batch jobs or scripts In addition, you can use line

mode when a graphical interface is unavailable

Starting Server Manager in Line Mode

Start Server Manager in line mode by entering the appropriate command at

the operating system prompt For example, on some systems the command

svrmgrl starts up Server Manager in line mode.

You can also start Server Manager in line mode and execute a script

Note

The exact command for starting Server Manager in line mode depends on

your platform For information about starting Server Manager in line mode,

see your operating system–specific Oracle documentation

Starting Server Manager in

Line Mode

Starting and executing a script:

svrmgrl command=@credb.sql svrmgrl command=“CONNECT scott/tiger”

Tiêu đề	Oracle 8 Database Administration Volume 1 Instruction Guide Part 2 Potx
Trường học	University of Example
Chuyên ngành	Database Administration
Thể loại	instruction guide
Năm xuất bản	1998
Thành phố	Unknown

Định dạng
Số trang	40
Dung lượng	794,59 KB