- •About the Authors
- •Contents at a Glance
- •Contents
- •Table of Exercises
- •Introduction
- •Assessment Test
- •Answers to Assessment Test
- •What Is ASM?
- •Working with the ASM Instance
- •Overview of ASM Data Dictionary Views
- •Using ASM Storage
- •Using RMAN with ASM
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Understanding the Oracle Database as It Relates to Backup and Recovery
- •Performing Oracle Offline Backups
- •Performing Oracle Online Backups
- •Backing Up the Control File
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Performing Incomplete Recoveries
- •Performing Other Types of Recoveries
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Why Use RMAN?
- •Exploring the RMAN Architecture
- •Connecting to RMAN
- •Configuring RMAN for Use
- •Backing Up Your Database with RMAN
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Introducing the Recovery Catalog
- •Creating the Recovery Catalog User and Schema Objects
- •Using a Recovery Catalog
- •Maintaining the Recovery Catalog
- •Using the RMAN Virtual Private Catalog
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •RMAN Database-Recovery Basics
- •Using Image Copies to Recover Your Database
- •Other Basic Recovery Topics
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Switching Between RMAN Incarnations
- •Overview of RMAN Database Duplication
- •Performing an RMAN Tablespace Point-in-Time Recovery
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Overview of Flashback Technology
- •Using Automatic Undo Management
- •Using Flashback Technologies
- •Using Additional Flashback Operations
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Diagnosing the Database
- •Managing Database Performance
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Managing Memory
- •Managing Space
- •Managing Resources
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Automating Tasks with the Scheduler
- •Exploring the Scheduler Architecture
- •Exploring Common Administration Tools
- •Using Scheduler Jobs
- •Using Scheduler Programs
- •Using Schedules
- •Creating Lightweight Jobs
- •Using Job Chains
- •Using Scheduler Windows
- •Creating and Using Job Classes
- •Using Advanced Scheduler Concepts to Prioritize Jobs
- •Using Scheduler Views
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •An Overview of Globalization Support
- •Using NLS Parameters
- •Using Datetime Datatypes
- •Using Linguistic Sorts and Searches
- •Summary
- •Exam Essentials
- •Review Questions
- •Answers to Review Questions
- •Lab 1.1: Creating an ASM Instance
- •Lab 1.2: Creating ASM Disk Groups
- •Lab 1.3: Using ASM Disk Groups from a Database
- •Lab 2.2: Putting the Database in ARCHIVELOG Mode
- •Lab 2.3: Executing a Manual Online (Hot) Backup
- •Lab 3.1: Executing a Time-Based Point-in-Time Recovery
- •Lab 3.2: Recovering from Control-File Loss with a Backup Control File
- •Lab 3.3: Recovering from Loss of the Current Online Redo Log
- •Lab 4.1: Creating an RMAN Offline Backup
- •Lab 4.2: Creating an RMAN Incremental Backup
- •Lab 4.3: Creating an Image-Copy Backup
- •Lab 5.1: Implementing RVPC
- •Lab 6.1: Restoring a Datafile Online
- •Lab 6.2: Performing a Change-Based Recovery with RMAN
- •Lab 6.3: Restoring a Control File from an Autobackup
- •Lab 7.1: Monitoring RMAN Backups
- •Lab 7.2: One of My Backups Is Missing!
- •Lab 8.1: Duplicating a Database Using Active Database Duplication
- •Lab 8.2: Duplicating a Database Using Backup-Based Duplication to a Different Point in Time
- •Lab 9.1: Using the Recycle Bin
- •Lab 9.3: Using Flashback Data Archive
- •Lab 10.1: Using Support Workbench to Report a Problem to Oracle Support
- •Lab 11.1: Exporting a Transportable Tablespace
- •Lab 11.2: Testing Resumable Space Allocation
- •Lab 11.3: Manually Configuring the SGA
- •Lab 12.1: Creating a Local External Job
- •Lab 12.2: Creating a Job Window
- •Lab 13.1: Using the Locale Builder to Create a New Linguistic Sort
- •Lab 13.2: Setting NLS Parameters
- •Lab 13.3: Performing Linguistic Sorts
- •What You’ll Find on the CD
- •System Requirements
- •Using the CD
- •Troubleshooting
- •Glossary
- •Index
Glossary
788 Glossary
A
active database duplication Active database duplication is an RMAN duplication process that occurs over the network and does not use RMAN backups as the source of the duplication.
active online redo log group This is an online redo log that is not currently in use by the database but has not been archived.
archived redo logs Copies of the online redo logs. Critical to database recovery when the database is in ARCHIVELOG mode.
ARCHIVELOG mode When in ARCHIVELOG mode, the database will generate archived redo logs and, by applying those files and the online redo logs, can be recovered to any point other than the point of the last backup.
archiver process (ARCH) The process responsible for copying an online redo log to an archived redo log after a log switch has completed.
ASM See Automatic Storage Management.
ASM fast disk resync Method of quickly recovering from hardware failures that impact disk availability but do not corrupt the data on the disk itself.
ASM instance Running Oracle instance specific to ASM functionality.
ASM preferred mirror read When one set of disks is local and the other remote, this allows you to indicate to ASM that it should read from a specific failure group set of disks (typically the local set).
ASM redundancy Method of protecting data on ASM disks by mirroring the data on one or more ASM failure groups.
ASM_DISKSTRING Parameter that indicates to the Oracle ASM instance where to look for ASM disks upon instance startup.
ASMCMD Command -line tool used to manage ASM instances.
automated channel failover The ability of other channels to automatically restart failed RMAN activities from a channel that has failed.
Automatic Diagnostic Repository (ADR) File -based repository for database diagnostic data.
automatic instance Temporary instance created automatically when performing RMAN tablespace point-in-time recovery.
Automatic Memory Management Default memory-management model in Oracle 11g; when you set a nonzero value for MEMORY_TARGET and MEMORY_MAX_TARGET, Oracle will manage SGA and PGA memory pools, caches, and work-area sizes dynamically.
Glossary 789
Automatic PGA Memory Management When you set the value of MEMORY_TARGET and MAX_ MEMORY_TARGET to zero and set a nonzero value for PGA_AGGREGATE_TARGET, Oracle will manage PGA memory components automatically, resizing as needed within the value specified.
Automatic Shared Memory Management When you set the value of MEMORY_TARGET and
MAX_MEMORY_TARGET to zero and set a nonzero value for SGA_MAX_SIZE and SGA_TARGET, Oracle will manage SGA memory components automatically, resizing as needed within the value specified.
Automatic Storage Management An Oracle native file system management system that provides a volume manager, fault tolerance, and load balancing of disks assigned to Oracle databases. Both stand-alone and clustered databases are supported.
Automatic Undo Management (AUM) Oracle functionality that simplifies management of undo tablespaces. AUM works to meet undo retention goals while ensuring that adequate space is available for new transactions.
auxiliary instance Instance manually created when database duplication is performed or automatically created during tablespace point-in-time recovery. This instance is used as the destination database and will become the duplicated database during a database duplication, or is temporarily used during a tablespace point-in-time recovery.
auxiliary set The set of database datafiles required to create the automatic instance during a tablespace point-in-time recovery.
B
backup-based database duplication Database duplication dependent on the presence of RMAN backups and archived redo logs.
backup optimization When enabled, will prevent unnecessary backups of read-only database datafiles.
backup set A logical entity representing a single backup of specific database tablespace datafiles. Individual database datafiles are always contained within a single RMAN backup set. A backup set comprises one or more physical files called backup-set pieces.
backup-set compression Zip -like compression of RMAN backup sets used to reduce the size of the RMAN backup-set pieces, and thus the entire backup.
backup-set piece Default physical file used by RMAN to back up a database. Backup-set pieces are physical files. Many backup-set pieces may belong to one logical structure called a backup set.
binary sorting Ordering character strings based on their binary coded values.
790 Glossary
block-change tracking file Physical file used to track changed blocks. Level-1 incremental backups use these files to improve performance of incremental backups by avoiding unnecessary datafile IO.
block media recovery A technique for restoring and recovering individual data blocks that have been identified as corrupt while all database files remain online and available.
byte semantics Assumes a single-byte character set, where one character always requires one byte of storage; treatment of strings as a sequence of bytes.
C
chain Two or more Scheduler programs that are linked together to meet an objective, where the outcome of one job determines the next steps in the chain.
change-based recovery Recovery of the database based on a specific system change number (SCN).
channel A connection from the database server to the backup destination (disk or MML layer). Each RMAN backup or recovery has at least one channel allocated. Multiple channels provide for parallel backup and recovery in RMAN.
character semantics Allows character data to be specified in terms of the number of characters regardless of the number of bytes required.
character set A collection of elements that represent textual information for a specific language or group of languages.
complete database recovery Recovering the database to the point of failure by applying all archived redo logs and all redo in the online redo logs.
complete recovery A recovery of the database to the point of the last completed transaction. This kind of recovery involves no data loss. Complete recovery is synonymous with point-of-failure recovery.
consistent shutdown A shutdown of the database that leaves the datafiles in a consistent state.
control file A critical database file that contains metadata related to the database, such as the location of the database datafiles, redo logs, and archived redo logs.
control-file autobackups An RMAN feature providing automatic backup and restore of the database control files and spfiles. Also provides for easy recovery of these databaserelated files.
convert The RMAN command that allows the DBA to transport databases and tablespaces from one platform to another where the source and destination platforms have different endianness.
Glossary 791
corrupt block An Oracle block that is not in a recognized Oracle format or whose contents are not internally consistent. Corruption is usually caused by hardware or operating-system problems. In Oracle, block corruption is classified as either logical, caused by an Oracle internal error, or physically corrupt media, meaning the block format is not correct.
crash or instance recovery A recovery from a database that has been shut down in an inconsistent manner. This kind of recovery does not require user intervention because the Oracle database uses the online redo logs to bring the database to a consistent state.
cumulative incremental backup An incremental backup that contains all database data blocks changed since the last level 0 backup.
current online redo log group The online redo log group that is currently in use by the database.
D
datafile recoveries The recovery of specific database datafiles instead of the entire database or a specific tablespace.
database A collection of datafiles that is used to store data.
database character set The character set that will govern default text storage in the database.
database control file The file that contains configuration information on the database associated with the control file, such as the location of datafiles, redo logs, and RMANrelated information.
database datafiles The physical media used to store database data. Tablespaces are assigned to one or more database datafiles.
database duplication The process of creating one database from another.
database incarnation Indicates the logical life of a given database. The first incarnation begins at the creation of a database and ends at the point the resetlogs command is used when creating a database.
database parameter file (pfile) A text-based file that contains database-related parameters and their settings.
Database Replay An Oracle feature that allows the capture of a workload and replay on a similar database in a test environment.
Database Resource Manager The Oracle feature that allows the DBA to manage resource allocation by creating directives, plans, and groups.
database writer process (DBWR) The process responsible for writing database changes to the database datafiles.
792 Glossary
DBID Stands for database ID. This number uniquely identifies an Oracle database.
dependency scheduling Scheduling jobs based on the outcomes of previous jobs. In the Oracle Scheduler, job chains are used to implement dependency-based scheduling.
destination host The host that is the destination of a duplicate operation.
differential incremental backup An incremental backup that contains all database data blocks changed since the last level-1 differential incremental backup.
disk group Equivalent to a logical volume in ASM. Individual LUNs (disks) are assigned to disk groups. Fault tolerance is supported at the disk-group level through mirroring.
disk-group attributes Attributes assigned individually to specific disk groups, such as disk repair time.
duplexing A method of creating more than one copy of a database backup during the backup.
dynamic performance data dictionary views Views that provide near-real-time information on the database, including metadata about database structures. Much of the data in dynamic performance views comes from the database control file.
E
endianness Or endian format. The byte order used by a particular hardware platform. When the sequence of bytes with increasing significance is stored with increasing memory addresses, this is referred to as little endian. If the most significant byte is stored first, this is referred to as big endian. When bytes are exchanged between platforms with different endianness, a conversion process must occur before the bytes will make sense to the destination computer.
event-based scheduling The concept of executing a job based on an event that signals the Scheduler, regardless of time or dependencies.
F
failure groups (or FAILGROUPS) Failure groups provide for data redundancy in an ASM disk group. Each failure group is assigned one or more disks. Data in one failure group is redundantly copied to the other failure groups in the disk group.
flash recovery area (FRA) File system dedicated to back-up and recovery purposes. Can contain many recovery-related components, such as backup set pieces, archived redo logs, and copies of the database control files.
Glossary 793
Flashback Data Archive An Oracle feature that allows one to archive all DML changes to a table and retain those changes for a specified retention period. The archive is queryable; the retention period is automatically enforced.
Flashback Database An Oracle feature that allows a point-in-time logical recovery of the entire database to a timestamp, SCN, or named recovery point. Flashback Database uses flashback logs stored in the flash recovery area. It can’t be used for media recovery.
Flashback Drop An Oracle feature that allows dropped objects to be recovered from the Recycle Bin instead of requiring an expensive incomplete recovery process to be performed.
Flashback Table An Oracle feature that allows tables (and their dependent objects) to be recovered to an earlier point in time while the tablespace remains online and without the overhead of an incomplete recovery.
Flashback Transaction Query An Oracle feature that allows users to identify changes made to tables at a transactional level.
Flashback Versions Query A feature that allows users to view all versions of data that have existed over a period of time for a specific table.
I
image copies Exact byte-for-byte copies of database datafiles backed up by RMAN. Image copies are made with the backup as copy command. Can be used for quick database restores.
inactive online redo log group One of the online redo log groups of the database that is not in use and has been archived.
incident In Oracle 11g, a single occurrence of a problem.
Incident Packaging Service (IPS) Enables you to automatically gather the diagnostic data pertaining to a critical error into a zip file for transmission to Oracle Support.
incomplete recovery Also called point-in-time recovery. Recovery of the database to a earlier point in time, SCN, or log sequence number than the current ones.
inconsistent shutdown A shutdown of the database that leaves the datafiles in an inconsistent state. An inconsistent shutdown of the database will definitely result in an instance recovery. In some cases, an inconsistent shutdown (due to loss of a datafile, for example) could lead to media recovery.
incremental database backups Backups of an Oracle database by RMAN. Includes a level-0 backup, which is a full backup of the database, and a level-1 backup, which incrementally backs up only changed database blocks.
794 Glossary
incrementally updated backups Combination of a backup image copy and incremental backups that allow for very fast recovery of the Oracle database.
instance Collection of database processes and memory.
J
job An instruction to the Scheduler to execute a specific program at a specific time on a specific date, for example.
job class A logical way of grouping jobs that have similar business or performance attributes.
job coordinator The Oracle background process that is responsible for ensuring that jobs are run on schedule.
job table The master table in the Oracle database that contains the information for all enabled jobs in the database.
L
length semantics Determines how you treat the length of a character string. The length can be treated as a sequence of characters or bytes.
level-0 incremental backup Essentially, a full backup of the Oracle database. A level-0 incremental backup is required to be able to perform subsequent level-1 incremental backups. Whole-database backups are not the same as incremental level-0 backups.
level-1 incremental backup Backups of changed blocks in Oracle databases. Can be based on either a differential backup strategy or a cumulative incremental strategy.
log sequenced–based recovery Recovery of the database based on a specific log sequence number.
log sequence number point-in-time recovery A method of incomplete recovery used to restore the database to a specific log sequence number.
log writer process (LGWR) The process responsible for writing redo data from the redo log buffer in the SGA to the online redo logs of the database.
logical corruption Data inconsistencies caused by user error (where a user can be a user, developer, DBA, or program that modifies data in the database).
Glossary 795
M
Media Management Library (MML) An Oracle-created API that allows media-product vendors to write interfaces into RMAN. RMAN channels that are allocated to SBT instead of disk will be directed to the MML layer. This layer must first be configured according to vendor instructions.
media recovery Recovery of the database, typically by means of database backups, that requires DBA intervention.
MOUNT mode One of three modes that the database can be in. When the database is open in MOUNT mode, it has read the database control file but not yet opened the database datafiles.
multibyte character A character whose character code consists of two or more bytes under a certain encoding scheme.
multiplexing The ability to create more than one copy of a database backup to a different location (but on the same type of media). It is essentially making parallel copies of the backup.
multiselection backups Multiselection backups provide the ability to chunk up large database datafiles into individual backup channels. This allows for parallelization of backups of individual datafiles.
N
national character set An character set that is an alternative to the database character set that governs NCHAR, NVARCHAR2, and NCLOB data.
National Language Support (NLS) Allows users to interact with the database in their native language.
NOARCHIVELOG mode When in NOARCHIVELOG mode, the database will not generate archived redo logs and can not be recovered to any point other than the point of the last backup.
O
online datafile recovery Recovery of a database datafile while the database is open.
online redo logs Persistent mechanism that stores redo copied from the redo-log buffer by the LGWR process.
online tablespace recovery Recovery of all tablespace datafiles while the database is open.
796 Glossary
Oracle data dictionary Views in the database that provide metadata information about the database, including database configuration, objects, users, and other information.
Oracle Database Configuration Assistant (DBCA) Oracle graphical interface used to create both ASM instances and Oracle databases.
orapwd The program that creates the Oracle database password file.
P
parameter file A file that defines global database settings such as memory allocations. Can be either a pfile, which is text-based, or an spfile, which is managed by the Oracle server.
point-in-time recovery Recovery of the database based on a specific point in time. Also called incomplete recovery. Recovery of the database to an earlier point in time, SCN, or log sequence number than the current ones.
point-of-failure recovery This is a recovery of the database to the point of the last completed transaction. This kind of recovery involves no data loss. Point-of-failure recovery is synonymous with complete recovery.
problem In Oracle 11g, a critical error in the database.
program Defines the action that will occur when a job runs.
R
recover command RMAN command used to recover a database, tablespace, or datafile. Causes the application of incremental backups and archived redo logs to complete the database recovery.
recovery catalog The recovery catalog is an optional database schema that maintains a record of all RMAN backup operations.
recovery catalog stored scripts These are scripts stored in the recovery catalog. These scripts can be called by RMAN for backup, recovery, or reporting purposes.
recovery catalog views Views that can be queried by the DBA to look at RMAN-related metadata in the recovery catalog.
recovery set Set of tablespaces to be recovered during tablespace point-in-time recovery.
Recycle Bin A logical container that stores objects dropped from the database. The Recycle Bin (in conjunction with the Flashback Drop feature) offers users a simple method of querying and recovering objects that may have been dropped by accident.
Glossary 797
redo-log buffer The memory area in the SGA to which redo is initially written.
redo log sequence number A unique number assigned to each online redo log to define the order in which it was written to.
redo logfile group A set of one or more online redo logs. Each redo logfile group is written to one at a time and may have one or more copies of the redo log called members.
Redo logfile member One or more files in a redo logfile group. These files are written to in parallel and are used to protect each online redo log from failure.
resource consumer group A logical grouping of users based on resource-consumption requirements and business needs.
resource plan A group of resource-plan directives that specify how resources should be distributed among the consumer groups.
resource-plan directive Defines resource allocation rules and connects resource plans to consumer groups.
restore command RMAN command used to restore the database from a previous RMAN backup. This command will cause RMAN to restore datafiles from backup-set pieces or image copies.
resumable space allocation The Oracle feature that enables transactions to suspend and wait for a space condition to be resolved within a specified time without aborting the transaction. When the space condition is resolved, the transaction will resume.
RETENTION GUARANTEE An option that, when enabled, will guarantee that unexpired undo records will never be removed from the undo tablespace. They will be maintained until they expire (at the end of the retention period), even at the expense of new transactions failing because of lack of undo space.
retention policies Retention policies determine how long database backups will be considered valid in RMAN. After the retention period expires, backups will be marked as obsolete and be eligible for removal.
RMAN Oracle’s provided backup and recovery tool.
RMAN backup-format specification The backup-format specification is used to indicate the file-naming convention to be used when creating an RMAN backup-set piece.
RMAN command-line interface (RCLI) Used to access RMAN and perform RMANrelated activities.
RMAN persistent configuration settings Provide the ability to configure backupand recovery-related settings that become the default value for RMAN backup and recovery operations.
798 Glossary
RMAN virtual private catalog An optional feature of the RMAN recovery catalog that provides the ability to limit access to records in the recovery catalog to specific sets of users.
run block A block of RMAN commands that starts with a run command. The commands are enclosed in braces.
S
schedule A schedule contains a start date, an optional end date, and a repeat interval.
SBT Device designation that indicates the backup, restore or maintenance operation will use the MML layer instead of a disk device.
SCN-based point-in-time recovery A method of recovery that restores the database to a specific point in time based on a database SCN.
Segment Shrink An online Segment Shrink operation reduces the size of a segment by moving rows and consolidating the space used, eliminating unused space above and below the high-water mark.
service request A request to Oracle Support to assist with a technical problem.
snapshot control file A consistent copy of the control file created at the beginning of an RMAN backup operation.
source host The database host machine where the target database is located.
space pressure A situation that occurs when Oracle cannot allocate any further extents in a tablespace without extending the tablespace. When this situation occurs, Oracle will purge objects from the Recycle Bin rather than extend the tablespace.
SQL Access Advisor The SQL advisor that analyzes the schema design for a workload and recommends indexes, partitions, and materialized views to improve performance.
SQL Tuning Advisor One of the SQL advisors; it takes one or more SQL statements as input and produces tuning advice.
SQL tuning set (STS) A database object that stores a set of SQL statements along with their execution context and statistics.
static data dictionary views Data dictionary views that provide metadata information on various database structures such as tables, indexes, and other database objects.
Glossary 799
substitution variables Variables used in place of literal values. Each time you execute code with the substitution variable, you can indicate the value of that variable or RMAN will prompt you for the value of that variable when you run the code.
Support Workbench An environment within Oracle Enterprise Manager that provides a workflow for investigating, diagnosing, reporting, submitting service requests to Oracle Support, and resolving problems.
suspended transaction A transaction that has encountered a space condition and is waiting for the space condition to be resolved. The transaction will abort if the suspend condition is not resolved within the time-out.
system change number (SCN) An internal counter that is used to maintain the order and dependency of transactions within a given database.
T
tablespace A logical storage area that is assigned to one or more database datafiles. Oracle objects (such as tables) are assigned to a tablespace when they are created.
tablespace point-in-time recovery The process of recovering one or more tablespaces in a database to a previous point in time.
tablespace recoveries See online tablespace recovery.
tag A specific name given to an RMAN backup. It can be referenced during subsequent RMAN operations such as recoveries.
time-based point-in-time recovery A method of recovery that restores the database to a specific point in time based on a time defined in the restore and recover commands.
transportable database In Oracle 11g, the feature that allows the DBA to copy an entire database from one platform to another.
transportable tablespace set The datafiles associated with a group of tablespaces along with the exported metadata that encapsulates the self-contained set of objects that are to be transported.
U
undo record A row stored in the undo tablespace that contains the data necessary to undo a transaction (or a piece of a transaction).
UNDO_RETENTION An Oracle parameter that governs the length of time that undo records will be retained in the undo tablespace after their associated transaction has completed.
Unicode The universal character set that supports all known written languages.
800 Glossary
W
white-space compression RMAN’s default behavior, attempting to reduce the size of backup sets by not including blocks that are unused in the backup.
whole-database backup A complete backup of an Oracle database using the backup command.
window A recurring block of time during which a specific resource plan should be enabled to govern resource allocation for the database.
window group A logical method of grouping windows to simplify management by allowing them to be managed as one object.