2 Oracle Database High Availability Solutions

2.1.1 Oracle Real Application Clusters

Oracle Real Application Clusters (RAC) allows the Oracle database to run any packaged or custom application unchanged across a set of clustered servers. This capability provides the highest levels of availability and the most flexible scalability. If a clustered server fails, the Oracle database will continue running on the surviving servers. When more processing power is needed, another server can be added without interrupting user access to data.

RAC enables multiple instances that are linked by an interconnect to share access to an Oracle database. In a RAC environment, the Oracle database runs on two or more systems in a cluster while concurrently accessing a single shared database. The result is a single database system that spans multiple hardware systems yet appears as a single unified database system to the application. This enables RAC to provide high availability, scalability, and redundancy during failures within the cluster. RAC accommodates all system types, from read-only data warehouse (DSS) systems to update-intensive online transaction processing (OLTP) systems.

High availability configurations have redundant hardware and software that maintain operations by avoiding single points-of-failure. To accomplish this, the Oracle Clusterware is installed as part of the RAC installation process. Oracle Clusterware is a portable solution that is integrated and designed specifically for the Oracle database. In a RAC environment, Oracle Clusterware monitors all Oracle components (such as instances and listeners). If a failure occurs, Oracle Clusterware will automatically attempt to restart the failed component. Other non-Oracle processes can also be managed by Oracle Clusterware. During outages, Oracle Clusterware relocates the processing performed by the inoperative component to a backup component. For example, if a node in the cluster fails, Oracle Clusterware will cause client processes running on the failed node to reconnect and resume running on a surviving node.

The Oracle Clusterware requires two files, the Oracle Cluster Registry (OCR) and the voting disk. To avoid single points-of-failure, the Oracle Clusterware automatically maintains redundant copies of these files. Oracle Clusterware also enables you to replace a damaged copy of the OCR online. Oracle's recovery processes quickly re-master resources, recover partial or failed transactions, and rapidly restore the system.

RAC provides the following benefits:

• Ability to tolerate and quickly recover from computer and instance failures

• Fast, automatic, and intelligent connection and service relocation and failover

• Rolling patch upgrades for qualified one-off patches

• Rolling release upgrades of Oracle Clusterware

• Load balancing advisory

• Runtime connection load balancing

• Flexibility to scale up processing capacity using commodity hardware without downtime or changes to the application

• Comprehensive manageability integrating database and cluster features

  See Also:

  Oracle Database Oracle Clusterware and Oracle Real Application Clusters Administration and Deployment Guide

2.1.2 Oracle Data Guard

Oracle Data Guard provides a comprehensive set of services that create, maintain, manage, and monitor one or more standby databases to enable production Oracle databases to survive failures, disasters, errors, and data corruption. Data Guard maintains these standby databases as transactionally consistent copies of the production database. Then, if the production database becomes unavailable due to a planned or an unplanned outage, Data Guard can switch any standby database to the production role, thus greatly reducing the downtime caused by the outage. The failover of data processing from the production to the standby database can be completely automatic and done without any human intervention, thereby reducing the management costs associated with the Data Guard configuration. Data Guard can be used with traditional backup, restore, and clustering solutions to provide a high level of data protection and data availability.

A Data Guard configuration consists of one production database and one or more physical or logical standby databases. The databases in a Data Guard configuration are connected by Oracle Net and may be dispersed geographically. There are no restrictions on where the databases are located if they can communicate with each other. For example, you can have a standby database in the same building as your primary database to help manage planned downtime and two or more standby databases in other locations for use in disaster recovery.

Oracle Data Guard provides the following benefits:

• Maintains real-time, transactionally consistent database copies to provide protection against unplanned downtime and disaster

• Complete data protection against computer failures, human errors, data corruption, and site failures

• Reduces planned downtime for hardware and system upgrades, and Oracle patch set and database upgrades

• Automatic detection and resolution of missing data following temporary loss of connectivity between the primary and standby database

• Multiple levels of data protection and performance to balance data availability against system performance requirements

• Allows efficient use of system resources by diverting reporting and backup operations from the production database to standby databases

• Ability to diverge a Redo Apply standby database for reporting or testing purposes and resynchronize it with primary database once complete

• Managed and automatic role transition and application notification to minimize planned and unplanned downtime

• Automatic resynchronization of a failed primary database following a failover

• All systems managed as a single configuration for simplified administration

  See Also:

  Oracle Data Guard Concepts and Administration

2.1.3 Oracle Streams

Oracle Streams enables the propagation and management of data, transactions, and events in a data stream, either within a database or from one database to another. Streams provides a set of elements that allow users to control what information is put into a data stream, how the stream is routed from node to node, what happens to events in the stream as they flow into each node, and how the stream terminates.

Streams can be used to replicate a database or a subset of a database. This enables users and applications to simultaneously update data at multiple locations. If a failure occurs at one of the locations, then users and applications at the surviving sites can continue to access and update data.

Streams can be used to build distributed applications that replicate changes at the application level using message queuing. If an application fails, then the surviving applications can continue to operate and provide access to data through locally maintained copies.

Streams provides granularity and control over what is replicated and how it is replicated. It supports bidirectional replication, data transformations, subsetting, custom apply functions, and heterogeneous platforms. It also gives users complete control over the routing of change records from the primary database to a replica database.

As with Streams, Oracle Data Guard in SQL Apply mode can capture database changes, propagate them to destinations, and apply the changes at these destinations. Although Streams and Data Guard in SQL Apply mode share much of the same underlying technologies for high availability, Data Guard in SQL Apply mode is easier to implement and manage than a Streams-based high availability solution.

Oracle Streams provides the following benefits:

• Data protection by maintaining a full or partial remote copy of the database

• Achieves little or no downtime during database upgrade or maintenance operations such as migrating a database to a different platform or character set, modifying database objects to support upgrades to user-created applications, and applying an Oracle software patch

• Data replication by capturing DML and DDL changes made to database objects and replicating these changes to one or more other databases

• Event management and notification by enqueuing messages or capturing events, propagating the messages and events through queues, and dequeuing and applying or acting upon the message or event

• Supports heterogeneous platforms across databases within the configuration

• Allows character sets to differ between replicas

• Permits fine-grained control of data sharing

2.1.4 Oracle Flashback Technology

Flashback technology provides a set of features to view and rewind data back and forth in time. The flashback features offer the capability to query past versions of schema objects, query historical data, perform change analysis, and perform self-service repair to recover from logical corruption while the database is online.

Flashback technology provides a SQL interface to quickly analyze and repair human errors. Flashback provides fine-grained analysis and repair for localized damage such as deleting the wrong customer order. Flashback technology also enables correction of more widespread damage, yet does it quickly to avoid long downtime. Flashback technology is unique to the Oracle Database and supports recovery at all levels including row, transaction, table, tablespace, and database.

Flashback technology includes the following features:

• Oracle Flashback Query

• Oracle Flashback Versions Query

• Oracle Flashback Transaction Query

• Oracle Flashback Table

• Oracle Flashback Drop

• Oracle Flashback Database

• Oracle Flashback Restore Points

2.1.5 Automatic Storage Management

Automatic Storage Management (ASM) provides a vertically integrated file system and volume manager directly in the Oracle kernel, resulting in:

• Significantly less work to provision database storage

• Higher level of availability

• Elimination of the expense, installation, and maintenance of specialized storage products

• Unique capabilities for database applications

For optimal performance, ASM spreads files across all available storage. To protect against data loss, ASM extends the concept of SAME (stripe and mirror everything) and adds more flexibility in that it can mirror at the database file level rather than the entire disk level.

More importantly, ASM eliminates complexities associated with managing data and disks; it simplifies the processes of setting up mirroring, adding disks, and removing disks. Instead of managing hundreds and possibly thousands of files (as in a large data warehouse), DBAs using ASM create and administer a larger-grained object—the disk group—which identifies the set of disks that will be managed as a logical unit. Automation of file naming and placement of the underlying database files save DBAs time and ensures adherence to standard best practices.

The ASM native mirroring mechanism (2-way or 3-way) is an option that is used to protect against storage failures. With ASM mirroring, an additional level of data protection can be provided with the use of failures groups. A failure group is a set of disks sharing a common resource (disk controller or an entire disk array) whose failure can be tolerated. Once defined, an ASM failure group will intelligently place redundant copies of the data in separate failure groups to ensure that the data will be available and transparently protected against the failure of any component in the storage subsystem.

ASM provides the following benefits:

• Provides the ability to mirror across drives and storage arrays

• Automatically re-mirrors from a failed drive to remaining drives

• Automatically rebalances stored data when disks are added or removed while the database remains online

• Allows for operational simplicity in managing a database storage grid

2.1.6 Recovery Manager

Recovery Manager (RMAN) is an Oracle utility to manage the backup and, more importantly, the recovery of the database. It eliminates operational complexity while providing superior performance and availability of the database.

Recovery Manager determines the most efficient method of executing the requested backup, restoration, or recovery operation and then submits these operations to the Oracle database server for processing. Recovery Manager and the server automatically identify modifications to the structure of the database and dynamically adjust the required operation to adapt to the changes.

RMAN provides the following benefits:

• Automated channel failover on backup and restore operations

• Automatic failover to a previous backup when the restore operation discovers a missing or corrupt backup

• Automated creation of new database and temporary files during recovery

• Automated recovery through a previous point-in-time recovery—recovery through resetlogs

• Block media recovery enables the datafile to remain online while fixing the block corruption

• Fast incremental backups using block change tracking

• Merge incremental backups into image copies in the background providing up-to-date recoverability

• Optimized backup and restore of required files only

• Retention policy ensures that relevant backups are retained

• Resumable backup and restore of previously failed operations

• Automatic backup of the control file and the server parameter file ensuring that backup metadata is available in times of database structural changes as well as media failure and disasters

• Online backup does not require the database to be placed into hot backup mode

  See Also:

  Oracle Database Backup and Recovery Advanced User's Guide

2.1.7 Flash Recovery Area

The flash recovery area is a unified storage location for all recovery-related files and activities in an Oracle database. After this feature is enabled, all RMAN backups, archive logs, control file autobackups, and datafile copies are automatically written to a specified file system or automatic storage management disk group, and the management of this disk space is handled by RMAN and the database server.

Making a backup to disk is faster because using the flash recovery area eliminates the bottleneck of writing to tape. More importantly, if database media recovery is required, then datafile backups are readily available. Restoration and recovery time is reduced because you do not need to find a tape and a free tape device to restore the needed datafiles and archive logs.

The flash recovery area provides:

• Unified storage location of related recovery files

• Management of the disk space allocated for recovery files to simplify database administration tasks

• Fast, reliable disk-based backup and restoration

2.1.8 Oracle Security Features

The best protection against human errors is to prevent their occurrence. The best way to prevent human errors is to restrict user access to data and services to only those they truly need to perform their business functions. Oracle provides a wide range of security tools to control user access to application data by authenticating users and then enabling administrators to grant users only those privileges required to perform their duties.

In addition, the security model of the Oracle database provides the ability to restrict data access at a row level using the Virtual Private Database feature, thereby further isolating users from data that they do not need to access.

Oracle security features include:

• Authentication control to validate the identities of entities using the networks, databases, and applications

• Authorization control to provide limits to user access and actions linked by user identities and roles

• Access control to objects, providing protection regardless of the entity seeking to access or alter them

• Auditing control to monitor and gather data about specific database activities, investigate suspicious activity, deter users (or others) from inappropriate activities, and detect problems with authorization or access control implementation

• Security policy management using profiles

• Encryption of data residing within the database and backups, or transferred to and from databases

2.1.9 Fast-Start Fault Recovery

Oracle provides fast and predictable recovery from system faults and database failures. The Fast-Start Fault Recovery technology included in the Oracle database automatically bounds database recovery time upon startup by using its self-tuned checkpoint processing. This makes recovery time fast and predictable, and improves the ability to meet service level objectives. Oracle's Fast-Start Fault Recovery can reduce recovery time on a heavily-laden database from tens of minutes to a few seconds.

Fast-Start Fault Recovery features include:

• Predictable, bounded recovery from computer failures

• Database checkpointing is self-tuning to maintain desired recovery time objective

2.1.10 LogMiner

Oracle log files contain useful information about the activities and history of the Oracle database. Log files contain all data necessary to perform a database recovery, and also record all changes made to the data and metadata within the database.

LogMiner is a fully relational tool that allows redo log files to be read, analyzed, and interpreted using SQL. Analysis of the log files with LogMiner can be used to:

• Track or audit changes to data

• Provide supplemental information for tuning and capacity planning

• Retrieve critical information for debugging complex applications

• Recover deleted data

LogMiner features include:

• Pinpoint when a logical corruption to the database—such as errors made at the application level—may have occurred

• Determine the necessary actions to perform fine-grained recovery at the transaction level

• Performance tuning and capacity planning through trend analysis

• Perform post-auditing

2.1.11 Hardware Assisted Resilient Data (HARD) Initiative

Oracle9i introduced the Hardware Assisted Resilient Data (HARD) Initiative, a program designed to prevent data corruption before it happens. Data corruption is very rare, but when it happens, it can have a catastrophic effect on a database, and therefore a business.

Under the HARD Initiative, Oracle works with selected system and storage vendors to build operating system and storage components that can detect corruption early and prevent corrupted data from being written to disk. The key approach is block checking where the storage subsystem validates the Oracle block contents.

To use HARD validation, all datafiles and log files are placed on HARD-compliant storage. The user must also enable the HARD validation feature on the storage, using the vendor-provided interface. When Oracle writes data to the storage, the storage system validates the data. If the data appears to be corrupted, then the write is either rejected with an error, or it is accepted with an error logged by the storage in the internal logs.

Figure 2-1 Oracle Data Validation

Description of "Figure 2-1 Oracle Data Validation"

Storage vendors may choose to implement some or all of the checks in their implementation. Also, each vendor's implementation is unique and their control interfaces may have different features. Please check with the HARD initiative page for the latest vendor and implementation information.

http://www.oracle.com/technology/deploy/availability/htdocs/HARD.html

2.2.1 Computer Failures

A computer failure outage occurs when the system running the database becomes unavailable because it has shut down or is no longer accessible. Downtime caused by computer failures can be reduced by employing rapid database recovery upon startup, or avoided by using cluster technology or data mirroring techniques.

Oracle offers the following high availability solutions to address computer failures:

• Fast-Start Fault Recovery

• Oracle Real Application Clusters

• Oracle Data Guard

• Oracle Streams

For information on the benefits and attainable recovery time for each solution, see Table 2-1.

2.2.2 Storage Failures

A storage failure outage occurs when the storage holding some or all of the database contents becomes unavailable because it has shut down or is no longer accessible. Downtime caused by storage failures can be reduced by keeping disk-based backups, copies, or replicas of the database, or avoided by using storage mirroring.

Oracle offers the following high availability solutions to address storage failures:

• Automatic Storage Management

• Recovery Manager

• Flash Recovery Area

• Oracle Data Guard

• Oracle Streams

For information on the benefits and attainable recovery time for each solution, see Table 2-1.

2.2.3 Human Errors

A human error outage occurs when unintentional or malicious actions are committed that cause data within the database to become logically corrupt or unusable. The service level impact of a human error outage can vary significantly depending on the amount and critical nature of the affected data. The best protection against a human error outage is to prevent humans errors from occurring where possible, and when prevention is not possible, to detect and undo the errors quickly.

Oracle offers the following high availability solutions to address human errors:

• Oracle Security Features

• Oracle Flashback Technology

• LogMiner

For information on the benefits and attainable recovery time for each solution, see Table 2-1.

2.2.4 Data Corruption

A data corruption outage occurs when a hardware or software component causes corrupt data to be read or written to the database. The service level impact of a data corruption outage may vary, from a small portion of the database (down to a single database block) to a large portion of the database (making it essentially unusable). If not prevented—or quickly detected and repaired—data corruption can disrupt the entire database or cause key business data to be lost.

Oracle offers the following high availability solutions to prevent—or detect and repair—data corruption:

• Hardware Assisted Resilient Data (HARD) Initiative

• Recovery Manager

• Flash Recovery Area

• Oracle Data Guard

• Oracle Streams

• Block checking db_block_checking and block checksumming with db_block_checksum

• DBVERIFY, ANALYZE, and DBMS_REPAIR detection tools

For information on the benefits and attainable recovery time for each solution, see Table 2-1.

2.2.5 Site Failures

A site failure outage occurs when an event causes all or a significant portion of an application to stop processing or slow to an unusable level. A site failure may affect all processing at a data center, or a subset of applications supported by a data center. Downtime caused by a site failure can be minimized by keeping copies or replicas of the database updated in real time.

Oracle offers the following high availability solutions to address site failures:

• Recovery Manager

• Oracle Data Guard

• Oracle Streams

For information on the benefits and attainable recovery time for each solution, see Table 2-1.

2.3.1 Dynamic Resource Provisioning

Oracle continues to broaden support for dynamic reconfiguration of the database, enabling it to adapt to changes in hardware demands without any service interruptions. The Oracle database dynamically accommodates various changes to hardware and database configurations:

• Add and remove processors from an SMP server

• Add and remove nodes and instances in an Oracle Real Application Cluster (RAC) environment

• Dynamically grow and shrink its shared memory allocation and automatically tune memory online using Automatic Shared Memory Management

• Add and remove database disks online without disturbing database activities using Automatic Storage Management (ASM)

• Add and remove storage arrays online without disturbing database activities using ASM

• Automatically rebalance I/O load across the database storage using ASM

• Move datafiles online when adding or dropping disks using ASM, which automatically rebalances database storage whenever the storage configuration is changed

These capabilities provide no-cost system changes and capacity on-demand provisioning, both of which are fundamental requirements of enterprise Grid computing.

Memory and storage management have improved significantly with the advent of Automatic Shared Memory Management and Automatic Storage Management (ASM). By setting the SGA_TARGET parameter to a nonzero value, the shared pool, large pool, Java pool, Streams pool, and buffer cache can automatically and dynamically resize as needed. ASM automates and simplifies the layout of datafiles, control files, and log files. Database files are automatically distributed across all available disks, and database storage is rebalanced whenever the storage configuration changes, including adding and removing disks or storage arrays. ASM provides redundancy through the mirroring of database files, and provides optimal performance by automatically distributing database files across all available disks. Rebalancing of the database storage automatically occurs whenever the storage configuration changes.

Another type of dynamic reconfiguration occurs when Oracle polls the operating system to detect changes in the number of available CPUs and reallocates internal resources. In addition, almost all initialization parameters can be changed without shutting down the instance. Simply use the ALTER SESSION statement to change the value of a parameter during a session, or the ALTER SYSTEM statement to change the value of a parameter in all sessions of an instance for the duration of the instance.

2.3.2 Rolling Upgrades

The Oracle database continues to reduce downtime required for system, software, and application upgrades. Oracle provides the following benefits:

• Zero downtime for system and hardware upgrades with RAC

• Zero downtime for operating system upgrades with RAC

• Zero downtime for qualified one-off database patches with RAC

• Zero downtime for storage migration with ASM

• Minimum downtime for system or cluster upgrades with Data Guard

• Minimum downtime for patchset or database upgrades with Data Guard

• Minimum downtime for database upgrade with Transportable Tablespace

• Minimum downtime for platform migration using Transportable Tablespace and potentially Data Guard

• Minimum downtime for database upgrade with Oracle Streams

• Minimum downtime for platform migration with Oracle Streams

Table 2-2 describes the various Oracle high availability solutions for planned downtime along with the recovery time that can be attained with each solution and their known considerations. For all cases, extensive testing is highly recommended prior to performing any rolling upgrade.

^Footnote 1 An example is migration from traditional storage to low cost storage

^Footnote 2 For more information, refer to the best practices white papers available at http://www.oracle.com/technology/deploy/availability/htdocs/maa.htm.

2.3.3 Online Reorganization and Redefinition

One way to enhance availability and manageability is to allow users full access to the database during a data reorganization operation. The Online Reorganization and Redefinition feature in Oracle Database 10g offers administrators significant flexibility to modify the physical attributes of a table and transform both data and table structure while allowing users full access to the database. This capability improves data availability, query performance, response time, and disk space utilization—all of which are important in a mission-critical environment and make the application upgrade process easier, safer, and faster.

This online architecture provides the following benefits:

• Tables can be reorganized and redefined online

  • Any physical attribute of the table can be changed online. The table can be moved to a new location, partitioned, and converted from one organization (such as heap-organized) to another (such as index-organized).

  • Many logical attributes can also be changed. Column names, types, and sizes can be changed. Columns can be added, deleted, or merged. One restriction is that the primary key of the table cannot be modified.

• All index operation can be performed online

  • Indexes can be created online and analyzed simultaneously. Online repair of the physical guess component of logical rowids (used in secondary indexes and in the mapping table for index-organized tables) can also be used.

  • An index-organized table and secondary indexes can be reorganized online to eliminate the reorganization maintenance window. Secondary indexes support efficient use of block hints (physical guesses). Invalid physical guesses of logical rowids stored in secondary indexes on index-organized table can also be repaired online.

  • An index-organized table or table partition can be reorganized without rebuilding its secondary indexes, resulting in a short reorganization maintenance window.

• Online move of a partitioned table

• Online reorganization support for advanced queues, clustered tables, materialized views, and abstract data types (objects)

Depending on the type of online reorganization that is required, the following types of data reorganization can be performed using the DBMS_REDEFINITION package or the SQL CREATE/ALTER TABLE and INDEX commands:

• Modify table storage parameters

• Move table to a different tablespace

• Add support for parallel queries

• Add or drop partitioning support

• Recreate table to avoid fragmentation

• Change from table to IOT or from IOT to table

• Add or drop a column

• Transform a column using a function

• Create indexes online

• Rebuild indexes online

• Coalesce indexes online

• Move index-organized tables online

• Copy dependent objects (such as triggers, constraints, and indexes)

• Convert LONG and LONG RAW columns to a LOB

• Use a unique key as an alternative to a primary key or rowid

• Specify columns to order data by

• Change a table without recompiling stored procedures

• Online segment shrink

• Reorganize a single partition

• Reorganize advanced queue and clustered tables

• Reorganize a table containing an ADT

• Retain and clone statistics

• Copies check and not null constraints

• Copies dependent objects for nested tables

2.4.1 Database Server Grid

The availability of low-cost and reliable blade servers, small multi-processor servers, and inexpensive open-source operating systems such as Linux, has made it possible to build a Database Server Grid that is highly available, scalable, flexible, and manageable.

Oracle Real Application Clusters (RAC) is the technology that enables a Database Server Grid by providing a single database that spans multiple low-cost servers yet appears to the application as a single, unified database system. RAC provides flexibility to dynamically provision resources and services within the Grid as computing needs change, and to add systems to the Grid as capacity demands increase. In addition, RAC provides protection from system failures by automatically recovering the processing of a failed node by any of the surviving systems running the database, and facilitating the reconnection of clients and redistribution of load affected by the failed system.

2.4.2 Database Storage Grid

The availability of low-cost ATA disk-based storage arrays and low-cost storage networks has made it possible to use a Database Storage Grid with the Oracle database at very low cost. A DBA can use the Automatic Storage Management (ASM) interface to specify the disks within the Database Storage Grid that ASM should manage across all server and storage platforms. ASM partitions the disk space and evenly distributes the data storage throughout the entire storage array. Additionally, ASM automatically redistributes the data storage as storage arrays are added or removed from the Database Storage Grid.

When used with Oracle Data Guard, Oracle Flashback, and Recovery Manager (RMAN), the Database Storage Grid can achieve high availability and performance for the Oracle database. Oracle Data Guard protects against site disasters and data corruption by automatically maintaining a standby database. Oracle Flashback protects against human errors using snapshot capabilities, allowing a DBA to rewind a table or the entire database to a specific moment in time before the error occurred. RMAN provides full and incremental backup to disk or tape for archival and protection against data corruption, while the Flash Recovery Area fully manages all disk-based recovery related files and database activities.

2.4.3 Resilient Low-Cost Storage Initiative

To help customers successfully deploy a Database Storage Grid, Oracle has launched the Resilient Low-Cost Storage Initiative. This initiative is analogous to the Oracle Storage Compatibility Program (OSCP) but with a focus on low-cost storage. Its goal is to ensure that low-cost storage performance can be manageable, reliable, and optimal.

This initiative partners Oracle with storage vendors to validate that their low-cost storage arrays meet the minimum price, functionality, and performance requirements for use in a Database Storage Grid. Vendors use Oracle-supplied tools to characterize their storage array performance. After verifying that all requirements can be satisfied, the storage array is certified for this initiative and the vendor—together with Oracle—produces a Best Practice white paper to instruct Oracle customers on how to optimally configure the storage array for a Database Storage Grid environment.