1 Oracle Database 10g New Features

Oracle Open Database Connectivity (ODBC) Driver Support for Linux

Support has been added for an Oracle ODBC driver for the Linux platform. Now, ODBC applications can be developed and used on Linux with an Oracle supported ODBC driver.

This support enables ODBC applications to run on the Linux platform, encouraging wider Oracle Database adoption by allowing Linux clients to access Oracle databases.

Oracle Call Interface (OCI) Client Version API

This feature adds client versioning macros to support version sensing.

This allows applications to be compiled for multiple versions of the Oracle client from a single source code environment. This leads to easier development of products with customizations based on versions.

Start Up and Shut Down of Database OCI API

This feature adds API calls for starting and stopping a database from inside an application.

This enables ISVs and testers to start up and shut down databases without using SQL*Plus or other tools. This streamlines the testing process for application developers.

Oracle C++ Call Interface (OCCI) Transparent Application Failover Support

This feature adds extensions to OCCI API to support Transparent Application Failover (TAF).

This enables more robustness with TAF. If a server node goes down, applications automatically reconnect to another surviving node, replaying transactions if possible.

Instant Client Light

Instant Client Light reduces the size of the Instant Client package by removing most of the character sets and leaving only English error messages.

This feature allows Instant Client to be used in a wider variety of environments, especially where storage space constraints are prominent.

Instant Client Software Development Kit (SDK)

The Instant Client SDK packages the OCI/OCCI header files necessary for developing Oracle client applications outside of an Oracle installation. Now, application developers can develop, test, and deploy Oracle client applications from an Instant Client deployment without the need for an ORACLE_HOME or full client installation.

This feature enables a larger development community to develop client applications for Oracle, encouraging more Oracle Database adoption.

Unicode 4.0

Unicode 4.0 is the latest release of the Unicode standard. Oracle's Unicode character sets, AL32UTF8 and AL16UTF16, have been updated to support Unicode 4.0 in Oracle Database 10g Release 2. Unicode 4.0 has 1,226 new characters.

This feature provides greater support of the Unicode character set for Oracle Database users.

Unicode Data Type Literal Enhancement

This feature supports SQL string literal INSERT and UPDATE operations on NCHAR without data loss regardless of what the database character set happens to be. Multilingual Unicode data can now be added without restrictions such as having to provide hex Unicode values. The support for this feature is available in SQL, PL/SQL, OCI, and JDBC.

This feature prevents data loss when performing INSERT and UPDATE operations on NCHAR.

Globalization Development Kit (GDK) 2.0

Globalization Development Kit (GDK) has added new locales and common locale mapping information into the GDK for the PL/SQL package.

Among other things, the mappings allow for picking the most appropriate Oracle Linguistic sort when ordering different users' language data.

Additional mapping functions have been added to allow developers to build global PL/SQL applications including Common Sort, Common Time Zone, Common Territory, and Translated Locale Names.

Character Set Scanner (CSSCAN) Enhancements

Two new parameters, QUERY and COLUMN, have been added in CSSCAN in Oracle Database 10g Release 2. These parameters offer finer control in performing selective analysis on the user data. Support for some remaining data types that CSSCAN does not handle has been added. They are multilevel VARRAY and multilevel nested tables.

The enhancements in CSSCAN provide more flexibility in defining the criteria for data scanning, which in turn helps to more accurately detect the effect of migrating the database to a new character set.

Language and Character Set File Scanner (LCSSCAN) Enhancements

In this release, enhancements include:

• Improved guessing for shorter length strings

• Support for HTML files

Oracle's Language and Character Set Detection (LCSD) is a high-performance, statistically-based technology for determining the language and the character set for unknown text in a file. It can automatically identify a wide variety of language and character set pairs. Introduced in Oracle Database 10g Release 1, LCSD is offered as part of the Migration Utilities and called the Language and Character Set File Scanner (LCSSCAN). LCSD is also offered as part of the Globalization Development Kit through Java API's.

Consistent Linguistic Searching and Sorting

A new NLS_COMP parameter, LINGUISTIC, has been added to support the existing and remaining SQL & PL/SQL operators and functions.

This feature ensures all SQL string comparisons are consistent, and that they follow the linguistic convention as specified in the NLS_SORT parameter.

Transliteration for Japanese Kana Conversion

This feature provides a transliteration mechanism to support various Japanese Kana conversions in different forms. The TRANSLITERATE function for the UTL_I18N PL/SQL package is provided for transliteration in this release.

This feature performs Japanese Kana conversions.

JDBC XA Connections Caching

This feature improves JDBC support for caching XAConnections on OracleXADataSources. Applications can now enable Implicit Connection Caching and Fast Connection Caching (in Real Application Clusters environments) on OracleXADataSources. When XAConnection caching is not enabled, applications obtain XAConnections directly from OracleXADataSource, and then obtain the associated XAResource objects.

This feature improves the performance of distributed transactions through the caching of XA connections in a single instance or multiple instance database environments.

JDBC Runtime Connection Load Balancing

This feature provides dynamic routing of connection requests to the least loaded instance. The JDBC connection cache manager uses Real Application Clusters (RAC) workload metrics and the specified load balancing policy to choose the best instance to process the connection request.

The benefit is efficient RAC and Grid database resource utilization by even distribution of the workload across RAC instances based on workload metrics and distribution policies.

JDBC Support for Data Manipulation Language (DML) in the RETURNING Clause

This feature enables you to issue a DML command (for example, INSERT, UPDATE, or DELETE) specified in the RETURNING clause upon a SQL query execution.

This feature increases performance by saving a client-to-database round-trip.

Support for JDBC 3.0 ResultSet Holdability

By default, the database automatically closes cursors and result sets upon transaction commit. A holdable cursor, or holdable resultset, is not closed when the transaction that contains the cursor is committed thereby reducing hard parses and library latches. JDBC 3.0 ResultSet holdability is specified using the createStatement(), prepareStatement(), or prepareCall() methods and one of the following two constants: HOLD_CURSORS_OVER_COMMIT or CLOSE_CURSORS_AT_COMMIT. Oracle Database supports HOLD_CURSORS_OVER_COMMIT. Any other value passed as an argument is ignored.

Keeping cursors open when a transaction is committed increases the performance of JDBC applications or J2EE components.

Support for JDBC 3.0 Retrieval of Auto-Generated Key

During a SQL INSERT command, this feature retrieves the value of the pseudo-column, ROWID, of the newly inserted row.

This feature allows JDBC applications and J2EE components to retrieve the key (ROWID) of the newly inserted row during the same database operation (INSERT). As a result, a network round-trip to the database is saved and the performance is increased.

Cluster Bar Charts

SVG charts now include Cluster Bar charts where bars are grouped into columns of data allowing for easy comparison.

The new Cluster Bar charts provide a method of visualizing data in a side-by-side comparison series.

Dial Charts

SVG charts now include a dial or gauge chart, allowing you to plot a single value relative to a range defined by a minimum and maximum value represented by a needle on a semicircular surface.

The new Dial chart provides a method of visualizing data commonly used in executive dashboards.

Percentage Bar Charts

SVG charts now include a Percent Bar chart used to show cumulative percentages.

Percent Bar charts provide a method of visualizing data that was previously unavailable in Oracle HTML DB.

Stacked Bar Charts

SVG charts now include Stacked Bar charts for showing multiple series of absolute values combined into single bars. They are useful when visualizing accumulations or cumulative data.

The addition of Stacked Bar charts allow visualizing of cumulative data in bar charts that was previously not possible.

Asynchronous Refresh for All Charts Based on Scalable Vector Graphics (SVG)

SVG charts can now be configured to refresh at an interval defined by the developer. The refresh causes the chart to rerun the query upon which its data set is based.

Every SVG chart used in an HTML DB application can now automatically refresh itself at regular intervals based on the most up-to-date data.

Create Master Detail With Choice of Layout

The Master Detail wizard creates a master detail mini-application divided into either 2 or 3 pages with the following features:

• Lost update detection

• Developer-defined navigation through master rows with Next and Previous buttons

• Editing of detail rows on a separate page or on a combined master detail form page

• Single click adding of detail rows in tabular form

The Master Detail wizard automates the manual steps previously required to build a master detail form.

Application From a Spreadsheet Wizard

The Create Application wizard has been enhanced in two ways. First, it now enables you to create an application on spreadsheet data. Previously, importing of spreadsheet data had to be done in a separate step using the Data Workshop. This task has now been integrated into the wizard. Secondly, the generated application both on imported spreadsheet data and on an existing table has more features. Those features include:

• Optional read-only version of the generated application

• Record-by-record navigation

• More analysis screens with summary reports and charts

The enhanced Create Application wizard simplifies the process of creating an application based on spreadsheet data and generates a more feature rich application.

Create Forms and Reports Based on a Web Service Reference

With the enhanced Web Services wizard, you can create forms that invoke web services and reports based on the results returned by the web service.

The enhancements to the Web Services wizard allow you to incorporate web services in an application without any programming.

Discover Web Services Through UDDI Repository

A Universal Description, Discovery, and Integration (UDDI) registry is a directory where businesses register their web services. Through UDDI, the new Web Services wizard can search for registered web services by name.

The UDDI feature of the new Web Services wizard simplifies the process of registering a web service for consumption by an Oracle HTML DB application.

Simplified and More Consistent User Interface

The user interface for Oracle HTML DB has been streamlined to improve usability. Specifically, a large number of options have been removed or reorganized to eliminate confusion. Additionally, commonly used tasks such as Run Application or Edit Application have moved to improve consistency.

Oracle HTML DB's simplified and more consistent user interface makes HTML DB easier to navigate and reduces the number of clicks to create and edit applications and components.

PL/SQL Conditional Compilation

Conditional compilation enables PL/SQL programmers to use new language features in a PL/SQL compilation unit without sacrificing the ability to compile the unit in an older database version that does not support these features. Other typical uses include embedding conditionally compiled debugging and tracing support code in PL/SQL programs.

The benefit is that one body of source code can be deployed in a range of Oracle Database versions to take advantage of new features where they are available and to use fallback approaches where they are not. Debugging and tracing code can be left in place without penalizing the size of the executable or runtime efficiency. Such code can be easily activated by a simple recompilation. This provides a feature in PL/SQL that programmers who use other languages have come to expect.

Obfuscation of Dynamically Generated PL/SQL Source Code

The existing DBMS_DDL supplied package is enhanced with two new subprograms. The command-line WRAP utility takes a plain text CREATE or REPLACE command that specifies creation of a PL/SQL unit and returns a CREATE or REPLACE command where the text of the PL/SQL unit has been obfuscated. The obfuscated representation is identical to what is produced if the same CREATE or REPLACE command is processed with the command-line WRAP utility. The CREATE_WRAPPED command is a shortcut for and has the same semantics as the EXECUTE IMMEDIATE DBMS_DDL.WRAP (input) command.

This feature allows the text of a PL/SQL unit that is created dynamically to be obfuscated. This is critical when an installed product generates new PL/SQL units, for example as part of a customization process, and when the generated code embodies methods that the product vendor considers to be valuable intellectual property. The command-line WRAP utility, available in Oracle Database 10g Release 1 and earlier, allowed the code that a vendor shipped to be obfuscated. This new feature allows custom code generated by shipped vendor code to be obfuscated.

Relaxation of Line Length and Overall Limits for the DBMS_OUTPUT PL/SQL Package

The declaration of the package type DBMS_OUTPUT.CHARARR is enhanced to support the increased maximum of 32,767 bytes. Previously, the VARCHAR2 constraint was 255 bytes. This determines the maximum line size that can be written by the DBMS_OUTPUT.PUT and DBMS_OUTPUT.PUT_LINE procedures. The procedure DBMS_OUTPUT.ENABLE (buffer_size integer defaults to 20000) has a new allowed value of NULL for the buffer_size. In this case, NULL means "unlimited." SQL*Plus, through the SET SERVEROUTPUT ON command, enables you to take advantage of the new "unlimited" meaning.

PL/SQL programmers frequently use DBMS_OUTPUT and, in Oracle Database 10g Release 1 and earlier, were constrained by the 255 byte limit. When using SQL*Plus, most programmers are regularly caught by the small default overall limit and sometimes by the current 1,000,000 maximum overall limit. In Release 2, the line length limit is increased to 32,767 bytes and the overall limit is removed altogether.

UTL_NLA PL/SQL Package for Matrix Math

C and Fortran APIs for matrix math are freely available. Moreover, efficient platform-specific implementations have been written for many of the most popular platforms. The new UTL_NLA package exposes a matrix PL/SQL data type and wrapper PL/SQL subprograms for two of the most popular of these C APIs; namely BLAS and LAPACK.

To do matrix manipulation in PL/SQL in Oracle Database 10g Release 1 and earlier, it was necessary to invent a matrix representation based on PL/SQL native data types and then write matrix manipulation routines from scratch. This required substantial programming effort and the performance of the resulting implementation was limited. Oracle Database 10g Release 2, through the new UTL_NLA package, removes the programming effort and delivers a fast implementation.

Rules Manager

Rules Manager is a new feature of Oracle Database 10g Release 2. It enables developers to create applications that process and respond to events of any complexity using rules and policies defined in the database. It can evaluate events using data from the application and from database tables. It stores intermediate results to quickly evaluate the next event in a long running composite event (an event made up of two or more simple events). Rules are defined using XML and SQL and can have complex conditions using conjunctions and disjunctions, and specify a set of events, time, and non-occurrence of events with or without a deadline. Event policies control how rules are processed and the duration of an event. Rules can trigger actions that are user-defined procedures running inside an Oracle Database or actions in another application.

The benefit of this feature is that rules that are managed in Oracle Database keep pace with changing business conditions and are always up-to-date; rules are easily changed with SQL and are not included in your application or loaded into a memory-based rules repository. Rules can be evaluated efficiently with the complete business context stored in your Oracle Database and data provided by your application. Event response is flexible; rules can trigger actions in Oracle Database or your application.

Additional advantages of Oracle Database over other approaches include:

• Manageability, by storing rules and event policies with your application data.

• Performance, by evaluating rules and coordinating multiple events and application threads with the full capabilities of Oracle Database.

• Scalability, by evaluating sets of rules of any size.

Regular Expression Enhancements

In Oracle Database 10g Release 2, Regular Expression is enhanced for Perl compatibility. Support for the common Perl REGEXP extensions that are not included and do not conflict with the POSIX standard are added.

Oracle's SQL Regular Expression provides a simple yet powerful mechanism for rapidly describing patterns and greatly simplifies the way in which you search, extract, format, and otherwise manipulate text in the database.

Collection Operator Performance Improvements

Collection operators operate on collections of data objects. For Oracle Database 10g Release 2, the EQUALITY and MEMBER operators for collections have been improved.

These enhancements have improved the performance for collection operators.

Distributed Large Objects (LOB) Support

Distributed LOBs support provides easy-to-use and efficient support for accessing unstructured data in a distributed environment. The data interface for LOBs can now INSERT, UPDATE, and SELECT LOBs across dblinks.

The benefit of this new support is the ability to access remote LOBs.

Expression Filter Enhancements

Expression Filter usability and performance enhancements include:

• Providing extensible optimizer support for the EVALUATE operator.

• Allowing users to specify default values for a subset of attributes in an attribute set.

• Allowing users to specify name spaces in stored XPath expressions.

• Indexing XPath range predicates.

• Allowing users to collect statistics on a set of XPath expressions stored in a table to create an optimal Expression Filter index structure.

Expression Filter features for Oracle Database 10g Release 2 include better usability and improved performance. You can specify default values for expression attributes and specify name spaces in XPath expressions.

XQuery

This feature provides a native implementation of XQuery, the World Wide Web Consortium's (W3C) emerging standard for querying XML, integrated into Oracle Database. It introduces two new functions, XMLQUERY and XMLTABLE.

You can use the XQuery language from an SQL client to query Oracle XML DB repository documents, XML views over relational data, and XMLType data, whether persistent or generated.

XMLTABLE maps the result of an XQuery to relational rows and columns so that results can be queried using SQL.

The benefit of this feature is that Oracle now supports querying of XML, relational, object-relational and repository data using XQuery, the W3C emerging standard for querying XML.

Advanced XPath Rewrite

Oracle Database 10g Release 2 extends the XPath Rewrite feature of XML DB to handle additional operations on more complex Schema Constructs. This feature reduces the number of conditions under which XPath Rewrite does not occur. The net effect is that applications using these constructs run faster without any recoding.

This feature allows faster querying and updating of schema-based XMLType data by extending the number of Schema Constructs and Xpath expressions that can be processed by XPath Rewrite.

New InsertXML(), AppendChildXML(), InsertXMLBefore(), and DeleteXML() Functions

The current UpdateXML() function can only manipulate the content of existing nodes within an XML document. It cannot be used to add or remove nodes from a document. The only way to add or remove nodes from a document is to use the methods provided by the DOM API.

This feature resolves this issue by introducing a family of SQL functions similar to UpdateXML() that make it possible to add and remove nodes from a document.

It also enables XPath Rewrite in the cases where the target of the function is a schema-based XMLType with structured storage. This means that operations can be executed in a more efficient manner than would be possible with the DOM API.

This feature improves programmer productivity by providing functions that make it possible to perform complex operations on the content of XML documents directly from SQL. Typically, these functions reduce the number of statements required to perform a particular operation when compared to typical DOM-based operations.

This new set of functions also improves application performance by allowing more complex manipulation of XML content to be performed by XPath Rewrite. XPath Rewrite makes it possible to perform operations on XML documents faster than can be achieved with conventional DOM or SAX-based programming techniques. It is now possible to do in one statement what previously took 10 to manipulate XML eliminating the need to use the DOM API to manipulate XML documents. These new functions improve application performance by allowing complex operations on XML documents to be performed using XPath Rewrite.

Simple Object Access Protocol (SOAP) 1.2 Implementation in C/C++

Implementation of the SOAP 1.2 W3C specification in C and C++.

Developers can build C or C++ SOAP Services to consume and generate SOAP messages using the Oracle XML stack.

SQL/XML Standard (SQL:2003 Standard Part 14) Compliance

This feature extends Oracle's support for the SQL:2003 Standard, adding support for comments, processing instructions, and CDATA and XML prologues. This helps ensure that the SQL/XML operators become the defacto approach to generated XML from a relational query, eventually replacing both XSU and the DBMS_XMLGEN package.

SQL/XML provides an intuitive, standards-based method for creating XML directly from a SQL query. The SQL/XML specification is Part 14 of the SQL:2003 Standard. This feature extends Oracle's support for the SQL/XML operators to bring it in compliance with the SQL:2003 Standard.

SQL/XML Standard Compliance

SQL/XML lets you query XML using SQL in a standard way and defines ways of importing and storing XML data in a SQL database, manipulating it within the database, and publishing both XML and conventional SQL data in XML form.

Oracle Database 10g Release 2 includes support for all the functions in SQL/XML 2003 (for example, XMLSERIALIZE and XMLPARSE), and some Oracle-proposed extension functions.

Easy XML Application Development Using JAXB Customization

The Java Architecture for XML Binding (JAXB) customization feature permits the XML-generated class names to be replaced with application or Java-specific names to ease manipulation of Java objects making development and maintenance more efficient and straightforward.

Application developers can more easily incorporate XML support into applications using JAXB. The customization feature enables modification of Java classes to speed development and reduce support costs.

Oracle XML Developer's Kit (XDK) PL/SQL Document Object Model (DOM) API Deprecation

This feature completes the work required to make it possible to deprecate the XDK PL/SQL packages and replace them with synonyms equivalent to Oracle XML DB PL/SQL packages without having a negative effect on applications that were originally developed using the XDK implementation.

This feature allows applications developed with the legacy XDK PL/SQL DOM API package (XMLDOM) to leverage the improved performance and functionality offered by the Oracle XML DB PL/SQL DOM API (DBMS_XMLDOM) without requiring significant application changes.

This release completes the deprecation of the XMLDOM, XMLPARSER and XSL_PROCESSOR packages in favor of DBMS_XMLDOM, DBMS_XMLPARSER and DBMS_XSLPROCESSOR.

XPath 2.0 and XQuery 1.0 Functions and Operators Support in Java

The World Wide Web Consortium (W3C) XPath 2.0 and XQuery 1.0 Functions and Operators adds support for the XML Schema data types along with a rich set of functions and operators to XPath that can be used in advance XSLT 2.0 style sheets and XQuerys.

This feature improves the performance and range of applications for style sheets and XQuerys. This reduces the need for compiled code, easing development and maintenance.

Custom Metadata for Oracle XML DB Repository Resources

This feature allows an application developer to define an XML Schema that defines the custom metadata and associate it with a particular XML DB resource. Once the XML Schema has been defined, application developers can create richer content management applications by adding customer metadata to the standard metadata recorded by the Oracle XML DB repository.

This feature makes better use of information contained in documents by making it possible to associate custom metadata with documents stored in the Oracle XML DB repository. This metadata improves your ability to classify, organize, manage, process, and search the vital corporate information stored in documents.

Secure Sockets Layer (SSL) Support

This feature enables the use of the secure HTTP protocol (HTTPS), in addition to the HTTP protocol, with Oracle XML DB.

SSL adds support for security standards to the Oracle XML protocols servers and allows the HTTPS protocol to be used to secure the transport of information between client and server.

Transportable Tablespace Support

This feature adds new support for XMLType to the Transportable Tablespace capability.

The benefit of this new feature is that it allows the Transportable Tablespace feature of Oracle Database to be used to move tablespaces containing schema-based XMLType tables and columns from one database to another. This method moving XML between different Oracle databases is faster than alternatives such as FTP and Import/Export.

Oracle XML DB Administration in the Enterprise Manager Web Console

Enterprise Manager can be used to manage the following Oracle XML DB features:

• XML DB parameters

• XML DB resources

• XML DB Access Control Lists (ACL)

• XML schemas

• XMLType tables and columns

It is now convenient to manage Oracle XML DB in the same place where Oracle Database is managed, which is in the Enterprise Manager Web console.

Fast-Start Failover

On a failure or outage of the primary server or site, or of the network, Data Guard can now rapidly and automatically fail over to a previously chosen standby database, without requiring manual intervention.

The benefit of fast-start failover is that it increases the degree of high availability, as well the robustness of disaster recovery, for the system infrastructure. It also reduces manual operations, thereby reducing management costs. This makes Data Guard particularly attractive as a high availability/disaster recovery solution for 24x7 applications where it is extremely critical to maintain uptime transparently, despite failures and outages.

Automatic Conversion of the Primary Database to a Standby Database Upon Failover

After a failover, this feature allows the old primary database to be automatically reinstated as a standby database, without requiring re-creation of the primary, or manually performing the Flashback Database operation. This feature is supported only for synchronous redo transports that enabled zero data loss (that is, Maximum Protection and Maximum Availability modes). It also requires Flashback Database to be enabled on both the primary and target standby databases. It is also possible to do this conversion using a single command or a simple push-button interface through Enterprise Manager.

The benefit of this feature is that it enables the Data Guard configuration to easily restore disaster protection in the configuration, without complex manual steps, improving the robustness of the disaster recovery features of Data Guard, as well as improving Data Guard manageability.

Optimized Asynchronous Redo Transmission

Asynchronous redo transmission using the log writer process is no longer limited by the size of the network buffer. The log writer process is no longer blocked on network time outs during asynchronous redo transmission. This allows a greater number of archiver processes to archive redo logs remotely.

The benefit of this feature is that it reduces any effect on the primary database during asynchronous redo transmission. It allows faster redo transmission to standby databases for bulk batch updates on the primary database.

Faster Redo Apply Failover

This feature allows the user to activate a physical standby database (that is, fail over to a primary database) that has never been opened read-only, without doing a database restart.

The benefit of faster Redo Apply failover is that it enables customers to recover from a failure or outage much faster, increasing the degree of high availability of the system.

Faster SQL Apply Failover

Failover to a logical standby database can now be completed much faster. It is no longer necessary to restart SQL Apply as part of the failover operation.

The benefit of faster SQL Apply failover is that it enables recovery from a failure or outage much faster, increasing the degree of high availability of the system.

Database Transport Across Same Endian Platforms

This transport feature provides a fast method to move a database from one platform to another.

The ability to migrate databases from one platform to another has become more important for Oracle customers. Oracle Data Pump can export the entire database, but it may be time consuming because it uses the SQL layer to process data. Database transport across the same Endian platforms provides a faster and easier way to move Oracle databases from one platform to another.

Add and Create a Transportable Tablespace to a Database

This feature allows a DBA to:

• Generate a tablespace set from the source database and plug it into the target database, essentially automating movement of tablespaces between the databases.

• Generate a tablespace set which can be used to transfer a tablespace to any compatible database.

• Plug-in the tablespace set (provided by the user) into the target database and perform an import operation.

Using this tool, database administrators can automate the process of transporting tablespaces across databases. Tablespaces can be transported between databases running on different platforms using different supported versions. This user-friendly tool performs checks and validations, which are needed to ensure valid and compatible tablespace sets for plugging into the target databases.

Additional Data Type Support in LogMiner and SQL Apply

SQL Apply, LogMiner, and Streams now support IOTs with overflow segments and LOBs.

This feature allows Data Guard SQL Apply and LogMiner to be used for protection of a broader range of data types. Similarly, Streams can be used for replication of a broader range of data types.

Automatic Deletion of Applied Archive Logs

Archived logs, once they are applied on the logical standby database, will be automatically deleted by SQL Apply.

This feature reduces storage consumption on the logical standby database and improves Data Guard manageability.

Data Guard: Improved Manageability for Redo Transport, Log Apply, and Broker

This feature deprecates unnecessary initialization parameters and attributes for Oracle Database 10g Release 2.

This feature simplifies the usage of initialization parameters and attributes, as well as certain SQL statements.

Dynamic RMAN Channel Allocation for Backing Up and Recovering RAC Database

By configuring parallelism when backing up or recovering a RAC database, RMAN channels are dynamically allocated across all RAC instances. This is similar to dynamic RMAN channel allocation for a single instance.

Channel failover allows a failed operation on one node to be continued on another node.

This feature eliminates manual allocation of RMAN channels for each RAC node. It also provides better tolerance of media failures (that is, when one node fails, another node can continue backup work).

RMAN Backup Encryption

RMAN now creates encrypted backups that cannot be restored by unauthorized people. There are 3 modes of backup encryption:

• Transparent encryption

• Password encryption

• Dual-mode encryption using either transparent or password encryption

Backup encryption is a regulatory requirement for many customers. The presence of this capability in Oracle obviates the need to purchase it elsewhere.

Easy Conversion of a Physical Standby Database to a Reporting Database

A physical standby database may be activated to be a primary database, opened read/write for reporting purposes, then flashed back in the past to be easily converted back to a physical standby, upon which Data Guard will automatically synchronize the standby database with the primary database. This mechanism also allows a physical standby database to be used as part of a clone, test, or development solution. Note that while this database is open for reporting, it will not be able to receive any redo data from the primary database. Therefore, such a configuration may be more suitable for customers with less demanding disaster recovery requirements.

This feature allows the physical standby database to be utilized for reporting, read/write, and cloning activities, enabling customers to extract more value out of their Data Guard investment.

Eliminate Control File Re-Creation

With the control file enhancements, there is no longer a requirement to re-create the control file when changes in the configuration parameters are made. These include the MAXLOGFILE, MAXLOGMEMBERS, MAXLOGHISTORY, MAXDATAFILES, and MAXINSTANCES parameters.

This feature eliminates database downtime when a configuration parameter change is made in the control file.

Enhanced Memory Corruption Checking

A new value of MAX has been added to the database parameter, DB_BLOCK_CHECKSUM, to compute checksums on the reading, updating, and deleting of a block.

This feature enhances the ability to validate the block when it is in memory and not just on disk.

Enhanced RMAN Backup Job Views

The new RMAN views correlate and simplify the metadata of an RMAN backup job. For example, a backup job can be associated with its backup sets, and a backup set with its composite control file, archived logs, or datafiles. Additional details such as I/O file sizes and compression ratio can be found in the new views.

This features enhances the manageability of RMAN backups by providing a SQL interface to find additional statistics on the backup job and files.

Enhanced Server Side Shutdown

When using the SHUTDOWN ABORT command with Oracle Database 10g Release 2 and higher clients, all processes (including, for Unix, the foreground process) will be terminated.

Also, when issuing the STARTUP FORCE command, in previous versions, "Starting ORACLE instance (FORCE)" was written to the alert log. In Oracle Database 10g Release 2 and higher, "Shutting down instance (ABORT)" followed by "Starting ORACLE instance (NORMAL)" are now written to the alert log.

Transparently handled both in the server and the client, all processes will be cleaned up after a SHUTDOWN is issued. This allows a DBA to startup in the same session.

Enhanced Space Management Verification

Additional checking has been added to the DBMS_SPACE_ADMIN package to provide better tools to debug space management problems. Extended syntax is added to the ALTER SYSTEM command to dump the space headers of locally managed tablespace files for added analysis.

This feature enhances the tools used to debug space management-related problems to detect corrupt metadata.

Failover Improvements for Real Application Clusters (RAC) Distributed Transactions

Distributed Transactions in a RAC environment detect failures and start the failover and failback processes automatically.

The benefit of this feature is reduced administration and development time because Oracle manages the distributed transaction, therefore, there is no need to write manual procedures.

Faster Instance Startup for Large Memory Configurations

This feature decreases the time it takes to open Oracle Database instance.

Accelerated instance startup makes Oracle Database available sooner when databases have multigigabyte shared memory configurations.

Fine-Grained, Automated Tracking of SQL Apply Runtime Performance

Several additional views have been provided, and existing views have been enhanced to enable administrators to dynamically track the progress and performance of SQL Apply at a granular level. Enterprise Manager also provides historical trend analysis of the progress and performance data.

This feature makes it considerably easier to maintain and tune SQL Apply configurations and improves Data Guard manageability.

Flashback Across Data Guard Switchovers

It is now possible to flashback the primary and standby databases to an SCN or a point in time prior to the switchover operation. Using flashback in this way on a physical standby database preserves the standby role. Using flashback in this way on a logical standby database changes the role of the standby database to what it was at the target SCN/time.

This extends the flashback window; that is, the possible SCN/time in the past to which any database can be flashed back to. This provides administrators more flexibility to detect and correct human error situations.

Flashback Database Through Resetlogs

This feature supports flexible data recovery that enables movement of the database back and forth in time even after you perform a resetlogs operation.

If you ever need to do a resetlogs operation for recovery purposes, you can take the database back to a point in time before the resetlogs if you find that you made a mistake. This provides administrators more flexibility to detect and correct human error situations.

Hardware Assisted Resilient Data (HARD) Compliant RMAN Backups

RMAN backup pieces can be stored on and validated by HARD-compliant disk hardware.

This prevents corruptions in RMAN backup pieces from being written to permanent storage.

Increase Supported Real Application Clusters Instances for a Database

For Oracle Database 10g Release 2, the control file supports 1,055 instances for a RAC database.

This feature provides enhanced support and flexibility for Grid Management.

Incremental Roll Forward of Database Copy

This feature can be used to create an RMAN incremental backup from a specified SCN. This incremental backup can be used to quickly move a standby database forward in time or load new data into a clone database for testing.

This feature minimizes the time data is unprotected by creating an incremental backup that can be applied to a standby database to quickly make the standby database transactionally consistent with the primary database. This also eliminates the need to continuously re-create a clone database with newer data values.

Online Redefinition Supports Clustered Tables, ADT's, MV Logs, AQ Tables, and Preserves Database Statistics

Online redefinition can now be performed on clustered tables, tables containing Abstract Data Types (ADT), Advanced Queuing (AQ) tables, Materialized View (MV) logs, and it retains all statistics on a table. It also clones statistics, checks and non-NULL constraints, and dependent objects on nested tables. PL/SQL packages do not have to be recompiled if the number and ordering of columns and data types in the table remain unchanged.

This feature means less downtime as more tables are now supported for online redefinition. Analysis of the table after reorganization is no longer required and PL/SQL packages remain available following an online redefinition, thus maintaining application availability. These enhancements are especially beneficial to customers using Oracle Applications.

Online Redefinition of a Single Partition

Individual partitions may now be reorganized online.

With this feature, system downtime is reduced and, during reorganization, storage requirements and rollback space requirements are significantly reduced.

Optimized Creation of Logical Standby Database

Creation of a logical standby database no longer requires the creation of a specialized logical standby control file, which could not be used by RMAN. Logical standby databases can now be created easily from a physical standby database.

This feature reduces specialized manual operations for creating a logical standby database and improves Data Guard manageability.

Unused Block Compression

By optimizing database backups, only blocks currently used by the database are backed up. In previous releases, NULL compression was utilized, whereby only never-used blocks were excluded from backup. With this new feature, all unused blocks, whether they have been used or not in the past, will be excluded from backup.

This enables faster backups and smaller backup files.

Reorganized Object Types Improves User Interface

This feature supports reorganization of additional database object types that are currently not supported (for example, nested tables), enhances capabilities like launch-in-context, and improves usability of the Re-org wizard.

Currently if a tablespace contains objects like nested tables, the tablespace cannot be reorganized. This new feature enhances support for more object types for tablespace reorganization. In addition, this feature provides enhanced support for launch-in-context so that applications such as Policy Violations can pass recommendations to the wizard directly and do not depend on the user to enter the data manually.

Restore Point

This feature provides the ability to associate a user-defined name, Restore Point, with a point-in-time. The Restore Point can be used with a Flashback Database or Flashback Table during recovery operations. Restore Point can be specified such that it guarantees the database can be recovered to a particular point-in-time.

Restore Point eliminates the need to manually record an SCN or timestamp to use for Flashback Database and Flashback Table operations. In addition, a special 'guaranteed' restore point ensures that the database can be flashed back to that particular point-in-time. Using guaranteed restore points can achieve efficient disk space usage, as only those flashback logs necessary to meet the guaranteed restore point are required to be retained.

Restore points can also be created as 'guaranteed', meaning it provides enhanced Flashback Database efficiency in recovery and disk storage requirements. And, it guarantees the database can be rewound to a specified point-in-time.

Temporary Datafiles Are Re-Created on RMAN Recovery

Temporary datafiles that belong to locally managed temporary tablespaces are automatically re-created during recovery operation.

This eliminates the need to manually create temporary tablespaces after recovery.

Enhanced Upsert Abilities in MODEL Clause

This feature provides new behavior supporting upsert of cells. Rules are now able to insert new cells even when the left hand side includes existential predicates such as ANY, IN, >, <, and LIKE.

The benefit is more expressive rules and greater productivity in specifying the MODEL clause.

Enhancements to MODEL Rules Which Use FOR Loops

This feature moves evaluation of MODEL clause rules using FOR loops from compile time to runtime. This permits rules with FOR loops to use reference spreadsheets, nested references, and the ITERATION_NUMBER variable. In certain situations, subqueries within FOR loops can have shared cursors and refer to tables defined in the WITH clause.

The benefit of this feature is that MODEL clauses with certain types of rules perform better, handle larger model sizes, and include more powerful rule expressions.

Use of Analytic Functions in the SQL MODEL Clause

This feature provides direct use of analytic functions like moving average and cumulative sums in SQL MODEL clause rules.

Native support of analytic functions is significantly faster than alternative techniques and simplifies expression of complex calculations, such as allocations and forecasts, in the MODEL clause.

Intra-Partition DML in the Presence of Bitmap Indexes

Intra-Partition DML (PDML) can now be fully leveraged in the presence of bitmap indexes.

With this feature in Oracle Database 10g Release 2, the last main restriction for PDML is lifted.

Increased Maximum Number of Partitions Per Object

The maximum number of partitions is now 1024K-1, increased from a previous limit of 64K-1.

Increasing the maximum number of possible partitions to objects enables finer granularities for partitioned objects and even broader flexibility in choosing the right partitioning approach for a business scenario.

Enhanced Dynamic Partition Pruning

This feature enhances partition pruning for complex queries.

This feature provides enhanced performance for a broader set of complex queries.