Oracle Database 10g : New Features For Administrators Oracle Database 10g: New Features for Administrators Volume III • Student Guide D17079GC20 Edition 2.0 August 2005 D19871 đ Authors Copyright â 2005, Oracle All rights reserved Christine Jeal Disclaimer Janet Stern This document contains proprietary information and is protected by copyright and other intellectual property laws You may copy and print this document solely for your own use in an Oracle training course The document may not be modified or altered in any way Except where your use constitutes "fair use" under copyright law, you may not use, share, download, upload, copy, print, display, perform, reproduce, publish, license, post, transmit, or distribute this document in whole or in part without the express authorization of Oracle Jean-Francois Verrier Technical Contributors and Reviewers Troy Anthony David Austin Mirella Avila The information contained in this document is subject to change without notice If you find any problems in the document, please report them in writing to: Oracle University, 500 Oracle Parkway, Redwood Shores, California 94065 USA This document is not warranted to be error-free John Beresniewicz Tom Best Harald van Breederode Mary Bryksa Timothy Chien Donna Cooksey Andy Fortunak Joel Goodman Donna Hamby Magnus Isaksson Pete Jones Pierre Labrousse Dominique Laugraud Stefan Lindblad Yi Lu Venkat Maddali Paul Manning Sabiha Miri Srinivas Putrevu Mark Ramacher James Spiller S Matt Taylor Branislav Valny Anthony Woodell Editors Atanu Raychaudhuri Aju Kumar Graphic Designer Samir Mozumdar Publisher Joseph Fernandez Restricted Rights Notice If this documentation is delivered to the United States Government or anyone using the documentation on behalf of the United States Government, the following notice is applicable: U.S GOVERNMENT RIGHTS The U.S Government’s rights to use, modify, reproduce, release, perform, display, or disclose these training materials are restricted by the terms of the applicable Oracle license agreement and/or the applicable U.S Government contract Trademark Notice Oracle, JD Edwards, PeopleSoft, and Retek are registered trademarks of Oracle Corporation and/or its affiliates Other names may be trademarks of their respective owners Contents I Introduction Overview I-2 Today’s IT Infrastructure I-3 How DBAs Spent Their Time Before Oracle Database 10g I-4 Grid Computing I-5 Oracle Ecosystem I-6 Enterprise Grid Computing I-7 Implement One from Many I-8 Manage Many as One I-10 Readily Available Product Controls I-12 Oracle Enterprise Manager 10g Grid Control I-13 Oracle Database 10g: The Database for the Grid I-14 Suggested Additional Courses I-15 Oracle Database 10g Manageability Goals I-16 Database Management Challenges I-17 Oracle Database 10g Solution: Self-Managing Database I-18 How Oracle Database 10g DBAs Spend Their Time I-19 Further Information I-20 Suggested Schedule I-21 Installation Objectives 1-2 Installation Methods 1-3 Installation: New Feature Support 1-4 Performance Enhancements to Installation 1-5 Checking the Preinstallation Requirements 1-6 Miscellaneous Installation Enhancements 1-7 Response File Improvements 1-8 CD Pack Contents 1-9 Management Options 1-10 File Storage Options 1-11 Backup and Recovery Options 1-12 Passwords 1-13 Grid Control Components 1-14 Managed Targets for Grid Control 1-16 Oracle Management Service 1-17 Oracle Management Repository 1-18 Grid Control Console: Home 1-19 Grid Control Console: Targets 1-20 Managing Very Large Grids 1-21 Summary 1-22 iii Server Configuration Objectives 2-2 Database Configuration Assistant (DBCA) Enhancements 2-3 Using Database Control for Management 2-5 SYSAUX and DBCA 2-6 Database Control: Overview 2-7 Important EM Agent Directories 2-8 Configuring Database Control 2-9 Managing Database Control Components 2-10 Database Home Page: Overview 2-11 Database Control Setup Pages 2-13 Database Control Preferences Pages 2-14 Net Services Administration Pages 2-15 Database Feature Usage: Statistics Collection 2-16 Database Feature Usage: EM Interface 2-17 Database Feature Usage: HWM Page 2-18 Policy Framework 2-19 Policy Violations Page 2-20 Manage Policy Violations Page 2-22 Rule Violation Page 2-24 Manage Policy Library Page 2-25 Simplified Initialization Parameters 2-26 Viewing Initialization Parameters 2-27 Irreversible Data File Compatibility 2-28 Enabling iSQL*Plus DBA Access 2-29 Accessing iSQL*Plus DBA from EM 2-30 Summary 2-31 Practice 2: Overview 2-32 Data Pump, SQL*Loader, and External Tables Objectives 3-2 Data Pump: Overview 3-3 Data Pump: General Architecture 3-4 Data Pump Export and Import: Overview 3-5 Data Pump Export and Import: Benefits 3-6 General Implementation: Overview 3-8 Data Pump Direct Path: Considerations 3-9 Data Pump File Locations 3-10 Data Pump File Naming and Size 3-12 Data Pump Utility: Interfaces and Modes 3-13 Data Pump and Filtering 3-14 Data Pump Import Transformations 3-15 Data Pump Job Monitoring Views 3-17 iv Database Control and Data Pump 3-18 Using EM to Monitor Data Pump Jobs 3-19 Parallel Full Export and Import: Examples 3-20 Limited Schema Export: Example 3-21 Network Mode Import: Example 3-22 Attaching to Existing Job: Example 3-23 Restarting a Stopped Job: Example 3-24 Data-Only Unload: Example 3-25 Loading Data from Flat Files by Using EM 3-26 External Table Population: Overview 3-27 External Table Population Operation 3-28 External Table Parallel Populate Operation 3-29 External Table Population: Example 3-30 External Table Projected Columns 3-31 External Table Projected Column: Examples 3-32 Summary 3-33 Practice 3: Overview 3-34 Database and Tablespace Data Movements Objectives 4-2 Cross-Platform Transportable Tablespaces 4-3 Minimum Compatibility Level 4-4 Transportable Tablespace Procedure 4-5 Determining the Endian Format of a Platform 4-6 Data File Conversion: Examples 4-7 CLOB Exception 4-8 Using Database Control to Transport Tablespaces 4-9 Generating Transportable Tablespace Sets 4-10 Specifying Destination Characteristics 4-11 Specifying Conversion Information 4-12 Specifying the Files Transfer Mode 4-13 Database Transportation: Overview 4-14 Database Transportation Procedure 4-15 Database Transportation: Conversion 4-16 Database Transportation: Example 4-17 Database Transportation: Example 4-18 Database Transportation: Generated Pfile 4-19 Database Transportation: Generated SQL Script 4-20 Database Transportation: Generated RMAN Script 4-22 Database Transportation: Considerations 4-23 DBMS_TDB Package 4-24 Database Cloning Using Database Control 4-25 Transport Tablespace from Backup: Overview 4-26 Transport Tablespace from Backup: Implementation 4-27 v Transport Tablespace from Backup: Example 4-28 Versioning Tablespaces 4-29 Versioned Tablespaces: Usage Scenarios 4-30 Tablespace Repository 4-31 CLONE_TABLESPACES 4-32 CLONE_TABLESPACES: Example 4-33 DETACH_TABLESPACES 4-34 DETACH_TABLESPACES: Example 4-35 ATTACH_TABLESPACES 4-36 ATTACH_TABLESPACES: Example 4-37 Managing a File Group Repository 4-38 File Group Repository Catalog 4-41 File Group Repository Catalog Views 4-42 File Group Repository Privileges 4-43 Privileges for Versioning Tablespaces 4-45 Summary 4-46 Practice 4: Overview 4-47 Automatic Management Objectives 5-2 Oracle Database 10g Solution: Self-Managing Database 5-3 Automatic Database Diagnostic Monitor: Overview 5-4 ADDM Performance Monitoring 5-5 ADDM Methodology 5-6 Top Performance Issues Detected 5-7 Database Control and ADDM Findings 5-8 ADDM Analysis Results 5-9 ADDM Recommendations 5-10 Database Control and ADDM Task 5-11 Changing ADDM Attributes 5-12 Retrieving ADDM Reports by Using SQL 5-13 Automatic Shared Memory Management: Overview 5-14 SGA Tuning Principles 5-15 Benefits of Automatic Shared Memory Management 5-16 Automatic Shared Memory Management 5-17 SGA Advisor 5-18 Manual Configuration 5-19 Behavior of Autotuned SGA Parameters 5-20 Behavior of Manually Tuned SGA Parameters 5-21 Using the V$PARAMETER View 5-22 Resizing SGA_TARGET 5-23 Disabling Automatic Shared Memory Management 5-24 Manually Resizing Dynamic SGA Parameters 5-25 Automatic Segment Advisor: Overview 5-26 vi AUTO_SPACE_ADVISOR_JOB 5-27 Controlling AUTO_SPACE_ADVISOR_JOB 5-28 Viewing Latest Automatic Segment Advisor Recommendations 5-29 Automatic Optimizer Statistics Collection: Overview 5-30 GATHER_STATS_JOB 5-31 Controlling GATHER_STATS_JOB 5-32 Changing the Default Schedule 5-33 Automatic Undo Retention Tuning 5-34 Automatic Checkpoint Tuning 5-35 Automatically Tuned Multiblock Reads 5-36 Summary 5-37 Practice 5: Overview 5-38 Manageability Infrastructure Objectives 6-2 Oracle Database 10g Solution: Self-Managing Database 6-3 Automatic Workload Repository 6-4 Automatic Workload Repository: Overview 6-5 Automatic Workload Repository Data 6-6 Active Session History 6-7 Generate ASH Reports 6-8 Base Statistics and Metrics 6-9 Workload Repository 6-10 Statistic Levels 6-11 AWR Snapshot Sets 6-12 AWR Snapshot Purging Policy 6-13 Database Control and AWR 6-14 Generating AWR Reports 6-15 Snapshot Sets and Period Comparisons 6-16 Compare Periods: Results 6-17 Statspack and AWR 6-18 Server-Generated Alerts 6-19 Server-Generated Alerts: Overview 6-20 Alert Models Architecture 6-21 Server-Generated Alert Types 6-22 Readily Available Server-Generated Alerts 6-23 Database Control Usage Model 6-24 Database Control Interface to Alerts 6-25 Setting Alert Thresholds 6-26 Alerts Notification 6-27 Metric Details Severity History 6-28 Metric and Alert Views 6-29 PL/SQL Interface for Threshold Settings 6-30 Alert Consumption: Manual Configuration 6-31 vii User-Defined SQL Metrics 6-32 Automatic Routine Administration Tasks 6-33 DBCA and Automated Tasks 6-34 Common Manageability Infrastructure: Advisory Framework 6-35 Advisory Framework: Overview 6-36 Typical Advisor Tuning Session 6-37 Database Control and Advisors 6-38 DBMS_ADVISOR Package 6-39 Dictionary Changes 6-40 Using PL/SQL: Example 6-41 Summary 6-42 Practice 6: Overview 6-43 Adaptive Thresholds and Metric Baselines Objectives 7-2 Cyclic Nature of Workloads 7-3 Thresholds and Metric Baselines Overview: Fixed Thresholds 7-4 Thresholds and Metric Baselines Overview: Adaptive Thresholds 7-5 Metric Baselines and Thresholds Concepts 7-6 Metric Baselines and Time Groups 7-8 Enabling Metric Baselining 7-9 Activating the Moving Window Metric Baseline 7-10 Setting Adaptive Alert Thresholds 7-11 Visualizing Metric Baseline Statistics 7-12 Creating Static Metric Baselines 7-13 Time Grouping: Considerations 7-14 Metric Baseline and Adaptive Threshold: Considerations 7-15 Classification of Eligible Metrics 7-16 Metric Baseline: Normalized View 7-17 Configuring Normalization Metrics 7-19 Adaptive Thresholds and the All Metrics Page 7-20 Summary 7-21 Practice 7: Overview 7-22 Application Tuning Objectives 8-2 Oracle Database 10g Solution: Self-Managing Database 8-3 Automatic Statistics Gathering 8-4 Enhanced Query Optimization 8-5 Statistics on Dictionary Objects 8-6 Dictionary Statistics: Best Practices 8-7 Miscellaneous Statistics-Related Changes 8-8 Locking Statistics 8-9 Using the DBMS_STATS Package 8-10 Automatic Statistics Collection: Considerations 8-11 viii History of Optimizer Statistics 8-12 Managing Historical Optimizer Statistics 8-13 Managing Optimizer Statistics Using EM 8-14 DML Table Monitoring Changes 8-15 Rule-Based Optimizer Obsolescence 8-16 Automatic SQL Tuning: Overview 8-17 Application Tuning Challenges 8-18 SQL Tuning Advisor: Overview 8-19 Stale or Missing Object Statistics 8-20 SQL Statement Profiling 8-21 Plan Tuning Flow and SQL Profile Creation 8-22 SQL Tuning Loop 8-23 Access Path Analysis 8-24 SQL Structure Analysis 8-25 SQL Tuning Advisor: Usage Model 8-26 Database Control and SQL Tuning Advisor 8-27 Implementing Recommendations 8-28 DBMS_SQLTUNE Package 8-29 DBMS_SQLTUNE: Examples 8-30 Exporting or Importing SQL Profiles 8-31 Automatic SQL Tuning Categories 8-32 SQL Access Advisor: Overview 8-33 SQL Access Advisor: Usage Model 8-34 Possible Recommendations 8-35 Typical SQL Access Advisor Session 8-36 Workload Source 8-37 Recommendation Options 8-38 Reviewing Recommendations 8-40 SQL Access Advisor: Procedure Flow 8-41 Performance Monitoring Solutions 8-42 Performance Management Approach 8-43 Database Home Page 8-44 Database Performance Page 8-45 Active Sessions Waiting Pages 8-46 SQL Details 8-47 Historical Data View 8-48 Summary 8-49 Practice 8: Overview 8-50 Services Objectives 9-2 Traditional Workload Dispatching 9-3 Grid Workload Dispatching 9-4 What Is a Service? 9-5 ix Quote Operator q • • Eliminates the need to use escape characters for single quotation marks in text literals Supports CHAR and NCHAR literals • Valid values are: – Any single or multibyte character – Paired [ ], { }, ( ), and < > SQL PL/SQL 21-22 SELECT cust_address FROM customers WHERE cust_last_name = q'X 'John's Bait Shop' X'; v_string1 := q'['So,' she said, 'It's finished.']'; v_string2 := q'!name LIKE '%DBMS_%'!'; Copyright © 2005, Oracle All rights reserved Quote Operator q You can eliminate previous additional quotation strings in character string literals by choosing your own quotation mark delimiter It supports both CHAR and NCHAR literals You choose any convenient delimiter, single or multibyte, or any of the [ ], { }, ( ), or < > character pairs The delimiter can even be a single quotation mark However, if the delimiter appears in the text literal itself, ensure that it is not immediately followed by a single quotation mark In the first example in the slide, X is used as the quotation mark delimiter You not need to prefix the single quotation mark inside 'John's Bait Shop' with another single quotation mark The second example shows PL/SQL using the paired brackets [ ] as the delimiter Therefore, you not need to use two single quotation marks inside the character string used to initialize the v_string1 variable This makes your SQL text much more readable The new quote operator can also be used with the new SQL_TUNE function, which takes a SQL statement as its argument For example: EXECUTE SQL_TUNE – ('select * from emp where name LIKE ''%DBMS_%''', ) With the new quote operator, you can rewrite this statement as: EXECUTE SQL_TUNE – (q'!select * from emp where name LIKE '%DBMS_%'!') Oracle Database 10g: New Features for Administrators 21-22 UTL_MAIL Package A user-friendly interface by which to send e-mail: • Hides the underlying SMTP API protocol • Performs a send e-mail operation in one PL/SQL call • Uses parameters for common e-mail elements: – – – – – 21-23 Attachments Sender or recipients Cc or Bcc Subject Message body Copyright © 2005, Oracle All rights reserved UTL_MAIL Package The supplied UTL_MAIL PL/SQL package is a user-friendly utility for managing e-mail It allows you to use well-defined and commonly used e-mail features such as attachments, Cc, and Bcc Using UTL_MAIL, you can send e-mail in a single PL/SQL call and use parameters to pass commonly used e-mail elements such as the sender, receivers, Cc, Bcc, subject, and message body As an example, if you have a bug system and you want an easy interface to send e-mail with a small attachment to a recipient, UTL_MAIL provides a good interface for this task For sending large attachments (larger than 32 KB) to hundreds of recipients, use the UTL_SMTP package The UTL_MAIL package will be enhanced in the future to include greater functionality Note: To install the UTL_MAIL package, you need to run the utlmail.sql and prvtmail.plb scripts located in the $ORACLE_HOME/rdbms/admin directory For more information about the UTL_MAIL package, see the PL/SQL Packages and Types Reference Oracle Database 10g: New Features for Administrators 21-23 UTL_MAIL: Examples execute utl_mail.SEND (sender => 'the.instructor@oracle.com' ,recipients=> 'some valid email address' ,subject => '10gNF, UTL_MAIL Demo' ,message => 'Note: 1st message, no attachment'); execute utl_mail.SEND_ATTACH_VARCHAR2 ( , message => '2nd message, attachment in line' , attachment=> 'The 2nd demo attachment text' , att_inline=> TRUE ); execute utl_mail.SEND_ATTACH_VARCHAR2 ( , att_inline => FALSE , att_filename =>'message.doc' ); 21-24 - - Copyright © 2005, Oracle All rights reserved UTL_MAIL: Examples The first example in the slide shows how you can use the SEND procedure to send an e-mail message without attachments The second example shows you how to use the SEND_ATTACH_VARCHAR2 procedure to send an e-mail with an in-line attachment The ATT_INLINE argument is set to TRUE The third example shows how you can send e-mail messages with out-of-line attachments Note: The ATT_FILENAME argument is the suggested file name if recipients are to save the attachment as a file To send a file as an out-of-line attachment, you need to copy it into a BLOB and then send it as an attachment Oracle Database 10g: New Features for Administrators 21-24 LogMiner Enhancements • Automatically add log for mining: EXECUTE DBMS_LOGMNR.START_LOGMNR( OPTIONS => DBMS_LOGMNR.CONTINUOUS_MINE, STARTTIME => ('25-Sep-2003 12:00:00')); • Generate reconstructed statements without a row ID: EXECUTE DBMS_LOGMNR.START_LOGMNR( OPTIONS => DBMS_LOGMNR.NO_ROWID_IN_STMT); • Remove redo logs from analysis: EXECUTE DBMS_LOGMNR.REMOVE_LOGFILE( LOGFILENAME => '/oradata/arch012.rdo'); 21-25 Copyright © 2005, Oracle All rights reserved LogMiner Enhancements If you use LogMiner against the same database that generated the redo log files, LogMiner scans the control file and determines the redo log files that are needed to satisfy your requested time or system change number (SCN) range LogMiner adds the redo logs from the mining database by default You no longer need to map your time frame to an explicit set of redo log files You must use the CONTINUOUS_MINE option and specify a STARTSCN or STARTTIME If you are not sure of the SCN or time, you can query V$ARCHIVED_LOG to obtain that information You can also use this view to determine what redo log files are available for automatic detection by LogMiner You can use the NO_ROWID_IN_STMT option to disable the generation of physical row identifiers in the reconstructed SQL statements With supplemental logging, the redo stream contains logically unique identifiers for the modified rows, so the physical row identifiers are not needed In previous releases, you used the REMOVEFILE option with ADD_LOGFILE to remove redo log files That option has been deprecated; you can now remove redo log files with the REMOVE_LOGFILE procedure LogMiner provides support for the new BINARY_FLOAT and BINARY_DOUBLE data types LogMiner now supports the mining of multibyte CLOB and NCLOB data types and has support for tables containing LONG and LONG RAW columns Additionally, LogMiner supports the mining of index-organized tables (IOTs) Note: You can now use LogMiner in a shared-server configuration LogMiner requests a dedicated server in this environment Oracle Database 10g: New Features for Administrators 21-25 Enhanced Initialization Parameters The following initialization parameters offer improved corruption diagnostics capability: • DB_BLOCK_CHECKSUM – Values: OFF, TYPICAL, FULL • DB_BLOCK_CHECKING – Values: OFF, PARTIAL, FULL 21-26 Copyright © 2005, Oracle All rights reserved Enhanced Initialization Parameters Oracle Database 10g offers several mechanisms for catching corruptions before they occur and also to back track after a corruption is detected The DB_BLOCK_CHECKSUM parameter determines whether DBWn and the direct loader calculate a checksum (a number calculated from all the bytes stored in the block) and store it in the cache header of every data block when writing it to disk This parameter defaults to FALSE at startup If set to TYPICAL, it enables the previous behavior (compute on write, verify on read) only If the value is set to FULL, the checksum is verified before applying any change and recomputed after the change is completed for an UPDATE or DELETE statement This is achieved by maintaining the checksum while the block is in memory Thus, if a block is corrupted because of a stray write in memory, the checksum will mismatch when doing the next change This mechanism detects corruptions and protects any stray writes from making it to disk The overhead associated with this is 8–10% The DB_BLOCK_CHECKING parameter defaults to FALSE at startup and controls whether the Oracle server performs block checking for data blocks Currently, the check is performed on a disk read or just before preparing to write However, as the size of the block increases, the performance overhead also increases You can set it to TYPICAL to limit any performance impact When set, the Oracle server does not perform the overlap and ordering checking in data and index blocks, which reduces any overhead to one-fourth However, it is possible that corruptions, which would have been caught by the skipped checks, could make it to disk Oracle Database 10g: New Features for Administrators 21-26 Faster Instance Startup for Ultralarge Buffer Caches • • • • 21-27 Database available at 10% initialization of buffer cache Remaining 90% formatted in the background by CKPT Leverages the dynamic buffer cache infrastructure Especially useful for very large buffer caches Copyright © 2005, Oracle All rights reserved Faster Instance Startup for Ultralarge Buffer Caches With ultralarge buffer caches (100 GB or greater), the time taken to initialize the buffer caches can be in the order of several minutes This adds considerably to the time to open the database even when no recovery is required In Oracle Database 10g, only a small portion of the buffer cache (10%) is initialized before startup, with the remaining initialization completed after the startup completes Partial initialization is possible because the entire buffer cache is not required immediately after the database opens; for a very large buffer cache, it takes some time before the workload ramps up to the point of occupying every buffer in the cache This functionality leverages the dynamic buffer cache infrastructure The buffer cache is subsequently formatted in the background by the CKPT process as is done during a normal buffer cache resize operation Note: Any automatic cache resizing cannot occur until the buffer pool has been fully initialized Oracle Database 10g: New Features for Administrators 21-27 Asynchronous COMMIT • • • • The default COMMIT behavior is to wait until redo is saved in the redo log files The default behavior can now be changed to “not wait.” Asynchronous COMMIT is useful for high transaction throughput However, transactions may be lost: – Machine crashes – File I/O problems with redo log files 21-28 Copyright © 2005, Oracle All rights reserved Asynchronous COMMIT When a transaction commits, the log writer (LGWR) process writes redo for the commit, along with the accumulated redo of all changes in the corresponding transaction, to disk By default, the Oracle database writes the redo to disk before the call returns to the client This behavior introduces a latency in the commit because the application must wait for the redo to be written to disk Suppose that you are writing an application that requires very high transaction throughput With Oracle Database 10g, if you are willing to trade commit durability for lower commit latency, then you can change the default COMMIT options so that the application does not need to wait for the Oracle database to write data to the online redo logs Thus, the redo generated for a transaction may not persist when the commit returns to the user This opens a small window of vulnerability where the transaction that purportedly committed could be rolled back The most obvious case is where the machine crashes In this case, any commit redo buffered in the redo log buffers before being written to the online redo log files is also lost Another case could be when you experience file I/O problems with the online redo logs at the point where LGWR actually attempts to force any redo buffered in the redo log buffer to disk If the redo logs are not multiplexed to provide a level of redundancy, then it is possible to lose the commit action Oracle Database 10g: New Features for Administrators 21-28 How to Use Asynchronous COMMIT • Possible combinations: – – – – • IMMEDIATE, WAIT IMMEDIATE, NOWAIT BATCH, WAIT BATCH, NOWAIT System- or session-level examples: ALTER SYSTEM SET COMMIT_WRITE = IMMEDIATE, WAIT ALTER SESSION SET COMMIT_WRITE = IMMEDIATE, NOWAIT • COMMIT statement examples: COMMIT WRITE BATCH WAIT COMMIT WRITE BATCH NOWAIT 21-29 Copyright © 2005, Oracle All rights reserved How to Use Asynchronous COMMIT You can change the commit behavior in the following locations: • COMMIT_WRITE initialization parameter at the system or session level • COMMIT statement The IMMEDIATE option ensures that the redo for the commit of the transaction is written out immediately by the LGWR process; that is, an I/O is initiated The BATCH option means that the redo is buffered and no I/O is initiated However, LGWR is still permitted to write the redo to disk in its own time The WAIT option ensures that the commit does not return until the redo corresponding to the commit is persisted in the online redo logs When you use the NOWAIT option, the commit returns without waiting for the commit redo to be written to the online redo logs In the absence of any options, the defaults are IMMEDIATE and WAIT, which is consistent with earlier database releases You cannot specify both BATCH and IMMEDIATE together, nor can you specify both WAIT and NOWAIT together After the initialization parameter is set, a COMMIT statement with no options conforms to the options specified in the parameter Note: The options in the COMMIT statement override the current settings in the initialization parameter Oracle Database 10g: New Features for Administrators 21-29 Diagnosis of Hung or Extremely Slow Databases • New functionality for problem analysis when the database is performing very slowly, or is hung: – Direct access to SGA for performance monitoring (memory access mode) – V$SESSION – V$SESSION_WAIT – V$SYSTEM_EVENT – V$SYSSTAT – Hang analysis using Enterprise Manager • 21-30 Also supported for Oracle9i databases Copyright © 2005, Oracle All rights reserved Diagnosis of Hung or Extremely Slow Databases Oracle Database 10g introduces this functionality for problem analysis when the database is performing very poorly, or is actually The system supports the collection of real-time performance statistics directly from the System Global Area (SGA) using optimized or lightweight system-level calls as an alternative to SQL In Enterprise Manager, this is referred to as the memory access mode There is one SGA collector thread per Oracle instance, and it is started automatically by the EM agent when it starts monitoring a database instance The V$ views shown in the slide are the main performance views used for high-level performance diagnostics and, therefore, are the ones for which direct access to SGA is available If more extensive drill-down information is required, then you must use SQL to retrieve it Host information, such as the number of CPUs and the host name, are also collected and made visible through the EM interface It also builds on the current ORADEBUG Hang Analysis utility to provide an Enterprise Manager’s hang analysis graphical user interface Oracle Database 10g: New Features for Administrators 21-30 Using Memory Access Mode 21-31 Copyright © 2005, Oracle All rights reserved Using Memory Access Mode You can access the memory access mode graphic page from the home page in the Related Links section by clicking the Monitor in Memory Access Mode link The link takes you to the Performance page in Memory Access view mode As shown on the slide, you have to Enable Memory Access Mode for the first time, and you can disable it later if you want When you want to return to SQL Access view mode, just click the Switch to SQL Access Modes button on the Performance page Memory Access mode avoids the computation associated with parsing and executing SQL statements, thereby making it robust for severe cases of library cache contention that can prevent the instance from being monitored using SQL You should switch to Memory Access mode for slow or systems Pages in Memory Access mode contain data sampled at a higher frequency than pages in SQL mode Charts may appear to be slightly different from SQL mode for this reason Consequently, the Memory Access mode page provides better information about where events begin and end, and you may also detect short-duration events that might otherwise be missed Note: Memory access mode is also called direct SGA attach mode Oracle Database 10g: New Features for Administrators 21-31 Using the Hang Analysis Page 21-32 Copyright © 2005, Oracle All rights reserved Using the Hang Analysis Page ORADEBUG Hang Analysis utility is provided by the database to capture the instantaneous system wait-for-graph This is the building block for Enterprise Manager’s (EM) hang analysis graphical user interface This functionality exists in all post-Oracle9i databases and can be used to perform both instancewide and clusterwide analysis of waiting sessions To analyze hangs and system slowdowns, you can use the Hang Analysis page in the Additional Monitoring Links section of the Performance page in EM You can also use the Blocking Sessions page to display a list of all sessions currently blocking other sessions You use the Hang Analysis page to: • Determine which sessions are causing bottlenecks • Examine session summaries for blocking or blocked sessions This page provides you with a graphical topology of waiting sessions in the system, with blocked sessions appearing below blocking sessions Enterprise Manager assesses the situation based on historical activity and determines which sessions are or likely to be hung, rather than being only instantaneously in a wait state To view summary information about a session, click a session ID in the topology The Session Summary page appears, which displays general information about the selected session You can click View Session Details in the Session Summary page to get further session information, and determine whether terminating the session is beneficial Oracle Database 10g: New Features for Administrators 21-32 Using the Hang Analysis Page (continued) The sessions are displayed in green, yellow, or red depending on the perceived seriousness of the state of the session within the topology You can also select a Zoom Factor percentage based on the desired size of the topology In cases where there are many blocked sessions, you can use a smaller factor to visualize the overall wait information for the system You can then zoom in on relevant portions and read the topology or drill down further You click in an area of the Zoom Factor microwindow to move the topology to the desired viewing position For instance, if you click the left side of the window, the topology shifts to the right to enable you to fully view the left side of the topology Oracle Database 10g: New Features for Administrators 21-33 Summary In this lesson, you should have learned how to: • Monitor transactions • Use SQL to flush the buffer cache • Enable resumable timeout at the instance level • Use regular expressions in your SQL • Use additional linguistic comparison and sorting methods in SQL • Monitor or extremely slow databases 21-34 Copyright © 2005, Oracle All rights reserved Oracle Database 10g: New Features for Administrators 21-34 Practice 21: Overview This practice covers the following topics: • Demonstrating the use of regular expressions for searching capabilities • Demonstrating the use of case-insensitive sorting to affect query output 21-35 Copyright © 2005, Oracle All rights reserved Oracle Database 10g: New Features for Administrators 21-35 ... Note: For more information about the possible library commands, refer to the Oracle Secure Backup Reference guide Oracle Database 10g: New Features for Administrators 15-16 Oracle Secure Backup:... interface Note: For more information about individual rights, refer to the Oracle Secure Backup Administrator’s Guide Oracle Database 10g: New Features for Administrators 15-10 Oracle Secure... © 2005, Oracle All rights reserved Oracle Database 10g: New Features for Administrators 15-2 Data Protection to Tape for the Oracle Stack Oracle Collaboration Suite configuration files Oracle