As Published In March/April 2008

TECHNOLOGY: OCP

Upgrading Manageability

Concepts and sample questions for OCP upgrade certification to Oracle Database 11g

Oracle Database 11g offers significant enhancements in manageability, performance, and high availability. This column focuses on Oracle Database 11g's new features for change and memory management and enhancements in storage management. It also presents sample questions of the type you may encounter when taking the Oracle Database 11g: New Features for Administrators exam, which enables Oracle Certified Professionals (OCPs) certified on Oracle Database 10g to upgrade their certifications to Oracle Database 11g.

Change Management: Database Replay

Change is constant, and it is also the most common cause of instability. Realistic testing of a changing system before its move to production has been impossible, because enterprise production systems are complex and it is very difficult to simulate actual workloads. Without realistic testing, organizations are reluctant to make changes, such as upgrading the database or modifying the configuration, which can prevent them from adopting new competitive technologies.

Oracle Database 11g introduces the database replay feature, which minimizes the risk of change. By capturing a workload on the production system and replaying it on a test system, database replay enables you to realistically test system changes by essentially re-creating the production workload environment on the test system.

Which two statements about database replay are correct?

A. Workload information is captured in an external table.
B. Files created during the workload capture are not modified.
C. The recording infrastructure forces the database to operate in READ_ONLY mode.
D. During recording, the background jobs and all internal clients continue their execution without their actions being recorded.

The correct answers are B and D. Files created during the workload capture are not modified. Instead, as the workload capture data is processed, new workload replay-specific metadata files are created. During recording, the background jobs and all internal clients continue their work without being recorded. Answer A is incorrect because the workload information is captured to a file. Answer C is incorrect because you need to start the database in RESTRICTED, not READ_ONLY, mode.

Which two database operations can be captured as part of the workload capture?

A. Local transactions
B. Session switching
C. Import/export
D. Flashback query

The correct answers are A and B. Local transactions and session switching operations can be captured as part of the workload capture. Answers C and D are incorrect because import and export operations and flashback queries are not supported as part of the workload capture.

Your company decides to move from a single-node installation of Oracle Database 11g to a two-node Oracle Real Application Clusters (Oracle RAC) database. What is the best way to analyze the performance of the new Oracle RAC database installation and test how clients will connect to it before moving to the new environment?

A. Create a test Oracle RAC database and re-create the workload on it.
B. Create a standby Oracle RAC database and re-create all the sessions on it.
C. Capture the workload on the production database and replay it on a test Oracle RAC database.
D. Export the whole database from the production database, import it to the test Oracle RAC database, and re-create all the client sessions.

The correct answer is C. The database replay feature allows you to capture the workload on the production system and replay it on a test system to analyze the performance implications. The options in answers A, B, and D can be used, but they will not be able to simulate the exact production database workload and are also more complex to implement compared to database replay.

Managing Memory: Automatic Memory Management

Oracle strongly recommends the use of automatic memory management to manage the memory on your system. Automatic memory management enables Oracle Database to automatically manage and tune the instance memory. The database instance tunes memory to the target memory size, redistributing memory as needed between the system global area (SGA) and the instance program global area (PGA). This indirect memory transfer relies on the OS mechanism of freeing shared memory. After memory is released to the OS, other components can allocate memory by requesting it from the OS. This helps in maximum memory utilization and eliminates out-of-memory errors.

Which two statements are true?

Configuring your database instance for automatic memory management

A. Helps eliminate out-of-memory errors
B. Allows the instance to automatically adapt to workload changes
C. Ensures uniform allocation of memory to the SGA and the PGA
D. Stops you from setting lower-bound values for various caches

The correct answers are A and B. Enabling automatic memory management enables the system to automatically adapt to workload changes by automating the sizing of the SGA and PGA. This results in maximum memory utilization and helps eliminate out-of-memory errors. Answer C is incorrect because there is an indirect transfer of memory from the SGA to the PGA (and vice versa) but the size of the SGA and the PGA is not uniform. Answer D is incorrect because you will be able to set lower-bound values for caches even if you enable automatic memory tuning.

LMN Corp. has Oracle Database 11g installed on its production system. The DBA has observed that the system is experiencing frequent out-of-memory conditions and has decided to enable automatic memory management to address this issue. Which two parameters must the DBA set to enable this?

A. MEMORY_TARGET
B. SGA_TARGET
C. SGA_MAX_SIZE
D. MEMORY_MAX_TARGET

The correct answers are A and D. To enable the database to automatically manage and tune memory, you have to set only the target memory size initialization parameter (MEMORY_TARGET) and the maximum memory size initialization parameter (MEMORY_MAX_TARGET).

Storage Management: ASM Enhancements

Oracle Database 11g introduces new features in Automatic Storage Management (ASM) that enhance performance and improve availability. The ASM fast mirror resync feature significantly reduces the time required to restore the redundancy of an ASM disk group after a transient disk path failure—a failure caused by a disk path malfunction such as a cable, host bus adapter, or controller failure. ASM fast mirror resync keeps track of pending changes to extents on an OFFLINE disk during an outage. The extents are resynchronized when the disk is brought back online.

In Oracle Database 11g, you can configure your database to read from a particular failure group extent by configuring preferred read failure groups. This feature is useful in extended clusters in which remote nodes have asymmetric access.

What are the advantages of setting up ASM fast mirror resync?

A. It automatically takes the failed disk in a disk group offline.
B. It helps a failure group recover faster from a transient failure.
C. It enables you to define the time interval for repairing a disk and bringing it back online.
D. It improves performance by reading from a copy of the extent closest to the node when there is a failure in one of the disk groups.

The correct answers are A, B, and C. Whenever there is a transient failure, the failed disk is taken offline and not dropped from the disk group, resulting in faster recovery from such a failure. The ASM fast mirror resync feature enables you to set a value to the DISK_REPAIR_TIME attribute that defines the time interval for repairing the disk and bringing it back online. Answer D is incorrect because reading from a copy of the extent closest to the node during a failure is a characteristic of another new feature, called ASM preferred mirror read.

When is the ASM preferred mirror read feature useful?

A. When you have an extended cluster configuration and the remote nodes have asymmetric access with respect to performance
B. When you want to enable faster file opens
C. When you want faster recovery from a disk failure
D. When you want to improve the overall rebalance throughput

The correct answer is A. The ASM preferred mirror read feature enables more-efficient use of network resources, by eliminating the need to use the network to perform read operations. Answer B is incorrect because the ASM preferred mirror read feature does not enable faster file reopens, although with Oracle Database 11g, ASM file extent management has been enhanced to improve performance. Answer C is incorrect because you can obtain faster recovery from disk failure by setting up ASM fast mirror resync. Answer D is incorrect because rebalancing is achieved through the ASM fast rebalance feature.

Conclusion

This column focused on new change and memory management features as well as storage enhancements introduced in Oracle Database 11g:  

• Database replay helps you replicate the system workload, reduce the cost of testing and validation, and identify the full impact of changes before going live with a system change or upgrade. 

• With automatic memory management, the system automatically adapts to workload changes by managing an indirect transfer of memory from the SGA to the PGA (and vice versa). Because the target memory initialization parameter is dynamic, you can change the target memory size anytime without restarting the database. 

• New features and utilities introduced in ASM improve performance and reduce downtime. ASM fast mirror resync quickly resynchronizes ASM disks within a disk group after transient disk path failures. The ASM preferred mirror read feature improves efficiency by allowing a node to read from the failure group extent that is closest to the node.

Send us your comments