13.9. Answers to Review Questions

  1. C. The data dictionary view DBA_ADVISOR_ACTIONS contains the SQL statement(s) that the Segment Advisor supplies to implement its recommendation for segment maintenance. DBA_ADVISOR_FINDINGS contains the results of the analysis, but no SQL. DBA_ADVISOR_RECOMMENDATIONS presents one or more findings and the benefits for performing the recommendation. DBA_ADVISOR_RATIONALE provides a more detailed set of reasons why the recommendation should be implemented, along with the impact of not performing the recommendation.

  2. A. The new access path in a sorted hash cluster is used only if an equality predicate is used.

  3. B. Even if the size of the undo tablespace is adjusted after an undo space problem, only one alert is sent for each 24-hour period. Therefore, the only way that the problem will be resolved promptly is for SCOTT to call the DBA, because the DBA will not receive another alert until the next day when another query fails.

  4. A. The new background process MMON checks for threshold violations every 10 minutes. An alert is triggered when the threshold is reached or is cleared.

  5. D. FAST_START_MTTR_TARGET specifies the desired time, in seconds, for instance recovery after a crash or an instance failure. Therefore, the Redo Logfile Size Advisor uses this value to determine the optimal log file size. OPTIMAL_LOGFILE_SIZE is not an initialization parameter but a column in the view V$INSTANCE_RECOVERY. The initialization parameter FAST_START_IO_TARGET specifies recovery at the I/O level, and LOG_CHECKPOINT_INTERVAL specifies the frequency of checkpoints in terms of redo log file blocks used.

  6. D. While some chained rows may be fixed with segment shrink functionality, it is not guaranteed that all chained rows will be fixed, because not all blocks may be read in a segment shrink operation.

  7. B. Only PCTFREE is used in the calculation, because it is the amount of space to leave free in the block for updates to existing rows. PCTUSED is not needed unless the segment space management is not AUTO. In addition, extent sizes calculated by this feature help assess the impact on the tablespace where this segment will be stored.

  8. C. While cluster key values in a sorted hash cluster are hashed, this is also true of regular hash clusters, and therefore is not a benefit unique to sorted hash clusters.

  9. A. While the segment shrink operation could combine steps 1 and 4, the impact to the users will most likely be lessened by performing two smaller operations instead of one by specifying SHRINK SPACE COMPACT before specifying SHRINK SPACE in a subsequent operation.

  10. D. The call to DBMS_SERVER_ALERT.SET_THRESHOLD must specify the metric TABLESPACE_PCT_FULL, the two thresholds, an object type of tablespace, and the tablespace name itself. Specifying NULL for the tablespace name will set the threshold for all tablespaces, not just the UNDOTBS1 tablespace.

  11. A. Because the ROWIDs are changed with a segment shrink operation, tables with ROWID-based materialized views cannot be shrunk unless the materialized views are dropped and re-created after the segment shrink operation.

  12. D. The Segment Advisor is not used to find tables with chained rows, but instead is used for finding segments that are good candidates for segment shrink or may be growing too fast.

  13. A, E. If a query on a sorted hash cluster retrieves rows and an ORDER BY clause specifies either non-sort columns or a suffix of the sort columns, additional sorting is required, assuming that indexes are not defined on the columns in the ORDER BY clause.

  14. B. For segments in tablespaces with automatic segment space management, LOB segments cannot be shrunk. In addition, tables with LONG columns, on-commit materialized views, and ROWID-based materialized view cannot be shrunk. In all cases, shrink operations cannot be performed on segments managed by freelists.

  15. A. Whether you rebuild the index offline or online, you temporarily need twice as much disk space. If you rebuild the index online, you also need disk space to support a journal table to hold the intermediate changes to the index while the index is being rebuilt.

  16. B. When adding log file members, specify the group number or specify all the existing group members.

  17. B. Hash clusters, not index clusters, use a hashing function to find a row in a cluster and perform best for equivalence queries.

  18. A, C. Unless an ORDER BY clause forces a sort operation and uses up temporary space, a SELECT command will not otherwise trigger Resumable Space Allocation. Dropping a table does not use any additional disk space in a tablespace; it frees up disk space. Therefore, it will not trigger Resumable Space Allocation.

  19. B, E. The AFTER SUSPEND system-level trigger can be defined only at the database level or at the schema level.

  20. B. An index-organized table (IOT) does not have a physical ROWID; instead, a logical ROWID is constructed based on the value of the primary key.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset