ETL_TERADATA_架构设计考试.doc

Exam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Question: 1For which two reasons would Partitioned Primary Index (PPI) be considered on a table? (Choose two.)A. to eliminate ranges of data values from a query and thereby reduce scan timeB. to reduce response time by placing all data in the same partition on the same AMP C. to reduce the amount of I/O needed to process queriesD. to reduce the number of columns in a table to those needed by most queriesAnswer: A, CQuestion: 2How is system availability guaranteed, if a node fails in a MPP environment with multi-node cliques?A. The AMPs on the failed node migrate to the other nodes in the clique.B. The remaining nodes migrate to an alternate clique.C. The RAID groups owned by the AMPs on the failing nodes are shut down.D. The RAID groups owned by the AMPs on the failing nodes migrate to AMPs already running on the remaining nodes in the clique.Answer: AQuestion: 3On the subject of FastLoad, which two statements are correct? (Choose two.)A. no journalingB. can maintain NUSIs but not USIsC. no secondary indexes on the target tableD. loads duplicate rows into a multiset tableAnswer: A, CQuestion: 4Which of the following applies to Non-Partitioned Primary Index (NPPI) tables when determining the size of a row in Teradata? (Choose two.)A. two bytes for variable length accom modationB. eight bytes per row for the rowidC. two bytes per row for row lengthD. one byte per column for com pressionE. four bytes for the primary indexAnswer: B, CQuestion: 5Which two statements referring to a dimensional data model are correct? (Choose two.)A. It often implements fewer tables.B. It is extremely flexible.C. It requires significantly less ETL activity. D. It reduces the need for summary tables.E. It can be adapted to optimize performance.Answer: A, EPage 1 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Question: 6What type of query population is helped by a sparse index?A. queries that often redistribute a large tableB. queries that repeatedly use the same subset of the tableC. queries that often return a large percent of the rows of a large table based on the primary keyD. queries that often eliminate a large percent of the rows of a large table using a superset of the tableE. queries that rarely redistribute a large table in com plex join processingAnswer: BQuestion: 7When is a fully covering join index advantageous to a query?A. when the base table is small enough to fit in cacheB. when there is an unpredictable selection of columns from the base tableC. when relatively few columns are in the result and join or selection criteria includes values other than the base tables primary index(es)D. when the result set is produced through either join or selection criteria on the base tablesprimary index(es)Answer: CQuestion: 8The allocation of system CPU resources in a mixed workload environment can be improved by executing which three of the following tasks? (Choose three.)A. use the MODIFY USER statement to change the ACCOUNT for a userB. modify the Default Resource Partition to a lower weight than other Resource PartitionsC. use the MODIFY USER statement to give a user more spool space D. maintain development and test users in separate Allocation Groups E. define Allocation Groups for different times of the dayAnswer: A, D, EQuestion: 9Which of the following are two true statements regarding soft Referential Integrity (RI)? (Choose two.)A. The RI is presumed to be enforced outside of the database.B. The RI is enforced by the database.C. The PK columns can appear anywhere in the WHERE clause. D. Redundant joins between soft RI defined tables are eliminated.Answer: A, DQuestion: 10A join index would be added to a set of tables for what reason?A. to enforce Referential Integrity across tablesB. to ensure same distribution of both tablesC. to enforce data uniquenessD. to decrease response time on well known queriesPage 2 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Answer: DQuestion: 11Which are two features of the Multi Load utility? (Choose two.)A. Up to ten tables can be specified in a single execution.B. The upsert function is supported.C. Insert, update, and delete are all supported. D. It uses macros to improve performance.E. The processing rate can be controlled.Answer: B, CQuestion: 12Which two of the following statements are correct in relation to data marts? (Choose two.)A. Dependent data marts are isolated entities entirely separate from the enterprise datawarehouse.B. A data mart is generally a relatively small application-specific subset of the data warehouse.C. A logical data mart is a form of dependent data mart that is constructed virtually from the physical data warehouse.D. Independent data marts are derived from the enterprise data warehouse.Answer: B, CQuestion: 13A Partitioned Primary Index (PPI), NUSIs, and a USI is defined on Table X. which two of the following can you employ for daily maintenance? (Choose two.)A. Multi LoadB. TPumpC. FastLoad into an empty Table Y, then INSERT/SELECT into Table X D. Teradata Warehouse Builder Load OperatorAnswer: B, CQuestion: 14Which three of the following should be included in an analysis that defines TDQM restrictions?(Choose three.)A. a list of users whose access must be above restrictionB. a listing of the requirements for critical applications and the times of day that are most criticalC. the client resources available during the peak periodsD. an understanding of the timing requirements for maintenance of specific tables or databasesE. a projection of the growth pattern for major tables in the databaseAnswer: A, B, DQuestion: 15What are you calculating when you analyze volatility as a factor for selecting indexes?A. how unique the column values areB. how frequently the column value changesC. the difference between the maximum and minimum valuesD. how frequently the column is used for accessPage 3 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Answer: BQuestion: 16An empty target table is required by which two utilities when loading data? (Choose two.)A. FastExportB. FastLoadC. MultiLoadD. Teradata Warehouse Builder Load operatorAnswer: B, DQuestion: 17A RAID configuration for system availability has what advantage?A. If a node fails, the AMPs running on that node migrate to the other nodes in a different clique.B. If a single drive fails, work can proceed because the disk controller can reconstruct the data.C. If multiple drives fail in the same drive group, work can proceed using the redundant copies.D. A disk array cabinet failure does not interrupt operation since RAID forces redundancy acrossRAID cabinets.Answer: BQuestion: 18If you are defining Roles in Teradata, what are two questions that you should ask yourself?(Choose two.)A. Will this user need more than just SELECT access to the data?B. From which network path will this user access Teradata?C. What characteristics are important for this users password controls? D. Will this user be assigned to perform more than one job function?Answer: A, DQuestion: 19For which two reasons would you choose a NUPI? (Choose two.)A. It allows clustering of data rows.B. It facilitates AMP-local joins of matching sets. C. It is a closer match for the primary key.D. It allows the user to generate a much larger table. E. It allows duplicate rows in a set table.Answer: A, BQuestion: 20Fast Export is best described by which three of the following? (Choose three.)A. exports from multiple tablesB. uses Support EnvironmentC. uses multiple sessionsD. exports data to multiple files without use of OUTMODAnswer: A, B, CPage 4 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Question: 21The difference between the COPY and RESTORE operations of the ARC utility is best described by which two of the following statements? (Choose two.)A. A RESTORE operation moves the data from an archived file back to any Teradata systemand creates a new table if one does not already exist on the target database.B. A COPY operation moves the data from an archived file back to the same Teradata system from which it was archived or moves data to a different Teradata system as long as the database DBC is already restored.C. A COPY operation moves the data from an archived file back to any Teradata system and creates a new table if one does not already exist on the target database.D. A RESTORE operation moves the data from an archived file back to the same Teradata system from which it was archived or moves data to a different Teradata system as long as the database DBC is already restored.Answer: C, DQuestion: 22Which are three factual statements concerning the creation and use of indexes? (Choose three.)A. All secondary, hash, and join indexes are stored in sub tables, so they require additionalstorage space.B. Index sub tables are updated each time an indexed column value in the base table is updatedor deleted.C. Hash indexes are created in order to allow the Optimizer to choose hash joins.D. A Unique Secondary Index entry might not reside on the same AMP as the base table row.Answer: A, B, DQuestion: 23For what two reasons are allowing columns to follow consistent data types administered at the domain level important? (Choose two.)A. reduces data redistributionB. reduces data storage requirementsC. avoids data conversionsD. eliminates data com pressionE. eliminates need for constraintsAnswer: A, CQuestion: 24Com s data warehouse environment has critical data that needs to be available 24hours a day, 7 days a week. How can continual availability be ensured?A. archive data and keep a copy of the archived data offsiteB. develop and implement a strategy for simultaneously or near simultaneous loading of critical data to an offsite dual systemC. send archived data to the offsite system to be loaded on a regular basisD. keep a dual system offsite and restore the data as neededAnswer: BQuestion: 25Page 5 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Sufficient data protection can be provided by a backup of your entire database using the ARCutility, in which three of the following circumstances? (Choose three.)A. when there are accidentally dropped tables, views, or macrosB. when there is a loss of the data on multiple AMPs within the same clusterC. when there is a loss of data on an AMP for non-Fallback tablesD. when global temporary tables are in use at the time of the archive operationAnswer: A, B, CQuestion: 26A table that has a join index can be updated using which two methods? (Choose two.)A. MultiLoadB. TPumpC. FastLoad into an empty table and INSERT/SELECT into the target tableD. Teradata Warehouse Builder Apply OperatorAnswer: B, CQuestion: 27When you run a query against a table without Fallback if an AMP is down, how will it be affected?A. The query will execute only if it does not require data from the down AMP.B. The query will always execute and return data from only the AMPs that are available. C. The query will not execute under these conditions.D. The query will execute and return data from all AMPs.Answer: AQuestion: 28Which of the following is an accurate statement when creating an identity column asGENERATED ALWAYS?A. The column may be updated by an SQL statement.B. The column may be used to define the primary index.C. The column is used to facilitate copying data from one table into another. D. The column may be used in a join index.Answer: BQuestion: 29For columns with which two characteristics should you consider the com pression allowed inTeradata, when designing tables? (Choose two.)A. numeric columns with a large number of leading zerosB. columns with a large number of null occurrencesC. variable length character columns with nullsD. columns with a large number of default occurrencesAnswer: B, DQuestion: 30You are a database administrator at Com . When determining the workload rules forTDQM, which of the followingare two sets of information provided by Teradata tools that youPage 6 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110could use? (Choose two.)A. the RDBMS Setup report from Teradata ManagerB. the DBSControl Utility - Performance SettingsC. objects accessed by users in DBQL reportsD. Logon / Off usage reports from Teradata ManagerAnswer: C, DQuestion: 31Which three of the following aspects should be taken into consideration when a column is chosen as a primary index? (Choose three.)A. frequency of accessB. distribution across the AMPsC. data typeD. stability of the data valuesAnswer: A, B, DQuestion: 32Which two of the following statements are true with regard to lock levels? (Choose two.)A. A read lock ensures that no modification of that resource is permitted while a SELECToperation is in progress.B. A write lock ensures that the requester has exclusive rights to the locked resource so that no other process can read from, write to, or access the locked resource in any way.C. A write lock ensures that the requester has exclusive rights to the locked resource except for readers not concerned with data consistency.D. A read lock permits modifications on the underlying data while a SELECT operation is inprogress.Answer: A, CQuestion: 33You are an administrator at Com . Com sapplication finds the Partitioned Primary Index to be valuable for avoiding scans on large tables, and it has also established that primary index access (without specifying the partition key) takes significantly longer. Which of thefollowing is an option that will improve this?A. COLLECT STATISTICS on the partitioning columnB. create a non-unique secondary index on the partitioning keyC. COLLECT STATISTICS on the primary index columns individuallyD. create a unique secondary index on the columns of the primary indexAnswer: DQuestion: 34Which two of the following are features of the SyncScan operation in Teradata? (Choose two.)A. allows users to control which AMPs are to be processed first during a retrieval operationB. is enabled by the memory allocation specified in the DBSControl utilityC. allows the system to reduce the processing speed of inserts into tablesD. is especially powerful for transaction processing applicationsE. allows more than one user to share read buffers for a tablePage 7 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110Answer: B, EQuestion: 35Which formula is appropriate for calculating the base table size with Fallback?A. (Number of blocks*Typical block size) - (Number of header bytes*2)B. (Number of blocks*Typical block size) + (Number of header bytes*2) C. 2(Number of blocks*Typical block size) - Number of header bytesD. 2(Number of blocks*Typical block size) + Number of header bytesAnswer: DQuestion: 36Which are two pieces of information from DBQL that may be used to define a workload policy forTDQM? (Choose two.)A. the typical processing time for queries during the defined timeB. the occurrence of full table scans during the specified timeC. the size of the tables involved in queries during the defined timeD. the number of AMPs running at 100% CPU during the specified timeAnswer: A, BQuestion: 37Which two are inaccurate statements about Permanent Journals? (Choose two.)A. They can store both before and after images (copies) of rows changed in one or more tables.B. They store an image of each table that has been altered using any DML statement.C. They store images (copies) of each table both before and, optionally, after they have been updated.D. They store images (copies) of each data row that has been changed with an insert, update or delete statement.Answer: B, CQuestion: 38Permanent Journaling allows continuous operation In which two of the following ways? (Choose two.)A. All archive related activities on a Permanent Journal take significantly less time than anarchive of the referenced table.B. An archive of a Permanent Journal sub table places locks on the underlying table(s).C. If the Permanent Journal includes After Images, the table can be archived without impeding concurrent update activity in progress.D. It is not necessary to checkpoint a Permanent Journal when using a Group Read Lock.Answer: A, CQuestion: 39Which of the following are com patible types of Teradata locks? (Choose two.)A. a write lock and a write lockB. a read lock and a read lockC. a read lock and a write lockPage 8 of 24 SelfExamEExam Name:Teradata Design Architecture V2R5Exam TypeTeradataExam Code:NR0-015Total Questions:110D. a write lock and an access lockAnswer: B, DQuestion: 40Permanent space has which two features? (Choose two.)A. It is limited at the table lev