A score of 20 of the 25 is normally considered a passing score. Use this information to assess your readiness to take the Teradata Certification Exams. When you are satisfied with your answers, click on the SUBMIT button at the end of the test. You will receive immediate notification of your score. 1. Which two design choices can make the joining of a partitioned primary index (PPI) table to a non-partitioned primary index (NPPI) table more efficient? (Choose two.) defining PPI tables to have fewer rather than more rows per partition defining the partitions so that each partition has the same number of rows defining primary index columns that minimize row redistributions defining the partitions so that the most common WHERE conditions eliminate the most partitions 2. Which two statements about a sparse index are true? (Choose two.) It is especially useful for joining to tables that have few rows. It references values that are rarely used in queries. It is a type of secondary index. It is a type of join index. It reduces the storage requirements for the index. 3. What improves the performance of accessing and updating a join index? collecting statistics on the primary index columns of the join index adding the BEFORE JOURNAL option to the join index definition defining the join index in the same database as the base table it references setting the OptimizeJoinIndex flag to TRUE in DBSControl 4. Which three statements are true about hash indexes and single-table join indexes? (Choose three.) The hash index does not need to be dropped if loaded by a MultiLoad. Both implicitly define a direct access path to base table rows. Both can be FALLBACK protected. Both receive their space from permanent space and are stored as separate tables. Both are maintained by the system when their associated base table is updated. 5. Which three applications typically benefit from the use of join indexes? (Choose three.) OLTP OLAP Active Data Warehousing Decision Support Batch Loading 6. What is a potential complication of implementing a multi-table, partially covering join index? The cost plan for this join index is more costly than other plans. It does not have support for TPump. It is not optimized for use on joined tables. There is additional overhead for accessing a base table row through the index. 7. What are two potential advantages of implementing partitioned primary indexes (PPI)? (Choose two.) A more even distribution of rows to the AMPs may result. Load processing may be greatly improved. Partition eliminations might improve full outer join processing. Range-based queries might need fewer disk I/Os. 8. A session is active with a query in progress. Database Query Log (DBQL) has already cached some data for that session. What are two outcomes of ending query logging for that session? (Choose two.) Cached data is flushed and committed immediately. Cache is not immediately flushed. Subsequent queries during that session are not logged. Rows collected for that session are deleted from the log tables. 9. Which three types of limits or rules does the Teradata Dynamic Query Manager (TDQM) allow a DBA to define? (Choose three.) workload limits disk capacity limits object access rules query resource rules network capacity limits 10. Which two sources of information provided by Teradata tools might be used when designing workload rules for Teradata Dynamic Query Manager (TDQM)? (Choose two.) the DBSControl Utility - Performance Settings objects accessed by users from Database Query Log (DBQL) the RDBMS Setup report from Teradata Manager Logon / Logoff usage reports from DBC.Events 11. Which two statements are true about performance periods in Priority Scheduler? (Choose two.) There are five types of performance periods allowed: time, session, period, query, and cycle. A performance period is assigned to only one performance group. Performance periods allow for users assigned to a performance group to be automatically shifted from one allocation group to a different allocation group. Performance Periods allow for a DBA to be notified that a user is using more system resources than planned. 12. Which three are Teradata Index Wizard reports? (Choose three.) Primary Index Analysis Query Cost Analysis Workload Analysis Disk Space Analysis 13. Which three tasks can you perform using Teradata Manager utilities? (Choose three.) analyze crashdumps analyze table headers analyze errors in the event log system tables analyze audit information in the access logging system tables analyze performance 14. An EXPLAIN output contains the following phrase: "by way of a traversal of index #8 extracting row ids only" Which statement is true? A spool file is built from the rowhash found in the cylinder index. A temporary file is accessed with rowID values found in a secondary index. A spool file is built from rowID values found in a secondary index. RowID values are accessed in memory, without the need for a spool file. 15. For which two conditions are sampled statistics adequate? (Choose two.) small tables with high volatility highly skewed data uniformly distributed data very large tables 16. What can you do to store data efficiently? (Choose three.) specify the minimum needed precision for DECIMAL columns use row compression use VARCHAR use value compression use data block compression 17. Consider the following SQL statement: SELECT cust_id, cust_name, cust_addr, SAMPLEID FROM customer_table SAMPLE 10, 20, 30, 20, 10; Which SQL clause can be added to ensure that rows are sampled randomly across the AMPs? WITH RANDOM REPLACEMENT RANDOMIZED ALLOCATION PROPORTIONAL ALLOCATION WITH REPLACEMENT 18. Which statement is true about UPDATE ELSE INSERT (Upsert) processing? Upsert processing supports the use of subqueries. Update and insert privileges on the base table are required regardless of which portion of the statement is performed. When updating a partitioned primary index (PPI) table, the UPDATE clause may modify the value of a partitioning column. When updating a partitioned primary index (PPI) table, the INSERT clause may specify a different partition than the UPDATE clause. 19. Within ordered analytic functions, what is signified by the syntax element UNBOUNDED PRECEDING? the single row before the current row the entire partition before and after the current row the entire partition before the current row the number of rows before the current row 20. Which statement is true about identity columns? System-generated primary keys always ensure row uniqueness. You can add an identity column with the ALTER TABLE command. The identity column used with the DEFAULT column attribute is beneficial for system-generated primary keys. System-generated primary keys are useful when no quality natural key is available. 21. Which two statements are true about macros as compared to stored procedures? (Choose two.) Macros allow for greater parallelism with multi-statement requests. Macros cannot return EXPLAIN text while stored procedures can. For very simple requests, macros are usually a better choice than stored procedures. Macros require a certain overhead that stored procedures do not. 22. A row is inserted into a referencing table and foreign key columns are defined to be NOT NULL. What two verifications does the system make? (Choose two.) A row exists in the referenced table with the same value(s) as those in the primary key of the referencing table. The foreign key column value is NOT NULL. A row exists in the referenced table with the same value(s) as that in the foreign key column(s). A row exists in the reference index subtable with a zero count. A row exists in the referenced table that is NOT NULL. 23. What are three characteristics of TPump? (Choose three.) can change the rate at which statements are sent to be processed can utilize stored procedures supports conditional processing has no session limit 24. Which three are characteristics of Active Data Warehousing? (Choose three.) provides scalability to support a large amount of detailed data allows users to directly update the Operational Data Store provides an integrated environment that supports strategic and tactical queries must only be loaded using real-time data feeds 25. What are three benefits of using queues to support near real-time loading? (Choose three.) presequence the data so the TPump SERIALIZE option is not needed alleviates the need to merge data from multiple locations guaranteed delivery fewer TPump sessions control the flow of data to the load utility