单选题You are responsible for a data warehouse application that uses records from an external table to update one of the dimension tables periodically. The records in the external table may contain data for new rows in the dimension table， or for updates to 

第1题：

You are planning to import customer data from the Customer table and the Country table in your SQL Server 2000 database into a SQL Server 2005 database. You want to ensure that during the import process，each value in the CountryCode column of the Customer table has a corresponding record in the CountryCode column in the Country table of the SQL Server 2005 database. You define a foreign key between these two tables. Thus，referential integrity will guarantee that the import routine fails if there are any records where a CountryCode value does not exist in the Country table but does exist in the Customer table. You need to ensure that the import process does not fail if records are encountered that do not exist in the Country table. What should you do？（）

• A、Drop the foreign key. Import the data by using the script. Re-create the foreign key.
• B、Create a CHECK constraint.
• C、Create a DML INSTEAD OF trigger that writes the failed records to a file or table.
• D、Create a DML AFTER trigger that writes the failed records to a file or table.

第2题：

Which three are true?（）

• A、A MERGE statement is used to merge the data of one table with data from another.
• B、A MERGE statement replaces the data of one table with that of another.
• C、A MERGE statement can be used to insert new rows into a table.
• D、A MERGE statement can be used to update existing rows in a table.

第3题：

You are responsible for a data warehouse application that uses records from an external table to update one of the dimension tables periodically. The records in the external table may contain data for new rows in the dimension table， or for updates to its existing rows. Which type of SQL command would you use to transfer the data from the external table to the dimension table as efficiently as possible？（）

• A、MERGE
• B、SELECT …CROSS JOIN
• C、INSERT ALL …SELECT
• D、CREATE VIEW …CONSTRAINT

第4题：

In which scenario will you create a sorted hash cluster？ （）

• A、 if the application uses queries joining tables only occasionally
• B、 if the full table scan is executed often on only one table of the clustered tables
• C、 if the data for all the rows of a cluster key value exceeds one or two Oracle blocks
• D、 when you want the data to be returned automatically in the chronological order

第5题：

You added a PHONE_NUMBER column of NUMBER data type to an existing EMPLOYEES table. The EMPLOYEES table already contains records of 100 employees. Now, you want to enter the phone numbers of each of the 100 employees into the table. Some of the employees may not have a phone number available. Which data manipulation operation do you perform?（）

• A、MERGE
• B、INSERT
• C、UPDATE
• D、ADD
• E、ENTER
• F、You cannot enter the phone numbers for the existing employee records.

第6题：

You need to ensure the Order Control team is able to track the number of orders that reach the production department. What should you do?（）

• A、Create a new Time dimension based on the Holiday table. Use derived columns for the month and the year.
• B、Create a Server Time dimension. Within this dimension， enable the time periods year， month， and date. Select the Additional Manufacturing calendar option.
• C、Create a new Time dimension based on the date values within the ProductOrderStatus table. Flag those days on which no orders were received as Holiday.
• D、Create a new Time dimension based on a data source view named query. Derive the data for this data source view from unique dates within the ProductOrderStatus table. Use a subquery to identify the holidays and derived columns for the month and the year.

第7题：

第8题：

第9题：

第10题：

第11题：

第12题：

第13题：

You design a Business Intelligence （BI） solution by using SQL Server 2008. You create a SQL Server 2008 Integration Services （SSIS） package to perform an extract， transform， and load （ETL） process to load data to a DimCustomer dimension table that contains 1 million rows. Your data flow uses the following components： A SQL Destination data flow task to insert new customers An OLE DB Command transform that updates existing customers On average， 25 percent of existing customer records is updated each night. You need to reduce the amount of time required to update customer records. What should you do？（）

• A、 Modify the UPDATE statement in the OLE DB Command transform to use the PAGLOCK table hint. 
• B、 Modify the UPDATE statement in the OLE DB Command transform to use the TABLOCK table hint. 
• C、 Stage the data in the data flow. Replace the OLE DB Command transform in the data flow with an Execute SQL task in the control flow.
• D、 Stage the data in the data flow. Replace the UPDATE statement in the OLE DB Command transform with a DELETE statement followed by an INSERT statement.

第14题：

In which case would you use a FULL OUTER JOIN?（）

• A、Both tables have NULL values.
• B、You want all unmatched data from one table.
• C、You want all matched data from both tables.
• D、You want all unmatched data from both tables.
• E、One of the tables has more data than the other.
• F、You want all matched and unmatched data from only one table.

第15题：

Which two statements are correct regarding the Flashback Versions Query feature？（）

• A、 You can use this feature to identify the committed versions of the rows， but not the uncommitted versions.
• B、 You can use this feature to identify the versions of V$ views.
• C、 You can use this feature to identify the inserts， deletes， and updates performed on a particular row but not the data definition language （DDL） operations performed on the table.
• D、 You can use this feature for a table only if row movement is enabled for the table.
• E、 You can use this feature to identify the versions of external and fixed tables.

第16题：

In which scenario will you create a hash cluster？（）

• A、 when you want the data to be returned automatically in chronological order
• B、 if the application uses queries joining tables only occasionally
• C、 if the queries against the clustered table use the equality operator （=） to retrieve the desired row
• D、 if the full table scan is executed often on only one of the clustered tables

第17题：

An application uses Windows Azure Table storage. The application uses five tables.  One table used by the application is approaching the limit for storage requests per second.  You need to recommend an approach for avoiding data access throttling.  What should you recommend？（）

• A、 Use a single partition key for the table.
• B、 Compress data before storing it in the table.
• C、 Create additional partition keys for the table.
• D、 Continually remove unnecessary data from the table.

第18题：

第19题：

第20题：

第21题：

第22题：

第23题：

第24题：