CDW 320 Clarity for Caboodle

Question 1

True or False: You should always create new DMCs when bringing in data into Caboodle.

Answer 1

False. If you are only adding a few columns into Caboodle and there exists a DMC with a similar
granularity, it may be better to extend an Epic released DMC as opposed to making your own.

Question 2

True or False: Because Caboodle is a data warehouse, all data should be brought into Caboodle.

Answer 2

False: While Caboodle can store data from many different sources, some data should intentionally
not be brought into Caboodle to make sure the database stays healthy and organized.

Question 3

What extraction modes are used by Clarity packages?

Answer 3

Backfill and Incremental, and Backfill Only.

Question 4

What is the main difference between the WAREHOUSE_CHANGES2# and
WAREHOUSE_UNIQUE2# tables?

Answer 4

Since a single base ID can be flagged for incremental extract due to changes in multiple Clarity tables, WAREHOUSE_CHANGES2# tables may contain duplicate rows. These duplicate rows have been removed in the WAREHOUSE_UNIQUE2# table. Any base IDs that are also in the current backfill range will also have been removed.

Question 5

Explain the role of the DCIQ queries.

Answer 5

DCIQ queries convert the granularity of the temp tables into the BaseId of the load package.
Without the DCIQ query, the load package wouldn’t know which rows to update.

Question 6

What different SQL queries are necessary for a Clarity load package?

Answer 6

DMCs with a backfill only extraction method will just have a backfill query. The backfill query will
contain «LowerBound» and «UpperBound» tags to identify the records to load during backfill.

DMCs with an incremental extraction method will also have an incremental query, which uses a «WarehouseUniqueChangesTableName» tag to filter out the records that have not been updated in the last run of ETL. The incremental query knows which records to filter out thanks to the data changes insert queries (DCIQs). One DCIQ is necessary for each non‐ZC table in the incremental query.

Question 7

Suppose you’re creating a package that loads a type of record from Clarity that has never been
extracted to Caboodle. Which of the following are potentially appropriate choices for the package’s Allowed IdType?

a.﴿ EpicNewRecordId
b.﴿ EpicCustomRecordId
c.﴿ NewRecordId
d.﴿ CustomNewRecordId

Answer 7

c.) NewRecordId
d.) CustomNewRecordId

Question 8

Explain when the data lineage for StringBaseId should be left empty.

Answer 8

When filling out data lineage for Clarity load packages, you will either fill out StringBaseId or
NumericBaseId but not both. The one that gets filled out is determined by the primary base ID data type.

Question 9

Which of the following tags can appear in Backfill Extract Queries?
a.﴿ «LowerBound»
b.﴿ «UpperBound»
c.﴿ «EXECUTE»
d.﴿ «TempTable»
e.﴿ «WarehouseUniqueChangesTableName»

Answer 9

a.) «LowerBound»
b.) «UpperBound»
c.) «EXECUTE»

Question 10

True or False: Validation executions are a special type of execution that automatically fixes
inconsistencies between Clarity and Caboodle.

Answer 10

True

Question 11

How many columns are present in a Bridge table in the reporting database?
1
2
3
4

Answer 11

2 columns

Question 12

Which of the following types of lookup columns are available in data mart DMCs?
a.﴿ Key
b.﴿ DurableKey
c.﴿ ID
d.﴿ Durable ID

Answer 12

a.) Key
b.) DurableKey
c.) ID
d.) Durable ID