Teradata Teradata Architecture 2 — Questions and Answers
Question 1: What is a Primary Index in Teradata and why is it important?
- It determines data distribution across AMPs and is critical for query performance (Correct answer)
- It enforces uniqueness like a primary key only
- It creates a B-tree index for fast lookups
- It defines the physical sort order of the table
Correct answer: It determines data distribution across AMPs and is critical for query performance
The Primary Index determines how rows are distributed across AMPs; a well-chosen PI ensures even distribution and efficient row retrieval.
Question 2: What is the difference between a Unique Primary Index (UPI) and a Non-Unique Primary Index (NUPI)?
- UPI guarantees one row per AMP bucket; NUPI allows multiple rows to hash to the same AMP (Correct answer)
- UPI uses B-tree storage; NUPI uses hash storage
- UPI allows NULLs; NUPI does not
- UPI is defined at query time; NUPI is defined at table creation
Correct answer: UPI guarantees one row per AMP bucket; NUPI allows multiple rows to hash to the same AMP
A UPI ensures no two rows share the same hash value, guaranteeing a single row per hash bucket, while NUPI allows duplicates on the same AMP.
Question 3: In Teradata, what is a Secondary Index used for?
- Provides an alternate access path to data without redistributing rows (Correct answer)
- Replaces the Primary Index
- Stores data sorted by a non-PI column
- Partitions rows across different AMPs
Correct answer: Provides an alternate access path to data without redistributing rows
A Secondary Index creates an alternate access path so queries not using the PI can still access rows efficiently without a full-table scan.
Question 4: What is the purpose of Fallback in Teradata?
- Stores a second copy of each row on a different AMP for fault tolerance (Correct answer)
- Provides a rollback mechanism for failed transactions
- Backs up data to an external storage system
- Mirrors the entire Teradata system to a standby cluster
Correct answer: Stores a second copy of each row on a different AMP for fault tolerance
Fallback protection keeps a duplicate copy of every row on a different AMP, ensuring data availability if one AMP fails.
Question 5: What is a node in Teradata's hardware architecture?
- A physical server containing CPUs, memory, and disk that hosts multiple Vprocs (Correct answer)
- A single AMP process
- A network switch connecting AMPs
- A disk array shared across multiple servers
Correct answer: A physical server containing CPUs, memory, and disk that hosts multiple Vprocs
A node is a physical server with its own CPUs, memory, and I/O that hosts multiple Parsing Engines and AMPs (Vprocs).
Question 6: What does the Teradata Data Dictionary contain?
- Metadata about all database objects such as tables, views, users, and indexes (Correct answer)
- The actual data rows for all user tables
- Query execution plans cached for reuse
- Log files for all DML operations
Correct answer: Metadata about all database objects such as tables, views, users, and indexes
The Data Dictionary stores system-level metadata including definitions of tables, views, indexes, stored procedures, and user access rights.
Teradata Teradata Architecture 2
What is a Primary Index in Teradata and why is it important?