ETL Testing ETL Data Extraction Techniques 1 — Questions and Answers
Question 1: Which extraction method reads only the data that has changed since the last extraction, minimizing load on source systems?
- Full extraction
- Incremental extraction (Correct answer)
- Logical extraction
- Online extraction
Correct answer: Incremental extraction
Incremental extraction captures only changed or new records since the last ETL run, reducing processing time and source system load.
Question 2: What is a 'delta' in the context of ETL data extraction?
- A Greek letter used in database naming
- The difference or change in data since the last extraction (Correct answer)
- A full copy of the source table
- A staging area database
Correct answer: The difference or change in data since the last extraction
A delta represents the subset of records that have been inserted, updated, or deleted since the previous extraction cycle.
Question 3: Which technique uses database log files to capture data changes for ETL extraction?
- Timestamp-based extraction
- Trigger-based extraction
- Log-based Change Data Capture (CDC) (Correct answer)
- Full table scan
Correct answer: Log-based Change Data Capture (CDC)
Log-based CDC reads the database transaction logs to capture inserts, updates, and deletes without adding overhead to the source database.
Question 4: When testing ETL extraction, which validation confirms that no source records were dropped during the pull?
- Data type check
- Row count reconciliation (Correct answer)
- Null value check
- Duplicate check
Correct answer: Row count reconciliation
Row count reconciliation compares the number of records extracted against the source to ensure completeness.
Question 5: What does 'online extraction' mean in ETL testing?
- Extracting data from cloud-only sources
- Extracting data directly from a live, active production system (Correct answer)
- Extracting data via REST API only
- Extracting data during scheduled downtime
Correct answer: Extracting data directly from a live, active production system
Online extraction pulls data from an active source system while it is still handling live transactions, requiring careful concurrency management.
Question 6: Which extraction type is MOST suitable when source systems do not support timestamps or change tracking?
- Incremental extraction
- Log-based CDC
- Full extraction (Correct answer)
- Streaming extraction
Correct answer: Full extraction
Full extraction is used when no mechanism exists to identify changed records, so the entire dataset is re-read each cycle.
ETL Testing ETL Data Extraction Techniques 1
Which extraction method reads only the data that has changed since the last extraction, minimizing load on source systems?