A test case might be created either manually or automated to verify the functionality per the original acceptance criteria.
High Level Steps
Pre-Requisites
- Check connection details of source and target systems
- User access to various tools and access rights etc.
Checks
- Check the ETL script name(s) are aligning with naming conventions
- Tables/Files are present in landing zone and tables are not empty
- Take the count of entries of each table.
- Check field mappings as per the data dictionary
Data
- Reconciliation of number of records in a file/table
- Comparing values of all fields for a random sample rows (Source and target)
Steps
- Check the status of previous ETL job
- If previous job ran successfully and data is available then run the ETL script
- Job should be successful
- Number of entries at target should match with source (data in landing zone)
- Check the values of random sample rows of the table at target with source – values should match
IMPORTANT POINTS / QUESTIONS:
- Negative testing
- Appropriate comments are present in ETL script
- What will happen if the ETL job failed during run?
- Any error code is expected?
- What is business expectations on these occasions?
- When the truncate step is successful and the subsequent insert step fails, what will happen?