Skip to content

UC Data Warehouse Testing

TEST CASE DEVELOPMENT

The development of all test cases utilized for the different levels of testing will be a collaborative effort between IRAP and Data Services. It is expected that test cases will be based on file specification elements as laid out in business requirements and on all transformed fields. All test cases will be clearly defined in Atlassian JIRA by content area.

TESTING CATEGORIES

  • Unit Testing

It is important that developers perform their own unit testing so that problems that should have been identified early are not transferred to the business user as part of user acceptance testing. Unit testing will include functional and regression testing (if applicable). Through proper unit testing, each unit of the implementation can be isolated and tested to ensure that the individual parts are correct. Each unit must be tested against the business rules provided by IRAP to ensure that the developed logic satisfies the business requirements. For example, if an attribute such as residency has been specified in requirements to have three categories, the unit of the program responsible for fulfilling this requirement must be tested by the developer to ensure that this is the case. Unit testing is purely a responsibility of Data Services as it guarantees that the programs and processes being utilized are working as expected. The test cases utilized for unit testing must be based on the business requirements and file mapping specifications. All fields that have been transformed must be unit tested to ensure that expected outcomes for the fields have been met. IRAP will not participate in the unit testing process.

  • Performance Testing

Conducting performance testing on deliverables earmarked for production is necessary to ensure that any performance issues are identified before the deployment of data and reports to business users. Data Services must be able to preempt how the data warehouse will behave when multiple users hit it simultaneously.  Response times for executing pre-built reports in legacy systems must be compared with that of target systems to ensure that users will not experience degradation in performance. Data Services will be responsible for ensuring that data is organized well logically and that the database performs well. IRAP will work with Data Services in identifying and documenting test cases that will be included in the performance testing efforts. Performance testing will be performed by Data Services and results of these tests must be presented to IRAP.

  • Quality Assurance (QA) Testing

By implementing the necessary quality assurance processes, Data Services can ensure that mistakes and the resulting data quality issues are not transferred to IRAP. Proper quality assurance will help confirm that requirements are fulfilled, data being delivered is ‘fit for purpose’ and accurate the first time. IRAP will work with Data Services to help establish the test cases that will be utilized during the QA process.  Quality assurance is a function for Data Services. IRAP will not participate in the QA testing process. By implementing a reliable quality assurance system, both parties will experience an increase in client confidence and UCDW data credibility, to improve work processes and efficiency, and to enable better decision-making and the ability for UC to compete with other comparable higher education institutions.

  • User Acceptance Testing

User acceptance testing comes after the unit testing and QA processes have been completed. This is final testing process prior to the release of data for business reporting and is performed by IRAP who ultimately signs off on the deliverable. Test cases that will be employed as part of UAT have been partially implemented in the UAT Validation Framework and work continues on the framework as requirements evolve. UAT is a responsibility of the IRAP team. There is the expectation that errors that should have been detected during unit testing and QA will have been identified and addressed by the Data Services team prior to the commencement of the QA process.

  • Star Schema and Data Mart Testing

Once IRAP has certified the data prior to and after the load into the star-schema, the same data will not be tested again once the subsequent load into the data mart is complete. It is expected that the data within the data mart will mirror what is contained within the star-schema. However, if there are fields that are uniquely transformed within the data mart alone (fields that do not exist in the star-schema), IRAP will produce test cases to validate such fields after data marts are loaded.

For any questions about our data warehouse testing processes, please contact irap@ucop.edu.