Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Data Driven Testing Is More Than an Excel File

Digital & Social Media Marketing Summit İstanbul 2019 Event
Mehmet Gok

  • Be the first to comment

  • Be the first to like this

Data Driven Testing Is More Than an Excel File

  1. 1. Data Driven Testing Is More Than an Excel File İstanbul November 22, 2019 Test Automation & Digital QA Event Mehmet Gök #TAS19
  2. 2. About Me Bs. Indsutrial Engineer Ms. Computer Engineer Telco Industry Project Manager Technical Analyst Test Automation Test Lead Mehmet Gök
  3. 3. Agenda Frameworks Approaches Test Data Management Concepts Service Virtualization ApplicationTools
  4. 4. Framework selection depends on the context, the environment and even the team. Frameworks
  5. 5. Automation (or Manual?) Frameworks A scripting technique that uses data files to contain the test data and expected results needed to execute the test scripts. * Data Driven Testing * ISTQB gloassary A scripting technique in which test scripts contain high-level keywords and supporting files that contain low-level scripts that implement those keywords.* Keyword Driven Testing • Modular Testing • Hybrid Testing • Behavior Driven Development Other Frameworks
  6. 6. The maintenance of the test data becomes more difficultday by day with the complexity of your software under test increases and it becomes scary. Test Data Management Concepts
  7. 7. Test Data Management Concepts Subsetting It is the process of moving a data set from one data source to another, with all its relationships and dependencies. Synthetic Data Generation It is the process of producing nonsense data that does not represent a real person, institution or any other confidential data. Data Integrity A term describing that data is complete, reliable, referentially integrated, valid and usable. Masking To create a derivative that is structurally and formally similar to the actual data, but which cannot be estimated.
  8. 8. Data Integrity Duplicate records may result false positive bugs. Redundancy All those verifications are for accurate and usable test data. Accuracy & Usability Domain related data must remain same to protect the integrity of meaning. Domain Integrity Resource: Data Integrity Means Everything by Mike Butrym at Missing data may result false positive bugs. Completeness Orphan records may result false positive bugs. Referential Integrity Invalid data may result false positive bugs. Validity
  9. 9. The success of test automation begins with choosing the most appropriate test data management approach to your application and data. Approaches
  10. 10. If a subset is to be copied from one source to another, the data model must be fully constructed and given to the test data management tool. Data Modelling Data discovery helps to raise awareness about data and eliminate data-driven errors. Data Discovery The reduction of errors and costs caused by the test data is avoidable by appropriately preparing the test data. Profile Test Data Subsetting It is very likely that personal information will be transferred to the test environment as a result of subsetting. Masking
  11. 11. Static Data Masking Static data masking is the masking process at the database layer when creating a database copy or a subset of production data.
  12. 12. Dynamic Data Masking Dynamic data masking is the process of masking by putting a proxy between the database and the application layer which works on network layer or database layer.
  13. 13. Dynamic Data Masking With Postgres
  14. 14. Synthetic Data Generation Database level generation - Data structure should be known very well in order to generate data synthetically. - Synthetic data generation engines works with rule-sets. Application level generation - No need for data modelling & discovery - Service virtualization techniques needed on 3rd party integrations. - Synthetic data generation software (Probably API) should be designed Data Modelling Data Discovery Rule Set Design
  15. 15. Service virtualization is a must of independent and unlimited testing and synthetic data generation. Service Virtualization
  16. 16. Stubs, Hubs and Mocking vs Service Virtualization Stubs Hubs/Drivers Mocking
  17. 17. Tools
  18. 18. Tools
  19. 19. Tool Selection Criterias Learning Curve Is it simple enough to use immediately? Regulations Does it meet regulations like GDPR? Speed Will it keep up with your regression running frequency? Cost Is it worth the cost compared to its benefit? Decentralization Can anyone create test data within their authority wherever they want? Interoperability Does it work in your environment?
  20. 20. Application
  21. 21. On-the-fly Test Data Generation – A Simple Case Create an individual customer Test data need: Create a customer who is older than 18, is a Turkish citizen and is a man TR Identity Number Generator An Identity number generator method is needed to get a TR Identity Number that will return a person (>18, Male, Turkish Citizen) Create Customer Method You can use the same method that your application uses to create customers so you can create test data. Identity Share System (KPS-NVI) In order to gather a synthetically generated customer KPS service should be virtualized
  22. 22. Thanks! /mehmetg To get in touch;

    Be the first to comment

Digital & Social Media Marketing Summit İstanbul 2019 Event Mehmet Gok


Total views


On Slideshare


From embeds


Number of embeds