Year End Offer - Save up to 50% on tickets - Book Now

EuroSTAR Conference

Europe's Largest Quality Engineering Conference

Big Data

Are You Aware of the Hidden Cost of Test Data Governance?

by

What Does Lack of Test Data Governance Mean for Your Testing?

Inefficiency in test data management is one of the main causes of delays in companies that regularly conduct software testing activities. In these companies, highly skilled teams spend up to half of their workday on routine tasks such as creating or waiting for data, which not only significantly decreases productivity, but also impacts the quality of the software developed and the motivation of the team.

Software development and testing teams often resort to inefficient strategies that fail to provide the data that tests need. Among these, we can highlight spreadsheets, Business APIs, Reserved Data, and Automation Scripts.

Upon examining these alternatives, we can identify common problems among them, such as creating inconsistent, non-reusable data, simple data structures, etc. This leads to tests with unreliable results, which reduce test coverage and thus limit their effectiveness.

Failing to quantify these issues prevents organizations from realizing the true extent of the consequences and hinders their ability to implement effective Test Data Management strategies to reverse the situation. Recognizing and addressing the neglect of test data governance is, therefore, essential for optimizing resources, improving development cycles, and ensuring market competitiveness.

Digital Transformation in Test Data Management

Digital transformation in the realm of test data management marks a before and after in software quality. This allows you to have good test coverage and effective data management, which is fundamental for software quality.

The lack of software quality is directly related to the quality of test data. Therefore, a digital transformation necessarily encompasses not only processes and tools, but also the data management culture within the organization. By integrating advanced digital technologies for the creation, management, and deployment of test data, companies can significantly improve the accuracy of their tests, reducing data security and privacy challenges. Moreover, this transformation strengthens endusers’ trust in the digital solutions offered.

In this context, the integration of TDM tools not only increases efficiency and reduces Time To Market but also significantly improves the robustness and reliability of the software in complex production environments. Thus, automating test data management is the foundation for facing new technological challenges.

icaria TDM: Transforming Test Data Management

The constant evolutionary changes in technology have made efficiency and innovation fundamental in any technological process. icaria Technology positions itself as a leading vendor in Test Data Management soluctions with icaria TDM. Among its main features, we highlight the adaptation and extension to customer needs while staying within the philosophy of product-oriented design, scalability, and easy evolution.

icaria TDM is more than a tool, as it is based on a proprietary methodology; it is a robust solution that transforms the way testing teams access and use data. It provides accurate and secure data just when testers need it and perfect for their tests, radically transforming testing. By reducing waiting times and saving on storage costs and resources, icaria TDM not only optimizes operations but also significantly improves software quality.

What Features Should a Good TDM Tool Have?

  • Data Masking: Ensure compliance with European GDPR or similar regulations by protecting sensitive information.
  • Automatic Identification of PII: Locate sensitive data in databases, ensuring that no critical information is exposed.
  • Pseudonymization and Subsetting: Mask and move data subsets efficiently, maintaining coherence and security.
  • Self-Service and Synthetic Data: Facilitate on-demand data delivery and the generation of complex structures, streamlining the testing process.
  • Integration with Third Parties: Ensure perfect synchronization with external tools, improving flexibility and adaptability. Integrating into CI/CD processes.

The Impact of icaria TDM on Testing Teams

Testing teams that adopt icaria TDM experience a notable improvement in their daily work.

“Generating a test data used to take us about 3 days and was done manually… We have reduced the test data generation time by more than 70%.”

An extract from the Orange SQA team’s presentation during the ExpoQA’22 conference. The elimination of bottlenecks and the immediate availability of data not only accelerates Time to Market but also increases team satisfaction. With automated and predictable processes, icaria TDM improves SLAs and transforms data management challenges into a competitive advantage.

Where is icaria Technology Heading?

icaria Technology aims to continue being at the forefront of technological innovation. Participation in European events such as Eurostar is an opportunity to share knowledge and vision about TDM and also to learn from other experts in the field. With data tools like icaria TDM or icaria GDPR, icaria Technology aims not only to improve testing practices, but to redefine what software development is.

Author

Maarten Urbach

Enrique Almohalla, leading icaria Technology as CEO, brings a wealth of experience in TDM methodologies, cultivated through over twenty years of directing software development, deployment, and testing projects. His significant involvement in Test Data Management, marked by continuous innovation and application, underscores his deep understanding of the field. Additionally, his position as an Associate Professor at IE Business School in Operations and Technology melds his hands-on experience with academic insights, offering a comprehensive perspective on business management.

iCaria is an exhibitor at EuroSTAR 2024, join us in Stockholm.

Database Subsetting

by

The biggest technical challenge when using production data for development and testing is that the volume of data can be huge. Subsetting the data while keeping its semantic meaning is not an easy task.

In a practical sense relating to IT, subsetting is a process of making a large database smaller in order to ease the development and testing. Without it, teams would either have to works directly on production databases or each have his own copy of it, which when talking about enterprise solutions, could be well over the terabyte mark.

Ekobit's graphic illustration of subsetting a large database into several small ones
Figure 1 – Subsetting a large database into several small ones

Challenges of subsetting

With subsetting, each team or even team member can have his own database with the data that is relevant to the task. But because of the new process, a lot of technical skill are required, most of which are not needed and shouldn’t be required from a tester or, for example, a UX designer:

  • Knowledge of the SQL syntax
  • In depth knowledge of the database schema and relationships
  • The domain knowledge of the application, and what tables are needed for it to function

Because of all this, scarce resources have to be consumed, mostly database administrators and developers time. And it’s not a one-off process. Every time the database definition changes, the process must be updated, and the new subsets redistributed to users. Now add anonymization and pseudo anonymization because of the GDPR into the mix, and the task becomes extremely tedious.

What is a subset anyway?

We must decide what data we need, or to go down to the database level, what tables are the pivot for our subset building. Some of the possible strategies are:

  • First, top, bottom or random N records
  • Filter by columns of a table
  • List of primary keys to extract
  • Complex queries spanning over multiple tables

After that, we need to include all the related parent table data to the records we extracted so that when we bring the constraints in, the database is valid. We might also want some child records of the tables, with or without filters and then we must gather the remaining parents.

Ekobit's graphic illustration of parent-child relationships between tables
Figure 2 – Parent-child relationships between tables

Figure 2 – Parent child relationships between tables

 

Data explosion

There will probably be tables that have circular references back to itself, either directly or via some sequence of tables. Here we can easily experience data explosion, if we run into a record that is connected to many other records, and we end up with a useless subset, that is almost as big as the original one.

Here we need to identify such records and tables, and either copy them entirely, which is usable for smaller tables like organization hierarchy, but for a very active table, like transactions in a banking system we need a cutoff. Some of the basic cutoff strategies available are:

  • Don’t collect the data, we will deal with the constraint later, either by not collecting the records, or putting the appropriate key to null
  • Link all the records to the same parent record, be it a random, fixed or syntactic record

Ekobit's graphic illustration of circular references between tables
Figure 3 – Circular references between tables

Performance

There are many factors that can affect the performance, but the one we can influence is the number of queries to the database. We can make the queries smarter, try to minimize the number of visits of the same table, and perform micro optimizations by deciding how to transfer the data from one system to the other.

We can offload a lot of work to the machine running the subsetting, and have it log the Ids of row to be transferred, but then we run into a risk of needing a lot of memory. Or we can have the database engines run the queries for us utilizing linked servers, but then we can easily overextend the production database. It’s not an easy decision to make and each system is different.

Conclusion

Subsetting sounds trivial when first mentioned, but as you get deeper into it, it becomes more complex. As soon as you start working with real life database, a great number of challenges appear which are not easily solved. Considering that the users want an application that works and is thoroughly tested, subsetting should be an integral part of quality control, testing and DevOps.

If you’ve encountering the subsetting Gordian knot, and need help, or are just overwhelmed with the many details, come and find us at the Ekobit stand, and see how we, our software and our expertise in both data masking and subsetting can help you.