GenRocket

Fig 01 - GenRocket Project Dashboard

Synthetic data in GenRocket is generated using a model that consists of “domains”, “attributes”, “generators”, “receivers” and “scenarios”. Domains contain attributes, which are roughly equivalent to tables and columns respectively, and attributes are in turn equipped with one or more generators: methods for generating your synthetic data.

More than 650 generators are provided out of the box, many of which are customisable via built-in parameters. Notably, GenRocket provides several generators specifically to support machine learning via the creation of training data. It also has the ability to blend production data with synthetic data, as well as to generate synthetic data feeds. Generators can be linked together on an ad hoc basis, always create internally consistent data sets, and can even be assigned to your attributes automatically.

Each of your domains also has access to a variety of receivers, which determine the output format of your test data. Receivers exist for upwards of 70 different output formats, including such favourites as CSV, JSON, SQL, REST, SOAP and XML, and you can generate test data in several different formats simultaneously by attaching multiple receivers to a given domain.

Meanwhile, scenarios are sets of (configurable) instructions for generating your test data, created by combining domains, attributes, generators and receivers into a single specification. You can include multiple domains manually, or specify relationships that will allow GenRocket to bring in families of domains automatically. This guarantees referential integrity between those domains. You can also leverage rules-based validation to control the data (and thus the overarching structure of the data set) generated by your scenarios.

Scenarios can be grouped together into user stories, which can in turn be grouped into epics (a la Agile), all of which can be used (and reused) to repeatedly create test data via self-service, in real time, and on an essentially ad hoc basis. They can be executed centrally, facilitated by the GenRocket Multi User Server (GMUS) that allows large volumes of users to generate synthetic data simultaneously, but are also highly portable, and can be executed more or less anywhere (including on your local machine) via APIs or just by running the scenario file. Scenarios can even be modified in real time if necessary (perhaps as part of a dynamic testing workflow).

GenRocket also provides an XTS (Extract Table Schema) feature that will scan your database schema and automatically build out a test data model, including domains, attributes, generators and scenarios. Relationships between domains are established via a wizard, and you can proceed to either use the generated model as-is or customise it as you require. Similarly, GenRocket can automatically generate synthetic data for documents by leveraging a relevant XSD file.

Fig 02 - GenRocket architecture and use cases

GenRocket is an enterprise-ready, cost-effective point solution for test data management and automation that is well-equipped to handle use cases centered around synthetic data. It aims to be highly automated, to integrate with your existing pipelines, and to work almost behind the scenes to supply you with appropriate test data at the right place and at the right time. This is a laudable aim, and for the most part it achieves it.

The product offers a highly distributable, reusable, lightweight and portable system for generating synthetic data that is easily accessible and available wherever and whenever you need it. Compared to more traditional, monolithic test data management offerings, it is frequently leaner, easier to use, and – at least in terms of synthetic data – more sophisticated.

GenRocket is unique in that it takes an entirely synthetic approach to test data that aims to remove production data from your testing lifecycle completely. Under this paradigm, your synthetic data is not derived from your production data, but instead directly from your business requirements. This is no bad thing, since it means that your test data will be precisely, and reliably, what you need it to be at any given time: you do not need to fall victim to the whims of your production data. In short, GenRocket can leverage production data – for example, it can blend it into your synthetic data – but it does not require it. This makes GenRocket an ideal solution if you want to divest your testing processes of production data and instead rely 100% on the synthetic.

GenRocket does lack some functionality common to its competitors – data profiling, for instance – and its data subsetting and masking functions are fine yet hardly exceptional. But at the end of the day, these capabilities are largely irrelevant to GenRocket’s preferred mode of use. That said, if you want access to them regardless, GenRocket works excellently to augment other testing platforms with synthetic data: it’s lightweight, cost-effective, and offers a range of integration capabilities. Its recent partnership with Delphix is particularly noteworthy in this regard.

The Bottom Line

GenRocket is a fit-for-purpose test data platform that enables you to generate sophisticated synthetic data anytime and anywhere. It should be especially appealing if you want to remove production data from your testing process entirely.