Search

Due to its importance and widespread use in industry, automated testing of REST APIs has attracted major interest from the research community in the last few years. However, most of the work in the literature has been focused on black-box fuzzing. Although existing fuzzers have been used to automatically find many faults in existing APIs, there are still several open research challenges that hinder the achievement of better results (e.g., in terms of code coverage and fault finding). For example, under-specified schemas are a major issue for black-box fuzzers. Currently, EvoMaster is the only existing tool that supports white-box fuzzing of REST APIs. In this paper, we provide a series of novel white-box heuristics, including for example how to deal with under-specified constrains in API schemas, as well as under-specified schemas in SQL databases. Our novel techniques are implemented as an extension to our open-source, search-based fuzzer EvoMaster. An empirical study on 14 APIs from the EMB corpus, plus one industrial API, shows clear improvements of the results in some of these APIs.

In industry, RESTful APIs are widely used to build modern Cloud Applications. Testing them is challenging, because not only they rely on network communications, but also they deal with external services like databases. Therefore, there has been a large amount of research sprout in recent years on how to automatically verify this kind of web services. In this paper, we review the current state-of-the-art on testing RESTful APIs, based on an analysis of 92 scientific articles. This review categories and summarizes the existing scientific work on this topic, and discusses the current challenges in the verification of RESTful APIs.

The Graph Query Language (GraphQL) is a powerful language for APIs manipulation in web services. It has been recently introduced as an alternative solution for addressing the limitations of RESTful APIs. This paper introduces an automated solution for GraphQL APIs testing. We present a full framework for automated APIs testing, from the schema extraction to test case generation. In addition, we consider two kinds of testing: white-box and black-box testing. The white-box testing is performed when the source code of the GraphQL API is available. Our approach is based on evolutionary search. Test cases are evolved to intelligently explore the solution space while maximizing code coverage and fault-finding criteria. The black-box testing does not require access to the source code of the GraphQL API. It is therefore of more general applicability, albeit it has worse performance. In this context, we use a random search to generate GraphQL data. The proposed framework is implemented and integrated into the open-source EvoMaster tool. With enabled white-box heuristics, i.e., white-box mode, experiments on 7 open-source GraphQL APIs show statistically significant improvement of the evolutionary approach compared to the baseline random search. In addition, experiments on 31 online GraphQL APIs reveal the ability of the black-box mode to detect real faults.

Remote Procedure Call (RPC) is a communication protocol to support client-server interactions among services over a network. RPC is widely applied in industry for building large-scale distributed systems, such as Microservices. Modern RPC frameworks include for example Thrift, gRPC, SOFARPC and Dubbo. Testing such systems is very challenging, due to the complexity of distributed systems and various RPC frameworks the system could employ. To the best of our knowledge, there does not exist any tool or solution that could enable automated testing of modern RPC-based services. To fill this gap, in this paper we propose the first approach in the literature, together with an open-source tool, for white-box fuzzing modern RPC-based APIs with search. To assess our novel approach, we conducted an empirical study with two artificial and four industrial APIs selected by our industrial partner. The tool has been integrated into a real industrial pipeline, and could be applied to real industrial development process for fuzzing RPC-based APIs. To further demonstrate its effectiveness and application in industrial settings, we also report results of employing our tool for fuzzing another 50 industrial APIs autonomously conducted by our industrial partner in their testing processes. Results show that our novel approach is capable of enabling automated test case generation for industrial RPC-based APIs (i.e., two artificial and 54 industrial). We also compared with a simple grey-box technique and existing manually written tests. Our white-box solution achieves significant improvements on code coverage. Regarding fault detection, by conducting a careful review with our industrial partner of the tests generated by our novel approach in the selected four industrial APIs, a total of 41 real faults were identified, which have now been fixed. Another 8,377 detected faults are currently under investigation.

With several microservice architectures comprising of thousands of web services, used to serve 630 million customers, companies like Meituan face several challenges in the verification and validation of their software. This paper reports on our experience of integrating EvoMaster (a search-based white-box fuzzer) in the testing processes at Meituan over almost 2 years. Two user studies were carried out in 2021 and in 2023 to evaluate two versions of EvoMaster, respectively, in tackling the test generation for industrial web services which are parts of a large e-commerce microservice system. The two user studies involve in total 321,131 lines of code from five APIs and 27 industrial participants at Meituan. Questionnaires and interviews were carried out in both user studies with employees at Meituan. The two user studies demonstrate clear advantages of EvoMaster (i.e., code coverage and fault detection) and the urgent need to have such a fuzzer in industrial microservices testing. To study how these results could generalize, a follow up user study was done in 2024 with five engineers in the five different companies. Our results show that, besides their clear usefulness, there are still many critical challenges that the research community needs to investigate to improve performance further.

RESTful APIs are a type of web services that are widely used in industry. In the last few years, a lot of effort in the research community has been spent in designing novel techniques to automatically fuzz those APIs to find faults in them. Many real faults were automatically found in a large variety of RESTful APIs. However, usually the analyzed fuzzers treat the APIs as black-box, and no analysis of what is actually covered in these systems is done. Therefore, although these fuzzers are clearly useful for practitioners, we do not know what are their current limitations and actual effectiveness. Solving this is a necessary step to be able to design better, more efficient and effective techniques. To address this issue, in this paper we compare seven state-of-the-art fuzzers on 18 open-source and one industrial RESTful APIs. We then analyzed the source code of which parts of these APIs the fuzzers fail to generate tests for. This analysis points to clear limitations of these current fuzzers, listing concrete challenges for the research community to follow up on.

In the context of a large software testing company, we have deployed the model-based testing (MBT) approach to take the company's test automation practices to higher levels of maturity /and capability. We have chosen, from a set of open-source/commercial MBT tools, an open-source tool named GraphWalker, and have pragmatically used MBT for end-to-end test automation of several large web and mobile applications under test. The MBT approach has provided, so far in our project, various tangible and intangible benefits in terms of improved test coverage (number of paths tested), improved test-design practices, and also improved real-fault detection effectiveness. The goal of this experience report (applied research report), done based on "action research", is to share our experience of applying and evaluating MBT as a software technology (technique and tool) in a real industrial setting. We aim at contributing to the body of empirical evidence in industrial application of MBT by sharing our industry-academia project on applying MBT in practice, the insights that we have gained, and the challenges and questions that we have faced and tackled so far. We discuss an overview of the industrial setting, provide motivation, explain the events leading to the outcomes, discuss the challenges faced, summarize the outcomes, and conclude with lessons learned, take-away messages, and practical advices based on the described experience. By learning from the best practices in this paper, other test engineers could conduct more mature MBT in their test projects.

Context: With the rising complexity and scale of software systems, there is an ever-increasing demand for sophisticated and cost-effective software testing. To meet such a demand, there is a need for a highly-skilled software testing work-force (test engineers) in the industry. To address that need, many university educators worldwide have included software-testing education in their software engineering (SE) or computer science (CS) programs. Objective: Our objective in this paper is to summarize the body of experience and knowledge in the area of software-testing education to benefit the readers (both educators and researchers) in designing and delivering software testing courses in university settings, and to also conduct further education research in this area. Method: To address the above need, we conducted a systematic literature mapping (SLM) to synthesize what the community of educators have published on this topic. After compiling a candidate pool of 307 papers, and applying a set of inclusion/exclusion criteria, our final pool included 204 papers published between 1992 and 2019. Results: The topic of software-testing education is becoming more active, as we can see by the increasing number of papers. Many pedagogical approaches (how to best teach testing), course-ware, and specific tools for testing education have been proposed. Many challenges in testing education and insights on how to overcome those challenges have been proposed. Conclusion: This paper provides educators and researchers with a classification of existing studies within software-testing education. We further synthesize challenges and insights reported when teaching software testing. The paper also provides a reference ("index") to the vast body of knowledge and experience on teaching software testing.

This paper presents EvoMaster, an open-source tool that is able to automatically generate system level test cases using evolutionary algorithms. Currently, EvoMaster targets RESTful web services running on JVM technology, and has been used to find several faults in existing open-source projects. We discuss some of the architectural decisions made for its implementation, and future work., Comment: arXiv admin note: text overlap with arXiv:1901.01538

What is the impact of software engineering research on current practices in industry? In this paper, I report on my direct experience as a PhD/post-doc working in software engineering research projects, and then spending the following five years as an engineer in two different companies (the first one being the same I worked in collaboration with during my post-doc). Given a background in software engineering research, what cutting-edge techniques and tools from academia did I use in my daily work when developing and testing the systems of these companies? Regarding validation and verification (my main area of research), the answer is rather short: as far as I can tell, only FindBugs. In this paper, I report on why this was the case, and discuss all the challenging, complex open problems we face in industry and which somehow are "neglected" in the academic circles. In particular, I will first discuss what actual tools I could use in my daily work, such as JaCoCo and Selenium. Then, I will discuss the main open problems I faced, particularly related to environment simulators, unit and web testing. After that, popular topics in academia are presented, such as UML, regression and mutation testing. Their lack of impact on the type of projects I worked on in industry is then discussed. Finally, from this industrial experience, I provide my opinions about how this situation can be improved, in particular related to how academics are evaluated, and advocate for a greater involvement into open-source projects.

Nowadays, web services play a major role in the development of enterprise applications. Many such applications are now developed using a service-oriented architecture (SOA), where microservices is one of its most popular kind. A RESTful web service will provide data via an API over the network using HTTP, possibly interacting with databases and other web services. Testing a RESTful API poses challenges, as inputs/outputs are sequences of HTTP requests/responses to a remote server. Many approaches in the literature do black-box testing, as the tested API is a remote service whose code is not available. In this paper, we consider testing from the point of view of the developers, which do have full access to the code that they are writing. Therefore, we propose a fully automated white-box testing approach, where test cases are automatically generated using an evolutionary algorithm. Tests are rewarded based on code coverage and fault finding metrics. We implemented our technique in a tool called EVOMASTER, which is open-source. Experiments on two open-source, yet non-trivial RESTful services and an industrial one, do show that our novel technique did automatically find 38 real bugs in those applications. However, obtained code coverage is lower than the one achieved by the manually written test suites already existing in those services. Research directions on how to further improve such approach are therefore discussed.

Automatically generating test suites is intrinsically a multi-objective problem, as any of the testing targets (e.g, statements to execute or mutants to kill) is an objective on its own. Test suite generation has peculiarities that are quite different from other more regular optimisation problems. For example, given an existing test suite, one can add more tests to cover the remaining objectives. One would like the smallest number of small tests to cover as many objectives as possible, but that is a secondary goal compared to covering those targets in the first place. Furthermore, the amount of objectives in software testing can quickly become unmanageable, in the order of (tens/hundreds of) thousands, especially for system testing of industrial size systems. Traditional multi-objective optimisation algorithms can already start to struggle with just four or five objectives to optimize. To overcome these issues, different techniques have been proposed, like for example the Whole Test Suite (WTS) approach and the Many-Objective Sorting Algorithm (MOSA). However, those techniques might not scale well to very large numbers of objectives and limited search budgets (a typical case in system testing). In this paper, we propose a novel algorithm, called Many Independent Objective (MIO) algorithm. This algorithm is designed and tailored based on the specific properties of test suite generation. An empirical study, on a set of artificial and actual software, shows that the MIO algorithm can achieve higher coverage compared to WTS and MOSA, as it can better exploit the peculiarities of test suite generation.