SPL Conqueror: Toward optimization of non-functional properties in software product lines

Abstract. We introduce by means of an example a modular verification technique for analyzing the behavior of software product lines using the mCRL2 toolset. Based on feature-driven borders, we divide a behavioral model of a product line into a set of separate components with interfaces and a driver process to coordinate them. Abstracting from irrelevant components, we verify properties over a smaller behavioral model, which not only simplifies the model checking task but also makes the result amenable for reuse. This is a fundamental step forward for the approach to scale up to industrial-size product lines.

show abstract

“…However, more general and complex constraints, properties and objectives can be treated as well [8,37]. The feature diagram, i.e.…”

Section: A Product Line As Running Examplementioning

confidence: 99%

Towards Modular Verification of Software Product Lines with mCRL2

Beek

Vink

2014

Leveraging Applications of Formal Methods, Verification and Validation. Technologies for Mastering Change

View full text Add to dashboard Cite

show abstract

“…The impact each feature has on a quality attribute can be obtained through systematic measurement of different variants (e.g. [32,13]). For example, in Figure 1, the impact of feature PasswordProtection on memory consumption is quantified as 20.…”

Section: Attributed Feature Modelsmentioning

confidence: 99%

“…In most of the sample models, we compute the quality of a product variant by summing up the contributions of each feature present in the variant. Some of the models also involve more complex computations, including multiplicative terms (e.g., for computing reliability), and also terms that represent feature interactions [32]. The objective functions are then to either minimize or maximize the quality attribute of a product variant.…”

Section: Attributed Feature Modelsmentioning

confidence: 99%

“…These models cover a range of sizes, from a small one such as Berkeley DB, to large one with hundreds of features such as Eshop, or many objectives, such as ERS. Berkeley DB [32] describes the price, reliability, security and footprint of a database system. The attribute footprint was measured; the values for reliability and security were inferred (e.g., feature diagnostic would increase reliability); and the values for price were invented.…”

Section: Subject Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Comparison of exact and approximate multi-objective optimization for software product lines

Olaechea

Rayside

Guo

et al. 2014

Proceedings of the 18th International Software Product Line Conference - Volume 1

View full text Add to dashboard Cite

Software product lines (SPLs) allow stakeholders to manage product variants in a systematical way and derive variants by selecting features. Finding a desirable variant is often difficult, due to the huge configuration space and usually conflicting objectives (e.g., lower cost and higher performance). This scenario can be characterized as a multi-objective optimization problem applied to SPLs. We address the problem using an exact and an approximate algorithm and compare their accuracy, time consumption, scalability, parameter setting requirements on five case studies with increasing complexity. Our empirical results show that (1) it is feasible to use exact techniques for small SPL multi-objective optimization problems, and (2) approximate methods can be used for large problems but require substantial effort to find the best parameter setting for acceptable approximation which can be ameliorated with known good parameter ranges. Finally, we discuss the tradeoff between accuracy and time consumption when using exact and approximate techniques for SPL multi-objective optimization and guide stakeholders to choose one or the other in practice.

show abstract

“…In contrast, Siegmund et al [17] use interacting configurable features to predict non-functional properties like performance from a given configuration and also developed a method to automatically derive an optimized software variant [18]. As discussed earlier, it would be interesting to combine these results with our tailoring approach; for example, the generation of a tailored configuration could not only consider selecting as few features as possible, but rather select features optimal for non-functional properties deemed important for the target use case, e.g., power consumption in an automotive scenario.…”

Section: Related Workmentioning

confidence: 99%

Automatic feature selection in large-scale system-software product lines

2014

View full text Add to dashboard Cite

System software can typically be configured at compile time via a comfortable feature-based interface to tailor its functionality towards a specific use case. However, with the growing number of features, this tailoring process becomes increasingly difficult: As a prominent example, the Linux kernel in v3.14 provides nearly 14 000 configuration options to choose from. Even developers of embedded systems refrain from trying to build a minimized distinctive kernel configuration for their device -and thereby waste memory and money for unneeded functionality.In this paper, we present an approach for the automatic use-case specific tailoring of system software for special-purpose embedded systems. We evaluate the effectiveness of our approach on the example of Linux by generating tailored kernels for well-known applications of the Rasperry Pi and a Google Nexus 4 smartphone. Compared to the original configurations, our approach leads to memory savings of 15-70 percent and requires only very little manual intervention.

show abstract

SPL Conqueror: Toward optimization of non-functional properties in software product lines

Cited by 151 publications

References 36 publications

Towards Modular Verification of Software Product Lines with mCRL2

Towards Modular Verification of Software Product Lines with mCRL2

Comparison of exact and approximate multi-objective optimization for software product lines

Automatic feature selection in large-scale system-software product lines

Contact Info

Product

Resources

About