Towards an Integrated View on Architecture and its Evolution

Information about the evolution of a software architecture can be found in the source basis of a project and in the release history data such as modification and problem reports. Existing approaches deal with these two data sources separately and do not exploit the integration of their analyses. In this paper, we present an architecture analysis approach that provides an integration of both kinds of evolution data. The analysis applies fact extraction and generates specific directed attributed graphs; nodes represent source code entities and edges represent relationships such as accesses, includes, inherits, invokes, and coupling between certain architectural elements. The integration of data is then performed on a meta-model level to enable the generation of architectural views using binary relational algebra. These integrated architectural views show intended and unintended couplings between architectural elements, hence pointing software engineers to locations in the system that may be critical for on-going and future maintenance activities. We demonstrate our analysis approach using a large open source software system.     

An empirical investigation of architectural prototyping

Architectural prototyping is the process of using executable code to investigate stakeholders’ software architecture concerns with respect to a system under development. Previous work has established this as a useful and cost-effective way of exploration and learning of the design space of a system and in addressing issues regarding quality attributes, architectural risks, and the problem of knowledge transfer and conformance. However, the actual industrial use of architectural prototyping has not been thoroughly researched so far. In this article, we report from three studies of architectural prototyping in practice. First, we report findings from an ethnographic study of practicing software architects. Secondly, we report from a focus group on architectural prototyping involving architects from four companies. And, thirdly, we report from a survey study of 20 practicing software architects and software developers. Our findings indicate that architectural prototyping plays an important and frequent role in resolving problems experimentally, but less so in exploring alternative solutions. Furthermore, architectural prototypes include end-user or business related functionality rather than purely architectural functionality. Based on these observations we provide recommendations for effective industrial architectural prototyping.     

Two controlled experiments on model-based architectural decision making

ContextIn recent years, architectural design decisions are becoming more and more common for documenting software architectures. Rather than describing the structure of software systems, architectural decisions capture the design rationale and – often reusable – architectural knowledge. Many approaches and tools have been proposed in the literature to support architectural decision making and documentation (for instance, based on models, ontologies, or templates). In this context, the capturing, organization, and effective reuse of architectural knowledge has gained a lot of attention.ObjectiveHowever, there is little empirical evidence about the supportive effect of reusable architectural knowledge on the effectiveness and efficiency of architectural decision making.MethodTo investigate these aspects, we conducted two separate controlled experiments with software architecture students in which we tested the supportive effect of reusable decision models in decision making and documentation.ResultsOur results show that the use of reusable decision models can significantly increase both the efficiency and the effectiveness of novice architects.ConclusionWe can report, that our findings are in line with similar studies and support the claims regarding reusable architectural design decisions in principle.     

An Approach for Recovering Distributed System Architectures

Reasoning about software systems at the architectural level is key to effective software development, management, evolution and reuse. All too often, though, the lack of appropriate documentation leads to a situation where architectural design information has to be recovered directly from implemented software artifacts. This is a very demanding process, particularly when involving recovery of runtime abstractions (clients, servers, interaction protocols, etc.) that are typical to the design of distributed software systems. This paper presents an exploratory reverse engineering approach, called X-ray, to aid programmers in recovering architectural runtime information from a distributed system's existing software artifacts. X-ray comprises three domain-based static analysis techniques, namely component module classification, syntactic pattern matching, and structural reachability analysis. These complementary techniques can facilitate the task of identifying a distributed system's implemented executable components and their potential runtime interconnections. The component module classification technique automatically distinguishes source code modules according to the executables components they implement. The syntactic pattern matching technique in turn helps to recognise specific code fragments that may implement typical component interaction features. Finally, the structural reachability analysis technique aids in the association of those features to the code specific for each executable component. The paper describes and illustrates the main concepts underlying each technique, reports on their implementation as a suit of new and off-the-shelf tools, and, to give evidence of the utility of the approach, provides a detailed account of a successful application of the three techniques to help recover a static approximation of the runtime architecture for Field, a publicly-available distributed programming environment.     

Developing and evolving a DSL-based approach for runtime monitoring of systems of systems

Complex software-intensive systems are often described as systems of systems (SoS) due to their heterogeneous architectural elements. As SoS behavior is often only understandable during operation, runtime monitoring is needed to detect deviations from requirements. Today, while diverse monitoring approaches exist, most do not provide what is needed to monitor SoS, e.g., support for dynamically defining and deploying diverse checks across multiple systems. In this paper we report on our experiences of developing, applying, and evolving an approach for monitoring an SoS in the domain of industrial automation software, that is based on a domain-specific language (DSL). We first describe our initial approach to dynamically define and check constraints in SoS at runtime and then motivate and describe its evolution based on requirements elicited in an industry collaboration project. We furthermore describe solutions we have developed to support the evolution of our approach, i.e., a code generation approach and a framework to automate testing the DSL after changes. We evaluate the expressiveness and scalability of our new DSL-based approach using an industrial SoS. We also discuss lessons we learned. Our results show that while developing a DSL-based approach is a good solution to support industrial users, one must prepare the approach for evolution, by making it extensible and adaptable to future scenarios. Particularly, support for automated (re-)generation of tools and code after changes and automated testing are essential.     

Harmonizing architectural decisions with component view models using reusable architectural knowledge transformations and constraints

Architectural design decisions (ADDs) have been used in recent years for capturing design rationale and documenting architectural knowledge (AK). However, various architectural design views still provide the most common means for describing and communicating architectural design. The evolution of software systems requires that both ADDs and architectural design views are documented and maintained, which is a tedious and time-consuming task in the long run. Also, in lack of a systematic and automated support for bridging between ADDs and architectural design views, decisions and designs tend to become inconsistent over time. In our proposal, we introduce a reusable AK transformation language for supporting the automated transformation of reusable AK knowledge to component-and-connector models, the architectural design view used most commonly today. In addition, reusable consistency checking rules verify the consistency between decisions and designs. We evaluate our approach in an industrial case study and show that it offers high reusability, provides automation, and can, in principle, deal with large numbers of recurring decisions.     

Detecting deviations in the code using architecture view-based drift analysis

ContextOne of the key requirements for the code is conformance with the architecture. Architectural drift implies the diverging of the implemented code from the architecture design of the system. Manually checking the consistency between the implemented code and architecture can be intractable and cumbersome for large-scale systems.ObjectiveThis article proposes a holistic, automated architecture drift analysis approach that explicitly focuses on the adoption of architecture views. The approach builds on, complements, and enhances existing architecture conformance analysis methods that do not adopt a holistic approach or fail to address the architecture viewpoints.MethodA model-driven development approach is adopted in which architecture views are represented as specifications of domain-specific languages. The code in its turn, is analyzed, and the architectural view specifications are reconstructed, which are then automatically checked with the corresponding architecture models.ResultsTo illustrate the approach, we have applied a systematic case study research for an architecture drift analysis of the business-to-customer (B2C) system within a large-scale software company.ConclusionThe case study research showed that divergences and absences of architectural elements could be detected in a cost-effective manner with the proposed approach.     

Architecting Fault Tolerance with Exception Handling: Verification and Validation

When building dependable systems by integrating untrusted software components that were not originally designed to interact with each other, it is likely the occurrence of architectural mismatches related to assumptions in their failure behaviour. These mismatches, if not prevented during system design, have to be tolerated during runtime. This paper presents an architectural abstraction based on exception handling for structuring fault-tolerant software systems. This abstraction comprises several components and connectors that promote an existing untrusted software element into an idealised fault-tolerant architectural element. Moreover, it is considered in the context of a rigorous software development approach based on formal methods for representing the structure and behaviour of the software architecture. The proposed approach relies on a formal specification and verification for analysing exception propagation, and verifying important dependability properties, such as deadlock freedom, and scenarios of architectural reconfiguration. The formal models are automatically generated using model transformation from UML diagrams: component diagram representing the system structure, and sequence diagrams representing the system behaviour. Finally, the formal models are also used for generating unit and integration test cases that are used for assessing the correctness of the source code. The feasibility of the proposed architectural approach was evaluated on an embedded critical case study. Patrick Brito is supported by Fapesp/Brazil under Grant No. 06/02116–2 and CAPES/Brazil under Grant No. 0722–07–3. Cecília Rubira is partially supported by CNPq/Brazil under Grant Nos. 301446/2006–7 and 484138/2006–5.     

A fault-tolerant architectural approach for dependable systems

A system's structure enables it to generate its intended behavior from its components' behavior. A well-structured system simplifies relationships among components, which can increase dependability. With software systems, the architecture is an abstraction of the structure. Architectural reasoning about dependability has become increasingly important because emerging applications are increasingly complex. We've developed an architectural approach for effectively representing and analyzing fault-tolerant software systems. The proposed solution relies on exception handling to tolerate faults associated with component and connector failures, architectural mismatches, and configuration faults. Our approach, a specialization of the peer-to-peer architectural style, hides inside the architectural elements the complexities of exception handling and propagation. Our goal is to improve a system's overall reliability and availability by making it tolerant of nonmalicious faults.     

Extracting models from source code in software modernization

Model-driven software modernization is a discipline in which model-driven development (MDD) techniques are used in the modernization of legacy systems. When existing software artifacts are evolved, they must be transformed into models to apply MDD techniques such as model transformations. Since most modernization scenarios (e.g., application migration) involve dealing with code in general-purpose programming languages (GPL), the extraction of models from GPL code is an essential task in a model-based modernization process. This activity could be performed by tools to bridge grammarware and MDD technical spaces, which is normally carried out by dedicated parsers. Grammar-to-Model Transformation Language (Gra2MoL) is a domain-specific language (DSL) tailored to the extraction of models from GPL code. This DSL is actually a text-to-model transformation language which can be applied to any code conforming to a grammar. Gra2MoL aims to reduce the effort needed to implement grammarware-MDD bridges, since building dedicated parsers is a complex and time-consuming task. Like ATL and RubyTL languages, Gra2MoL incorporates the binding concept needed to write mappings between grammar elements and metamodel elements in a simple declarative style. The language also provides a powerful query language which eases the retrieval of scattered information in syntax trees. Moreover, it incorporates extensibility and grammar reuse mechanisms. This paper describes Gra2MoL in detail and includes a case study based on the application of the language in the extraction of models from Delphi code.     

Real-Time Reflexion Modelling in architecture reconciliation: A multi case study

ContextReflexion Modelling is considered one of the more successful approaches to architecture reconciliation. Empirical studies strongly suggest that professional developers involved in real-life industrial projects find the information provided by variants of this approach useful and insightful, but the degree to which it resolves architecture conformance issues is still unclear.ObjectiveThis paper aims to assess the level of architecture conformance achieved by professional architects using Reflexion Modelling, and to determine how the approach could be extended to improve its suitability for this task.MethodAn in vivo, multi-case-study protocol was adopted across five software systems, from four different financial services organizations. Think-aloud, video-tape and interview data from professional architects involved in Reflexion Modelling sessions were analysed qualitatively.ResultsThis study showed that (at least) four months after the Reflexion Modelling sessions less than 50% of the architectural violations identified were removed. The majority of participants who did remove violations favoured changes to the architectural model rather than to the code. Participants seemed to work off two specific architectural templates, and interactively explored their architectural model to focus in on the causes of violations, and to assess the ramifications of potential code changes. They expressed a desire for dependency analysis beyond static-source-code analysis and scalable visualizations.ConclusionThe findings support several interesting usage-in-practice traits, previously hinted at in the literature. These include (1) the iterative analysis of systems through Reflexion models, as a precursor to possible code change or as a focusing mechanism to identify the location of architecture conformance issues, (2) the extension of the approach with respect to dependency analysis of software systems and architectural modelling templates, (3) improved visualization support and (4) the insight that identification of architectural violations in itself does not lead to their removal in the majority of instances.