Refactoring of security antipatterns in distributed Java components
Seiten
2010
University of Bamberg Press (Verlag)
978-3-923507-67-2 (ISBN)
University of Bamberg Press (Verlag)
978-3-923507-67-2 (ISBN)
- Titel aus dem Verlags-Programm gestrichen
- Artikel merken
The importance of JAVA as a programming and execution environment has grown steadily over the past decade. Furthermore, the IT industry has adapted JAVA as a major building block for the creation of new middleware as well as a technology facilitating the migration of existing applications towards web-driven environments.
Parallel in time, the role of security in distributed environments has gained attention, as a large amount of middleware applications has replaced enterprise-level mainframe systems. The protection of confidentiality, integrity and availability are therefore critical for the market success of a product. The vulnerability level of every product is determined by the weakest embedded component, and selling vulnerable products can cause enormous economic damage to software vendors.
An important goal of this work is to create the awareness that the usage of a programming language, which is designed as being secure, is not sufficient to create secure and trustworthy distributed applications. Moreover, the incorporation of the threat model of the programming language improves the risk analysis by allowing a better definition of the attack surface of the application.
The evolution of a programming language leads towards common patterns for solutions for recurring quality aspects. Suboptimal solutions, also known as ´antipatterns´, are typical causes for quality weaknesses such as security vulnerabilities. Moreover, the exposure to a specific environment is an important parameter for threat analysis, as code considered secure in a specific scenario can cause unexpected risks when switching the environment.
Antipatterns are a well-established means on the abstractional level of system modeling to inform about the effects of incomplete solutions, which are also important in the later stages of the software development process. Especially on the implementation level, we see a deficit of helpful examples, that would give programmers a better and holistic understanding.
In our basic assumption, we link the missing experience of programmers regarding the security properties of patterns within their code to the creation of software vulnerabilities. Traditional software development models focus on security properties only on the meta layer. To transfer these efficiently to the practical level, we provide a three-stage approach:
First, we focus on typical security problems within JAVA applications, and develop a standardized catalogue of ´antipatterns´ with examples from standard software products. Detecting and avoiding these antipatterns positively influences software quality.
We therefore focus, as second element of our methodology, on possible enhancements to common models for the software development process. These help to control and identify the occurrence of antipatterns during development activities, i. e. during the coding phase and during the phase of component assembly, integrating one´s own and third party code.
Within the third part, and emphasizing the practical focus of this research, we implement prototypical tools for support of the software development phase. The practical findings of this research helped to enhance the security of the standard JAVA platforms and JEE frameworks. We verified the relevance of our methods and tools by applying these to standard software products leading to a measurable reduction of vulnerabilities and an information exchange with middleware vendors (Sun Microsystems, JBoss) targeting runtime security.
Our goal is to enable software architects and software developers developing end-user applications to apply our findings with embedded standard components on their environments. From a high-level perspective, software architects profit from this work through the projection of the quality-of-service goals to protection details. This supports their task of deriving security requirements when selecting standard components. In order to give implementation-near practitioners a helpful starting point to benefit from our research we provide tools and case-studies to achieve security improvements within their own code base.
Parallel in time, the role of security in distributed environments has gained attention, as a large amount of middleware applications has replaced enterprise-level mainframe systems. The protection of confidentiality, integrity and availability are therefore critical for the market success of a product. The vulnerability level of every product is determined by the weakest embedded component, and selling vulnerable products can cause enormous economic damage to software vendors.
An important goal of this work is to create the awareness that the usage of a programming language, which is designed as being secure, is not sufficient to create secure and trustworthy distributed applications. Moreover, the incorporation of the threat model of the programming language improves the risk analysis by allowing a better definition of the attack surface of the application.
The evolution of a programming language leads towards common patterns for solutions for recurring quality aspects. Suboptimal solutions, also known as ´antipatterns´, are typical causes for quality weaknesses such as security vulnerabilities. Moreover, the exposure to a specific environment is an important parameter for threat analysis, as code considered secure in a specific scenario can cause unexpected risks when switching the environment.
Antipatterns are a well-established means on the abstractional level of system modeling to inform about the effects of incomplete solutions, which are also important in the later stages of the software development process. Especially on the implementation level, we see a deficit of helpful examples, that would give programmers a better and holistic understanding.
In our basic assumption, we link the missing experience of programmers regarding the security properties of patterns within their code to the creation of software vulnerabilities. Traditional software development models focus on security properties only on the meta layer. To transfer these efficiently to the practical level, we provide a three-stage approach:
First, we focus on typical security problems within JAVA applications, and develop a standardized catalogue of ´antipatterns´ with examples from standard software products. Detecting and avoiding these antipatterns positively influences software quality.
We therefore focus, as second element of our methodology, on possible enhancements to common models for the software development process. These help to control and identify the occurrence of antipatterns during development activities, i. e. during the coding phase and during the phase of component assembly, integrating one´s own and third party code.
Within the third part, and emphasizing the practical focus of this research, we implement prototypical tools for support of the software development phase. The practical findings of this research helped to enhance the security of the standard JAVA platforms and JEE frameworks. We verified the relevance of our methods and tools by applying these to standard software products leading to a measurable reduction of vulnerabilities and an information exchange with middleware vendors (Sun Microsystems, JBoss) targeting runtime security.
Our goal is to enable software architects and software developers developing end-user applications to apply our findings with embedded standard components on their environments. From a high-level perspective, software architects profit from this work through the projection of the quality-of-service goals to protection details. This supports their task of deriving security requirements when selecting standard components. In order to give implementation-near practitioners a helpful starting point to benefit from our research we provide tools and case-studies to achieve security improvements within their own code base.
| Reihe/Serie | Schriften aus der Fakultät Wirtschaftsinformatik und Angewandte Informatik der Otto-Friedrich-Universität Bamberg ; 5 |
|---|---|
| Zusatzinfo | graph. Darst. |
| Sprache | englisch |
| Maße | 150 x 210 mm |
| Einbandart | Paperback |
| Themenwelt | Informatik ► Programmiersprachen / -werkzeuge ► Java |
| Schlagworte | Computersicherheit • Hardcover, Softcover / Informatik, EDV/Programmiersprachen • Java <Programmiersprache> • Java <Programmiersprache> • Online-Publikation |
| ISBN-10 | 3-923507-67-4 / 3923507674 |
| ISBN-13 | 978-3-923507-67-2 / 9783923507672 |
| Zustand | Neuware |
| Haben Sie eine Frage zum Produkt? |
Mehr entdecken
aus dem Bereich
aus dem Bereich
mit über 150 Workouts in Java und Python
Buch (2023)
Carl Hanser (Verlag)
CHF 41,95
Einführung, Ausbildung, Praxis
Buch | Hardcover (2023)
Rheinwerk (Verlag)
CHF 69,85
