quality assurance by means of feature models
play

Quality Assurance by Means of Feature Models David Gollasch FOSD - PowerPoint PPT Presentation

Faculty of Computer Science , Institute of Software - and Multimedia - Technology, Chair for Software Technology Quality Assurance by Means of Feature Models David Gollasch FOSD Meeting 2014, Dagstuhl, 07 .05.2014 Contents 1 2 3 Quality


  1. Faculty of Computer Science , Institute of Software - and Multimedia - Technology, Chair for Software Technology Quality Assurance by Means of Feature Models David Gollasch FOSD Meeting 2014, Dagstuhl, 07 .05.2014

  2. Contents 1 2 3 Quality Assurance Fundamentals Discussion by Means of Feature Models SaaS Quality as Attributes in Structural Analysis of Feature Models Feature Models Quality Assurance FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 2 of 21

  3. Motivation Modern business applications are getting increasingly distributed over the Internet as multi - tenant software as a service (SaaS). This leads to new challenges in terms of quality assurance when developing or maintaining such applications, because all customers are directly affected very often. Which effects do software changes have? How are those changes proactively determinable ? How useful are feature models to face this? FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 3 of 21

  4. 1. Fundamentals SaaS Quality Assurance FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 4 of 21

  5. [Schroeter et al. 2012] [Mietzner et al. 2011] [Linden 2007 , p. 6 ff.] Multi-Tenant Software as a Service Applications …can be seen as a special kind of SPL Tenant Config. Configurations possible through SaaS product lines Tenant Config. Instance (SPL) Tenant Config. Domain Engineering Application Engineering FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 5 of 21

  6. [Sneed 2012, EVOL - 21] Why to focus on quality assurance? Laws of software evolution (development of software in time) 1. Law of continous change 4. Law of diminishing productivity 2. Law of increasing complexity 5. Law of restricted growth 3. Law of decreasing quality key argument for quality assurance FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 6 of 21

  7. How to assure quality in general? Take a quality goal Prioritize development process: Remove unnecessary and try to reach it due to further features to avoid wasting development time and money . X exclusive G gradual analyse product structure quantized Q FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 7 of 21

  8. 2. Quality Assurance by Means of Feature Models Quality as Attributes in Feature Models Structural Analysis of Feature Models FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 8 of 21

  9. Qualities as Attributes in Feature Models ‣ Steps to analyze quality goals: Type Quality Goal Question/Interpretation Classification as X, G 1. or Q goal X goal Capabilities Are the requirements fulfilled? How efficient is the feature or G goal Efficiency Find metric or configuration? 2. heuristics How much memory needs the Q goal Resource usage calculation? Consolidation for 3. configuration FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 9 of 21

  10. Example: Comparing Configurations (Summation as Consolidation) „Measurement“ on feature 300 200 100 Consolidation Configuration 1 Configuration 2 300 500 200 100 200 300 300 < 500 FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 10 of 21

  11. Consolidation Methods ‣ simple approach for X goals: binary (achieved/not achieved) 
 ex: If there is one sub feature which does not achieve the X goal, the whole configuration does not achieve the goal. ! ‣ simple arithmetic operations for Q/G goals: e.g. summation ! ‣ complex consolidation method with dependencies , because not every feature set allows a simple summation of the quality values, 
 e.g. in terms of memory consumption. 
 If F A and F B → multiply sum with 0.5. FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 11 of 21

  12. [Siegmund et al. 2012] [Lettner et al. 2011] Further Measurement Approaches ‣ Not every quality goal can be measured easily, e.g. safety properties! ! ‣ Quality measurement at a concrete software instance respectively configuration ! ‣ usage of Benchmarks ‣ usage of model and code ! ‣ Quality determination by means of a business approach ‣ Assigning „costs“ to each feature ‣ Negotiation of „total costs“ for a configuration according to economical principles (discounts, price increase, ...) FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 12 of 21

  13. Structural Similarity of Configurations ‣ When does comparing configurations make sense? ‣ Comparison of similar configurations as a lead for further investigations Configurations are similar, if there are commonalities that can be identified. This does not necessarily refer to the selection of equal features. Similarity also involves structural commonalities . (derived definition) same parent feature FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 13 of 21

  14. Structural Similarity of Configurations ‣ When does comparing configurations make sense? ‣ Comparison of similar configurations as a lead for further investigations Configurations are similar, if there are commonalities that can be identified. This does not necessarily refer to the selection of equal features. Similarity also involves structural commonalities . (derived definition) same parent feature same group FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 13 of 21

  15. Structural Similarity of Configurations ‣ When does comparing configurations make sense? ‣ Comparison of similar configurations as a lead for further investigations Configurations are similar, if there are commonalities that can be identified. This does not necessarily refer to the selection of equal features. Similarity also involves structural commonalities . (derived definition) same parent feature same group same features of a group FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 13 of 21

  16. Structural Similarity of Configurations ‣ When does comparing configurations make sense? ‣ Comparison of similar configurations as a lead for further investigations Configurations are similar, if there are commonalities that can be identified. This does not necessarily refer to the selection of equal features. Similarity also involves structural commonalities . (derived definition) same parent feature same group same features of a group same features of a parent feature FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 13 of 21

  17. Structural Similarity of Configurations ‣ When does comparing configurations make sense? ‣ Comparison of similar configurations as a lead for further investigations Configurations are similar, if there are commonalities that can be identified. This does not necessarily refer to the selection of equal features. Similarity also involves structural commonalities . (derived definition) same parent feature same group same attributes same features of a group same features of a parent feature FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 13 of 21

  18. Draw Conclusions (1) Quality change due to software evolution FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 14 of 21

  19. Draw Conclusions (2) Quality difference of similar configurations …and its change due to software evolution FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 15 of 21

  20. Structural Analysis of Feature Models ‣ Comparisons only based on a feature model and existing derived configurations. ‣ Complete automation possible ! ‣ Conceptual implementation in Java as Eclipse plugin within the extFM - Tooling project (https://github.com/ extFM/extFM - Tooling/) FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 16 of 21

  21. 3. Discussion ‣ Possible starting points for further research Analysis of the Further research in shown methods Extension of analysis attributed feature without the SaaS tools models context Integration of the Empirical Combination of prototypical investigation in attributed and implementation in a practicability in real structural analysis practically usable tool projects FOSD Meeting, 07 .05.2014 Quality Assurance by Means of Feature Models 17 of 21

  22. (References on next slides) Faculty of Computer Science , Institute of Software - and Multimedia - Technology, Chair for Software Technology Thank you for your attention! 1 2 3 Quality Assurance Fundamentals Discussion by Means of Feature Models SaaS Quality as Attributes in Structural Analysis of Feature Models Feature Models Quality Assurance FOSD Meeting 2014, Dagstuhl, 07 .05.2014

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend