NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
External Dependencies-Driven Architecture Discovery and Analysis of Implemented Systems A method for architecture discovery and analysis of implemented systems (AIS) is disclosed. The premise of the method is that architecture decisions are inspired and influenced by the external entities that the software system makes use of. Examples of such external entities are COTS components, frameworks, and ultimately even the programming language itself and its libraries. Traces of these architecture decisions can thus be found in the implemented software and is manifested in the way software systems use such external entities. While this fact is often ignored in contemporary reverse engineering methods, the AIS method actively leverages and makes use of the dependencies to external entities as a starting point for the architecture discovery. The AIS method is demonstrated using the NASA's Space Network Access System (SNAS). The results show that, with abundant evidence, the method offers reusable and repeatable guidelines for discovering the architecture and locating potential risks (e.g. low testability, decreased performance) that are hidden deep in the implementation. The analysis is conducted by using external dependencies to identify, classify and review a minimal set of key source code files. Given the benefits of analyzing external dependencies as a way to discover architectures, it is argued that external dependencies deserve to be treated as first-class citizens during reverse engineering. The current structure of a knowledge base of external entities and analysis questions with strategies for getting answers is also discussed.
Document ID
20180001172
Acquisition Source
Goddard Space Flight Center
Document Type
Reprint (Version printed in journal)
External Source(s)
Authors
Ganesan, Dharmalingam
(Maryland Univ. College Park, MD, United States)
Lindvall, Mikael
(Maryland Univ. College Park, MD, United States)
Ron, Monica
(Honeywell, Inc. Annapolis, MD, United States)
Date Acquired
February 13, 2018
Publication Date
March 1, 2014
Publication Information
Publication: ACM Transactions on Software Engineering and Methodology (TOSEM)
Publisher: Association for Computing Machinery
Volume: 23
Issue: 2
Subject Category
Computer Programming And Software
Computer Systems
Report/Patent Number
LEGNEW-OLDGSFC-GSFC-LN-1169
Funding Number(s)
CONTRACT_GRANT: NNX08AZ60G
Distribution Limits
Public
Copyright
Public Use Permitted.
No Preview Available