Efficient regression testing of distributed real-time reactive systems in the context of model-driven development

Regression testing is indispensable, especially for real-time distributed systems to ensure that existing functionalities are not affected by changes. Despite recent advances, regression testing for distributed systems remains challenging and extremely costly. Existing techniques often require runni...

Full description

Saved in:
Bibliographic Details
Published in:Software and systems modeling Vol. 22; no. 5; pp. 1565 - 1587
Main Authors: Babaei, Majid, Dingel, Juergen
Format: Journal Article
Language:English
Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.10.2023
Springer Nature B.V
Subjects:
ISSN:1619-1366, 1619-1374
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Regression testing is indispensable, especially for real-time distributed systems to ensure that existing functionalities are not affected by changes. Despite recent advances, regression testing for distributed systems remains challenging and extremely costly. Existing techniques often require running a failing system several times before detecting a regression. As a result, conventional approaches that use re-execution without considering the inherent non-determinism of distributed systems, and providing no (or low) control over execution are inadequate in many ways. In this paper, we present MRegTest , a replay-based regression testing framework in the context of model-driven development to facilitate deterministic replay of traces for detecting regressions while offering sufficient control for the purpose of testing over the execution of the changed system. The experimental results show that compared to the traditional approaches that annotate traces with timestamps and variable values MRegTest detects almost all regressions while reducing the size of the trace significantly and incurring similar runtime overhead.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1619-1366
1619-1374
DOI:10.1007/s10270-023-01086-5