Back to search

University of British Columbia

Dockerfile flakiness : characterization and repair

Abstract

dc:description

Docker, a platform utilizing OS-level virtualization, packages applications into lightweight, portable images defined by instructions in a Dockerfile. However, the reliability of these Dockerfiles can be compromised by inconsistent and unpredictable behaviors, known as flakiness, which present significant challenges to the stability of Continuous Integration and Delivery (CI/CD) pipelines. This issue can lead to unreliable deployments and increased debugging efforts, yet it remains underexplored in current research. We conduct a systematic analysis of Dockerfile flakiness, presenting a comprehensive taxonomy of common flakiness categories, including dependency-related errors and server connectivity issues. Furthermore, we introduce FlakiDock, a tool leveraging large language models and retrieval-augmented generation techniques with dynamic analysis and an iterative feedback loop to automatically repair flaky Dockerfiles. Our evaluation shows that FlakiDock achieves a 73.55% repair accuracy, outperforming existing tools such as PARFUM by 12,581%, GPT-4-based prompting by 94.63%, and Llama3.1-based prompting by 110.87%. These results underscore the effectiveness of FlakiDock in addressing Dockerfile flakiness and improving build reliability.

Degree

thesis:*
Name thesis:degree_name
Master of Applied Science - MASc
Level thesis:degree_level
master's
Discipline thesis:degree_discipline
Electrical and Computer Engineering
Grantor dc:publisher
University of British Columbia
Year dc:date
2024

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • ShabaniMirzaei, Taha
Contributors dc:contributor
  • Mesbah, Ali

Rights

dc:rights
Statement dc:rights
  • Attribution-NonCommercial-NoDerivatives 4.0 International
Language dc:language
eng

Identifiers

dc:identifier.*
Handle dc:identifier
http://hdl.handle.net/2429/89482

Chain of custody

source
Harvested from
University of British Columbia
Base URL
circle.library.ubc.ca/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

ShabaniMirzaei, Taha. Dockerfile flakiness : characterization and repair. master's thesis, University of British Columbia, 2024. http://hdl.handle.net/2429/89482