AI for DevSecOps: A Landscape and Future Opportunities

ACM Transactions on Software Engineering and Methodology, Год журнала: 2025, Номер unknown

Опубликована: Янв. 16, 2025

DevOps has emerged as one of the most rapidly evolving software development paradigms. With growing concerns surrounding security in systems, DevSecOps paradigm gained prominence, urging practitioners to incorporate practices seamlessly into workflow. However, integrating workflow can impact agility and impede delivery speed. Recently, advancement artificial intelligence (AI) revolutionized automation various domains, including security. AI-driven approaches, particularly those leveraging machine learning or deep learning, hold promise automating workflows. They have potential reduce manual efforts be incorporated support consistent speed while aligning with principles paradigm. This paper seeks contribute critical intersection AI by presenting a comprehensive landscape techniques applicable identifying avenues for enhancing security, trust, efficiency processes. We analyzed 99 research papers spanning from 2017 2023. Specifically, we address two key questions (RQs). In RQ1, identified 12 tasks associated process reviewed existing problems they addressed, 65 benchmarks used evaluate approaches. Drawing insights our findings, RQ2, discussed state-of-the-art highlighted 15 challenges research, proposed corresponding future opportunities.

Язык: Английский

---

Control and Data Flow in Security Smell Detection for Infrastructure as Code: Is It Worth the Effort?

Опубликована: Май 1, 2023

Infrastructure as Code is the practice of developing and maintaining computing infrastructure through executable source code. Unfortunately, IaC has also brought about new cyber attack vectors. Prior work therefore proposed static analyses that detect security smells in files. However, they have so far remained at a shallow level, disregarding control data flow scripts under analysis, may lack awareness specific syntactic constructs. These limitations inhibit quality their results. To address these limitations, this paper, we present GASEL, novel smell detector for Ansible language. It uses graph queries on program dependence graphs to 7 smells. Our evaluation an oracle 243 real-world comparison against two state-of-the-art detectors shows syntax, flow, enables our approach substantially improve both precision recall. We further question whether additional effort required develop run such justified practice. end, investigate prevalence indirection across more than 15 000 scripts. find over 55% contain data-flow indirection, 32% require whole-project analysis detect. findings motivate need deeper tools vulnerabilities IaC.

Язык: Английский

---

Automated Infrastructure as Code Program Testing

IEEE Transactions on Software Engineering, Год журнала: 2024, Номер 50(6), С. 1585 - 1599

Опубликована: Май 1, 2024

Infrastructure as Code (IaC) enables efficient deployment and operation, which are crucial to releasing software quickly. As setups can be complex, developers implement IaC programs in general-purpose programming languages like TypeScript Python, using PL-IaC solutions Pulumi AWS CDK. The reliability of such is even more relevant than traditional because a bug impacts the whole system. Yet, though testing standard development practice, it rarely used for programs. For instance, August 2022, less 1% public on GitHub implemented tests. Available program techniques severely limit velocity or require much effort.

To solve these issues, we propose Automated Configuration Testing (ACT), methodology test many configurations quickly with low ACT automatically mocks all resource definitions uses generator oracle plugins generation validation. We ProTI, tool type-based oracle, support application specifications. Our evaluation 6 081 from artificial benchmarks shows that ProTI directly applied existing programs, finds bugs where current infeasible, reusing generators oracles thanks its pluggable architecture.

Язык: Английский

---

What Do Infrastructure-as-Code Practitioners Discuss: An Empirical Study on Stack Overflow

Опубликована: Окт. 26, 2023

Background. Infrastructure-as-Code (IaC) is an emerging practice to manage cloud infrastructure resources for software systems. Modern development has evolved embrace IaC as a best consistently provisioning and managing using various tools such Terraform Ansible. However, recent studies highlighted that developers still encounter challenges with tools. Aims. We aim in this paper understand the different analyze trend of seeking assistance on Q&A platforms context IaC. To end, we conduct large-scale empirical study investigating developers' discussions Stack Overflow. Method. first collect IaC-relevant tags Overflow, constituting dataset comprises 52,692 questions 64,078 answers. Then, group into specific topics Latent Dirichlet Allocation (LDA) method, which optimize Genetic Algorithm (GA) parameter's fine-tuning. Finally, gain better insights, identified based criteria popularity difficulty. Results. Our findings reveal average yearly increase 150% terms IaC-related 135% users between 2011 2022. Furthermore, observe revolve around seven main topics: server configuration, policy networking, deployment pipelines, variable management, templating, file management. Notably, found configuration management are most popular topics, i.e., discussed among developers, while pipelines templating difficult. Conclusions. results shed light often encountered by platforms. These important implications practitioners support real-world settings researchers community needs further investigate aspects.

Язык: Английский

---

Detecting Inconsistencies in Software Architecture Documentation Using Traceability Link Recovery

Опубликована: Март 1, 2023

Documenting software architecture is important for a system's success. Software documentation (SAD) makes information about the system available and eases comprehensibility. There are different forms of SADs like natural language texts formal models with benefits purposes. However, there can be inconsistent in same system. Inconsistent then cause flaws development maintenance. To tackle this, we present an approach inconsistency detection SAD models. We make use traceability link recovery (TLR) extend existing approach. utilize results from TLR to detect unmentioned (i.e., model elements without documentation) missing described but not modeled elements). In our evaluation, measure how adaptations on affected its performance. Moreover, evaluate detection. benchmark multiple open source projects compare baseline approaches. For TLR, achieve excellent F 1 -score 0.81, significantly outperforming other approaches by at least 0.24. Our also achieves (accuracy: 0.93) detecting good 0.75). These outperform competing baselines. Although see room improvements, show that inconsistencies using promising.

Язык: Английский

---

Infrastructure-as-Code Ecosystems

Springer eBooks, Год журнала: 2023, Номер unknown, С. 215 - 245

Опубликована: Янв. 1, 2023

Язык: Английский

---

An empirical study of task infections in Ansible scripts

Empirical Software Engineering, Год журнала: 2023, Номер 29(1)

Опубликована: Дек. 29, 2023

Язык: Английский

---

Anonymization-as-a-Service: The Service Center Transcripts Industrial Case

Lecture notes in computer science, Год журнала: 2023, Номер unknown, С. 261 - 275

Опубликована: Янв. 1, 2023

Язык: Английский

---

Towards a Taxonomy of Infrastructure as Code Misconfigurations: An Ansible Study

Communications in computer and information science, Год журнала: 2024, Номер unknown, С. 83 - 103

Опубликована: Окт. 18, 2024

Язык: Английский

---

Deep learning with class-level abstract syntax tree and code histories for detecting code modification requirements

Journal of Systems and Software, Год журнала: 2023, Номер 206, С. 111851 - 111851

Опубликована: Сен. 15, 2023

Язык: Английский

---