Understanding Software Vulnerabilities in the Maven Ecosystem: Patterns, Timelines, and Risks

Md Fazle Rabbi Rajshakhar Paul Arifa Islam Champa Minhaz F. Zibran · 2025

Read Understanding Software Vulnerabilities in the Maven Ecosystem: Patterns, Timelines, and Risks by Md Fazle Rabbi Rajshakhar Paul Arifa Islam Champa Minhaz F. Zibran (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-22391.txt.

Introduction

Md Fazle Rabbi, Rajshakhar Paul, Arifa Islam Champa, and Minhaz Zibran dissect security flaws in the Maven ecosystem, the library hub behind much Java software. Reused components carry risk when flaws linger, so the team ran a large-scale analysis of vulnerability types, documentation delays, and fix timelines.

They counted over seventy-seven thousand vulnerable releases tied to two hundred twenty-six weakness types. On average, a flaw waited nearly five years just to be documented, and four point four years to be resolved; some stayed open past a decade. The authors warn that these delays leave library users exposed and call for faster, more careful vulnerability handling across the ecosystem.

Catalog record

CREATORMd Fazle Rabbi Rajshakhar Paul Arifa Islam Champa Minhaz F. Zibran
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE20 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-22391

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Understanding Software Vulnerabilities in the Maven Ecosystem: Patterns, Timelines, and Risks Md Fazle Rabbi Rajshakhar Paul Arifa Islam Champa Minhaz F. Zibran Abstract Vulnerabilities in software libraries and reusable components cause major security challenges, particularly in dependency-heavy ecosystems such as Maven. This paper presents a large-scale analysis of vulnerabilities in the Maven ecosystem using the Goblin framework. Our analysis focuses on the aspects and implications of vulnerability types, documentation delays, and resolution timelines. We identify 77,393 vulnerable releases with 226 unique CWEs. On average, vulnerabilities take nearly half a decade to be documented and 4.4 years to be resolved, with some remaining unresolved for even over a decade. The delays in documenting and fixing vulnerabilities incur security risks for the library users emphasizing the need for more careful and efficient vulnerability management in the Maven ecosystem. I Introduction Modern software applications heavily rely on libraries and external dependencies [ 1 ] . These components help developers speed up progress and reduce costs by providing reusable code. Studies have found that third-party libraries often contribute between 30% to 90% of the source code in typical applications [ 2 , 3 ] . However, while these libraries offer significant advantages, they also introduce substantial security risks [ 4 ] . Vulnerabilities in libraries can spread across software ecosystems, affecting millions of users. High-profile incidents, such as Log4Shell vulnerability in Apache Log4j [ 5 ] and the SolarWinds supply chain attack [ 6 ] , demonstrate the serious impact of compromised dependencies. The prevalence of software vulnerabilities has raised significant concerns about the security and reliability of modern software systems. As software ecosystems expand, their attack surfaces increase, making the security of external libraries a critical concern for developers and organizations. Recent data from the National Vulnerability Database (NVD) shows a sharp increase in vulnerabilities. Projections estimate 35,000 new vulnerabilities in 2024, marking a 25% rise from the previous year [ 7 ] . Many of these vulnerabilities are found in widely used software libraries, pointing to increasing security risks in the software supply chain. If left unaddressed, these vulnerabilities can affect entire software ecosystems. The Common Weakness Enumeration (CWE) framework offers a structured approach for identifying and classifying software weaknesses [ 8 ] , which are foundational issues in code that can lead to security vulnerabilities if exploited. While CWE helps developers and organizations better understand and mitigate potential security risks, significant knowledge gaps remain regarding how vulnerabilities develop within the libraries. For instance, vulnerabilities often remain undocumented for extended periods, leaving users unaware of the risks associated with the libraries they rely on. Even when vulnerabilities are disclosed, the process of patching them can vary significantly across projects, leading to prolonged exposure. Additionally, understanding the common patterns of software weaknesses, such as those categorized under CWE, is critical for prioritizing mitigation efforts. Addressing these gaps is essential for improving security practices, particularly in ecosystems like Maven that rely heavily on dependencies. Vulnerabilities in libraries pose not only potential risks but also significant threats to applications across many industries. By better understanding the types of vulnerabilities, their discovery timelines, and the resolution processes, developers can make more informed decisions about which libraries to use, create more robust security policies, and enhance strategies for managing software supply chain risks. Towards this goal, we, therefore, examine the key aspects of software vulnerabilities in Maven ecosystem by investigating the following three research questions (RQs): RQ1: What are the most frequently occurring software weaknesses in libraries? RQ2: How long do vulnerabilities remain undocumented, and what percentage of releases are deployed with known vulnerabilities? RQ3: What are the typical resolution times for vulnerabilities, and what proportion of libraries remain unresolved? Our findings provide actionable insights for improving software security practices, particularly in dependency-heavy ecosystems like Maven. We make our scripts and data publicly available as a replication package [ 9 ] . II Methodology In this study, we use the Goblin framework [ 10 ] as the primary data source, which provides a Neo4j-based dependency graph of the Maven Central repository. The dataset includes 658,078 artifact or library nodes and 14,459,139 release nodes, where each library node has a unique Maven identifier and release nodes capture version details and timestamps. Additionally, the dataset version with_metrics_goblin_maven_30_08_24.dump [ 11 ] , dated August 30, 2024, includes 44,035,495 AddedValue nodes. These AddedValue nodes include enriched metrics such as Common Vulnerabilities and Exposures (CVEs) sourced from the OSV dataset [ 12 ] , along with other attributes like freshness, popularity, and speed. From the dataset, we identify 77,393 releases with vulnerabilities. These releases are associated with 1,411 unique libraries. We then retrieve all releases from these 1,411 libraries, regardless of whether they have vulnerabilities. This results in a refined dataset comprising 125,816 releases associated with the 1,411 libraries. This refined dataset (RD) serves as the basis for further analysis in our study. II-A Identifying Frequent CWEs To address RQ1, we analyze the structure of CVE information in RD. Each CVE entry comprises three elements: cwe , severity , and name . CWE provides a standardized classification of software weaknesses. Severity indicates the criticality of the vulnerability and is categorized as critical , high , medium , or low . The CVE name is a unique identifier for publicly disclosed vulnerabilities. We extract the CWE information from the CVE data and compute the frequency of each unique CWE. The frequencies are then ranked in descending order to identify the most common weaknesses. This ranking provides insights into the dominant security weaknesses in Maven libraries. II-B Analyzing Documentation Delays and Vulnerabilities We consider a vulnerability ‘documented’ when it is officially assigned a CVE identifier and its details are made publicly available in a vulnerability database, such as the NVD. The time to documentation is the period between the first appearance of a vulnerability in the release of a library and the date when the vulnerability is officially published with a CVE identifier in a vulnerability database. To determine how long vulnerabilities remain undocumented and the proportion of releases deployed with known vulnerabilities, we follow these steps: II-B 1 Data Preprocessing We convert the release timestamps in RD from Unix format to a human-readable date-time format using Python’s datetime module. II-B 2 Fetching CVE Publication Dates We extract CVE names from RD and retrieve their publication dates from the NVD through web scraping. The URL used for this process is: https://nvd.nist.gov/vuln/detail/{cve_name} , where {cve_name} is dynamically replaced with CVE names. Some CVEs in the dataset are listed as UNKNOWN , making it infeasible to fetch their publication dates. To maintain data quality, we exclude these entries, reducing the dataset from 1,411 to 1,381 libraries. Note that we exclude UNKNOWN CVEs only from the RQ2 analysis and not from other parts of this study. II-B 3 Calculating Documentation Delays For each library, we calculate the time difference between the publication date of a CVE and the release date of the library version where the CVE first appears. • Positive time differences indicate delays in documenting vulnerabilities after a release. • Negative time differences reveal releases deployed with known vulnerabilities. II-C Calculating Fixing Timelines and Unfixed Vulnerabilities To address RQ3, we analyze how long it takes

Read the complete text