Analysis

Whether researching new components or tracking open-source in production, you can scan your applications throughout your software development lifecycle (SDLC). Here we will review the Sonatype evaluation and make sure to use the best methods for your environment.

Every evaluation runs through the following process: 

1. Begin an evaluation using one of Sonatype's scanner integrations. 
   1. Recursively scan all the target files using a process called Advanced Binary Fingerprinting (ABF)
   2. Observe the dependencies declared in the package-lock and manifest files
   3. Include discovered SBOMs and modules.xml files for component identifiers
2. Scan results are submitted through the IQ Server to the Sonatype Data Services for live data and processing.
3. Component metadata and partial matches are returned to the IQ Server.
4. IQ Server applies the governance policy and waivers to the raw results.
5. Any policy violations and enforcement details are reported back to the scanner.
6. Notifications and webhooks are sent out when appropriate.

Advanced Binary Fingerprinting (ABF)

With ABF scanning, we examine binary fingerprints (similar to a truncated sha1 hash) of all of the files and not just the file names and manifests. ABF is highly accurate because it examines everything included in the application after the build, including any embedded dependencies. This means that an ABF scan will never return false positives in its report. Sonatype data is tied to the component fingerprints of any files where the vulnerability is discovered. When a vulnerability is reported it is because the component fingerprint is in your application. Depending on the environment, there are a few points where this might be confusing.

1. When the build includes development dependencies and clutter such as testing frameworks left in the source control repository. This can be corrected by rescoping the scan to only the artifacts that are deployed to production.
2. Embedded dependencies that are renamed within the application. These cannot be detected by name matching or manifest scanning but will be caught by ABF.
3. When vulnerable files are reused in other open source components with completely different names. This secondary expansion of discovered vulnerabilities is unique to Sonatype Evaluations.
4. ABF can be used to track inner-source components and apply policy to them.
5. For Java applications, the ABF scanner uses an additional unique “partial matching” technology, which is capable of identifying when a component is “similar” but not “identical” to the cataloged version. Sometimes modifications to open source are made by development teams to 'fix' the vulnerabilities within the project. One major issue with this is that it introduces technical debt and potential risk to the application when those changes are not documented. Modifying open source components will also trigger license risk for components with weak copyleft licenses.

Manifest Evaluation

Scans performed before the application has been built do not have open source dependencies to fingerprint and must therefore leverage package-lock and manifest files to scan. This happens when directly analyzing the source control repositories or software bill of materials earlier in the development lifecycle. The Lifecycle scanners can use the lock or manifest files to get an idea of what should be in the final application. This will include any transitive dependencies if they are included in the lock file. Otherwise, it will report on only what is requested in the manifest.

The Sonatype scanners default to ABF scans as they are accurate to what is in your application. They will then look for lock files followed by manifest files. As we provide feedback earlier in the development/design process, manifest scanning becomes the only option providing the earliest possible feedback during component selection. As a best practice, we recommend development teams include lock files and use fixed versions of components in manifest files. This will ensure repeatable builds and the most accurate feedback for your production application.

IQ Server Data Analysis

The Nexus IQ Server uses data derived from our automated vulnerability detection system — basically, a big funnel of sources (NVD, GitHub commits, OSS Index, Sonatype research, etc.) that is processed with automated techniques such as data filtering, aggregation, and machine learning algorithms.

Some supported ecosystems will have security, license, and identity data, while others only have security and identity data. License data includes OSS licenses identified in the package manifest, and in the case of Java, any licenses are also observed within the package itself. Identity refers to component details such as recommendations, version graphs, or cataloged data pulled from the package manager repository. We categorize these ecosystems as having either Premium or Standard data capabilities.

Premium Capabilities

For most ecosystems, Sonatype researchers triage incoming data and determine if there is a vulnerability, creating a research ticket for further investigation when necessary. Tickets are prioritized and then entered into our human-curated research process. When research is complete, it goes into our data mart which feeds Sonatype Data Services. Data from the Sonatype Data Services is what you’ll then see in the IQ Server Dashboard and Application Composition report after an application scan.

Standard Capabilities

For ecosystems with only Security/Identity data, we report any known security vulnerabilities but may not include in-depth research or license data.

Ecosystem Support

For more information, please see our guide on Understanding Sonatype Vulnerability Data.

The examples throughout this section use the IQ Server CLI for scanning. For additional scanning methods and more detailed information, please see our Comprehensive Guide to Lifecycle Scanning.