Stackable Docs Hub

Stackable

Stackable

Does your data platform need an SBOM? – supply-chain sovereignty with SBOM and VEX

Software supply chains have become a primary attack surface for modern infrastructure, and data platforms are not exempt. When your data stack runs dozens of open-source components, each with its own dependency tree, knowing exactly what is running, and whether any of it is vulnerable, is no longer optional. It is a prerequisite for operating responsibly. The concepts of Software Bill of Materials (SBOM) and Vulnerability Exploitability eXchange (VEX) have moved from niche security tooling into mainstream compliance conversations, accelerated by frameworks like the Cyber Resilience Act (CRA) and the NIS-2 Directive. We look at what these standards actually mean for data platform teams, and how the Stackable Data Platform (SDP) fits into that picture, specifically at the end.

The software supply chain risk hiding in your data stack

A modern data platform is not a single application. It is a composition of many open-source projects, each of which pulls in its own libraries, runtimes, and system dependencies. Apache Kafka®, Apache Spark™, Trino, and their surrounding tooling collectively represent hundreds of transitive dependencies. Any one of those can carry a known vulnerability, and without visibility into the full dependency graph, you are effectively operating blind.

The risk is compounded by the pace of change. Open-source projects release frequently, Common Vulnerabilities and Exposures (CVE) disclosures arrive continuously, and the gap between a vulnerability being published and a patch being deployed is where incidents happen. Supply chain attacks, where malicious code is introduced upstream and inherited downstream, have demonstrated that trusting a package name is not the same as trusting its contents. For data platform teams handling sensitive financial, healthcare, or operational data, that distinction matters enormously.

What an SBOM actually contains, and what it doesn’t

An SBOM is a structured, machine-readable inventory of every software component in a given artifact. Think of it as a nutritional label for software: it lists what is inside, its version, and its licence. Common formats include SPDX and CycloneDX, both of which are widely supported by tooling in the Kubernetes and container ecosystem.

A well-formed SBOM typically includes component names and versions, package identifiers such as CPE or PURL, licence information, and the relationships between components. What it does not include is any judgment about whether a given component is actually exploitable in your specific deployment context. That is where many teams run into trouble: an SBOM surfaces a component with a known CVE, but the CVE may affect a code path that your application never exercises. Without additional context, every flagged dependency looks equally urgent, which leads to alert fatigue and wasted remediation effort.

Where VEX fits into the vulnerability picture

Vulnerability Exploitability eXchange (VEX) is the mechanism that adds that missing context. A VEX document is issued by a software supplier or maintainer and states, for a specific component and a specific CVE, whether the vulnerability is exploitable in the product as shipped. The possible statuses are not affected, affected, fixed, or under investigation.

This matters practically. When a CVE is published against a library you use, your security team needs to know whether to act immediately, schedule a patch, or safely deprioritize. A VEX statement from the component maintainer, or from the platform vendor, provides that signal without requiring your team to reverse-engineer the code path themselves. The Common Security Advisory Framework (CSAF) is the underlying standard that structures how VEX documents are published and consumed, and it is increasingly referenced in European regulatory guidance under the CRA and NIS-2 Directive.

Together, SBOM and VEX form a feedback loop: the SBOM tells you what is present, and VEX tells you what is actually dangerous. Neither is sufficient alone.

How SBOMs support data sovereignty and regulatory compliance

Data sovereignty is not only about where data is stored. It extends to understanding and controlling the software that processes that data. An SBOM is a foundational artifact for that control: it makes the software supply chain traceable, auditable, and verifiable by parties other than the original vendor.

From a regulatory perspective, the Cyber Resilience Act (CRA), which applies to products with digital elements sold in the EU, explicitly references SBOMs as part of the documentation requirements for software manufacturers. The Digital Operational Resilience Act (DORA), which targets financial entities, similarly emphasises ICT supply chain risk management. The NIS-2 Directive extends supply chain security obligations to a broader set of critical sectors. In each case, the underlying expectation is the same: organisations must be able to demonstrate that they know what software they are running and that they have a process for responding to vulnerabilities in it.

For teams running open-source data platforms, this is actually an advantage. Open-source components can be inspected, their SBOMs can be generated from source, and the dependency graph is not hidden behind a proprietary build process. That transparency is a structural benefit compared to black-box commercial software, where you depend entirely on the vendor to disclose what is inside. Supply-chain Levels for Software Artifacts (SLSA, pronounced „salsa“) provides a framework for reasoning about the integrity of that build process, and it pairs naturally with SBOM generation in CI/CD pipelines.

Generating and consuming SBOMs in a Kubernetes-native platform

In a Kubernetes-native environment, SBOMs are typically generated at the container image level. Tools like Syft, Trivy, and Grype can scan OCI images and produce SPDX or CycloneDX output. These can be integrated into build pipelines so that every image produced carries an attached SBOM, either as a separate artifact or as an OCI attestation stored alongside the image in a registry.

Consuming SBOMs at scale requires a policy layer. Raw SBOM output for a complex data platform can run to thousands of components. The practical workflow involves ingesting SBOMs into a vulnerability management tool, correlating them against CVE feeds, and then applying VEX statements to filter out non-exploitable findings. Tools like Grype and Dependency-Track support this workflow, and the CSAF standard enables automated ingestion of VEX documents from suppliers who publish them.

For Kubernetes operators specifically, the operator pattern introduces an additional layer worth considering. An operator manages the lifecycle of a data application, including its configuration and upgrades. If the operator itself carries an SBOM, and if VEX statements are published for the operator’s dependencies, then the security posture of the managed application becomes significantly more transparent. This is the model that aligns with a fully traceable software supply chain for data infrastructure.

Operationalising SBOM and VEX across your data platform team

Generating an SBOM once is not a security practice. It is a starting point. Operationalising SBOM and VEX means embedding them into the regular workflows of platform engineers, security teams, and release processes.

A practical starting point is to define SBOM generation as a required output of every container build in your CI/CD pipeline. Store SBOMs as OCI attestations so they travel with the image and are available at deployment time. Establish a process for ingesting VEX documents from your upstream suppliers, and configure your vulnerability management tooling to apply them automatically before escalating findings to engineers.

Cross-team coordination is the harder problem. Security teams need to understand which CVEs are relevant to which platform components. Platform engineers need to understand the remediation options, whether that is upgrading a dependency, applying a configuration workaround, or accepting risk with documented justification. Data engineers need visibility into whether the tools they depend on are in a known-good state. A shared SBOM and VEX workflow creates a common language for that conversation, replacing ad hoc email threads with structured, auditable records.

For teams operating under regulatory frameworks, that auditability is not a nice-to-have. It is the evidence that demonstrates due diligence to auditors and regulators. Building the habit now, before an audit or an incident, is considerably less painful than reconstructing it after the fact.

How Stackable supports software supply chain security

The SDP is built on the principle that the software supply chain should be fully traceable. Every component in the platform is open source, which means the dependency graph is inspectable rather than hidden. Stackable publishes SBOMs for its operators and container images, and we are actively working on VEX publication through CSAF-compatible formats so that teams consuming the SDP can automate their vulnerability triage rather than manually reviewing every CVE disclosure.

Concretely, here is what that means for teams running the SDP:

  • SBOMs are generated and published for Stackable operator images, covering components like the Stackable Operator for Apache Kafka® and others across the platform.
  • The Kubernetes-native architecture means SBOM attestations can be stored and verified using standard OCI tooling, without proprietary lock-in.
  • Open-source licensing means security teams can inspect, reproduce, and validate the build process, supporting SLSA compliance goals.
  • The modular design of the SDP allows teams to scope their SBOM and vulnerability management effort to the components they actually deploy, rather than managing a monolithic platform inventory.
  • For regulated industries, the combination of open-source transparency and structured SBOM output provides concrete evidence for CRA, DORA, and NIS-2 Directive compliance requirements.

If supply chain security and data sovereignty are active concerns for your team, we are happy to walk through how the SDP fits your specific compliance context. Get in touch with the Stackable team to start that conversation.

Ähnliche Artikel

Comments are closed.