An archived open-source project can stay in production for years after its repository stops receiving fixes. Kubeless is a Kubernetes-native serverless framework now maintained under the VMware archive. Nine pages named Kubeless in ZoomEye do not prove that nine clusters still run it, but they justify a review.

On 2026-11-22 at 00:45 UTC, ZoomEye returned nine matches for title="Kubeless" and zero for app="Kubeless" at sub_type=all, page 1, and pagesize=1. The results are index matches, not a vulnerability count.

Open the title query in ZoomEye. The repository is archived, so a live deployment may have missed security fixes that arrived after maintenance ended. The count also needs semantic review because a page can document the project without running it.

Archived code is an ownership problem

The Kubeless repository is marked as archived and read-only. Its documentation still describes a Kubernetes-native framework that deploys small functions and provides API routing, monitoring, and troubleshooting. Those features can be useful enough that an old installation survives a platform migration freeze.

The security review should start with ownership. Which team deployed Kubeless, which cluster still contains its custom resources, and who receives alerts from the function routes? A team that cannot answer those questions has an inventory gap even if the public page count is low.

The archived project also changes remediation expectations. A new vulnerability may not receive an upstream fix. Operators need a migration plan rather than a recurring promise to wait for a patch. Until migration, network isolation, authentication at the ingress, and least-privilege service accounts become the main controls.

Separate a documentation page from a runtime

For each title match, check for live function routes, Kubernetes API references, custom resource names, or a UI that matches the project. Documentation pages and tutorials should be classified separately from deployments.

For authorized clusters, search for Kubeless custom resource definitions, controller deployments, function services, and ingress rules. Record the namespace, service account, image digest, and exposed paths. Check whether the runtime uses an old image that no longer receives registry security updates.

Test one function route with a harmless request. Determine whether the route is public by design or protected by the ingress. Review the Kubernetes service account used by the function and the secrets mounted into it. The platform's age increases the importance of these compensating controls.

Make the migration measurable

Track the number of Kubeless namespaces, functions, routes, and active owners. Set a date to remove the controller and migrate functions to a supported platform. A public-page count can show whether documentation and deployment references are still discoverable, but the internal inventory is the authoritative measure.

If a deployment is retired, remove DNS records, ingress rules, controller service accounts, and stored secrets. Do not leave the function images in a registry without an owner. An archived platform deserves an explicit end-of-life record, not an unexplained collection of running resources.

What the nine matches can mean

The count may contain project pages, old tutorials, or live installations. Its small size makes manual validation practical. The important output is a named owner and a migration state for each confirmed deployment.

ZoomEye gives a reproducible external lead. The operator still has to inspect Kubernetes, the ingress, and the deployment history. That is the correct division of evidence for an archived platform: the internet tells you where to look, but only the cluster inventory tells you whether a function still runs and what authority it holds.

References