Kubeless

Kubeless let developers run small pieces of application code in response to a web request or an incoming event, without first building a separate container image for each function. It supplied the surrounding runtime and Kubernetes machinery, making it useful for teams that wanted functions alongside their other cluster applications. This entry covers the retired Kubeless project; its runtime and trigger behavior matter when moving existing functions to another deployment model.

Current guidance

Its archived repository explicitly says VMware stopped active maintenance and response to issues and pull requests. The historical compatibility claims and language list are not a current security-support matrix for those bundled runtime images.

Export function definitions, source, dependency manifests, runtime versions, environment configuration and trigger resources before removing the controller. Inventory how callers authenticate, how events are acknowledged or retried, and whether functions rely on writable local storage or runtime-provided libraries. Build a reproducible container or another explicit deployment artifact rather than depending on an abandoned controller to assemble code at startup.

Choose a destination according to each function’s behavior: a request-serving workload, scheduled job and event consumer may require different designs. Test payload formats, response codes, timeouts, concurrency and retry-related duplicate effects. Keep consumers from competing for the same messages during a cutover unless that is intentional. No officially equivalent successor was established by the reviewed source; adopting another serverless platform requires its own trigger and operational compatibility review.