The Kubernetes community is launching a Workload Conformance Working Group. This group is tasked with drafting a shared standard detailing how workloads operate during runtime within Kubernetes. Input from both developers who build workloads and operators who manage the clusters is essential for this effort.
Current validation methods include Certified Kubernetes for cluster testing and Kubernetes AI Conformance for AI and ML workload assessment. However, no equivalent standard exists for the workloads themselves. Consequently, cluster operators must manually examine each application before routine maintenance or a Kubernetes upgrade.
The issues operators encounter are well-known. For instance, a standard node drain frequently reveals an application's inability to tolerate service interruption. Furthermore, an upgrade might cause an application failure if it relies on a deleted API version. Daily operational issues also appear, such as workloads requesting excessive privileges or reserving unused resources.
This new specification aims to provide both groups with a shared baseline using repeatable tests. The charter outlines three deliverables: a specification of workload behavior for each Kubernetes minor version; a methodology for testing each requirement against a live workload; and upstream conformance tooling to implement these tests.
The scope of coverage includes networking, storage, API usage, observability, security, and resilience during disruptions. A key constraint is that every requirement must possess an objective test executable by a test binary, precluding any subjective criteria. The specification focuses on *how* a workload executes, deliberately excluding its application logic.
Two Special Interest Groups (SIGs) are sponsoring this initiative. SIG Architecture will manage the specification and tooling via a subproject once initial artifacts are presented. SIG Apps will contribute the application-side viewpoint. The working group will engage with other SIGs when requirements involve areas like node behavior or networking.
The objective is to achieve a stable initial specification and working tests within two to three Kubernetes release cycles. Market participants are encouraged to contribute real-world operational data regarding workload assessment, upgrade failures, or current guidance gaps.
This development reflects a broader trend toward formalizing operational expectations within complex infrastructure. The establishment of this group signals a move toward standardized, verifiable runtime guarantees for workloads deployed on Kubernetes platforms.
