Kubernetes Workload Conformance Aims to Make Application Readiness Testable

Oct 27, 2026 515 views

Kubernetes has tests for cluster conformance, but operators still lack a shared way to check whether applications behave reliably on a cluster. The new Workload Conformance Working Group aims to close that gap with a standard and repeatable tests for workloads at runtime.

That could reduce repeated manual checks before maintenance and upgrades. Today, operators may discover during a node drain that an application cannot handle disruption, or learn during an upgrade that it still relies on a removed API version. Other problems include unnecessary privileges and resource requests that exceed actual use.

From cluster checks to workload checks

Certified Kubernetes evaluates the cluster itself. Kubernetes AI Conformance checks whether a cluster can run AI and machine learning workloads. Neither provides an equivalent assessment of workload behavior, leaving operators to review applications individually and maintainers to answer similar questions across environments.

Kubernetes already documents guidance on security and disruptions. The missing piece is a common specification backed by tests, so teams can use the same baseline instead of repeating the same assessment.

Objective tests define the scope

The group’s charter calls for three outputs: a workload behavior specification for each Kubernetes minor version, a method to test each requirement against a running workload, and upstream tooling that runs those tests.

The group still needs to decide which workloads and specific behaviors the specification will cover. Potential areas include security, disruption resilience, resource use, networking, storage, API usage and observability. One constraint is already clear: each requirement must have an objective test that a test binary can run. Subjective criteria won’t qualify.

The standard will assess how workloads run, not what their application logic does. It also won’t repeat checks covered by Certified Kubernetes or mandate a packaging format or vendor tool, such as a policy engine or security scanner. That focus keeps the work centered on observable workload behavior across Kubernetes environments.

SIGs will shape the specification

SIG Architecture will own the specification and tooling through a subproject, to be created after the working group proposes its first artifacts. SIG Apps will contribute the application perspective. For requirements involving other areas, such as nodes or networking, the group plans to consult the relevant Kubernetes SIGs.

The working group gives SIGs a place to shape the specification while it’s being written. Its target is a stable first specification and working tests within two to three Kubernetes release cycles. That timeline makes early review and testing important: requirements will need to hold up across real workloads and different environments.

Bring workload experience to the work

The group is looking for examples from both cluster operators and application maintainers. Operators can describe how they assess workloads today and what failed during recent upgrades or maintenance. Maintainers can explain what their applications require from the cluster and where existing guidance leaves unanswered questions.

  • Reviewers can assess proposed requirements; contributors can build tests and tooling.
  • Early testers can run the test binary against their workloads and report results that vary across runs or environments.
  • People from relevant SIGs can check whether a requirement and its test fit their area.
  • The planned subproject will need long-term maintainers after the working group’s work ends.

Join the discussion in #wg-workload-conformance on Kubernetes Slack; slack.k8s.io provides an invitation. The group also uses a mailing list and meets every other Wednesday at 15:00 UTC. The meeting notes contain the Zoom link and agenda, where participants can add topics. Joining the mailing list usually adds the meeting invite to your calendar.

The group’s proposal has already received reviews and support. Its next test is whether operators and maintainers can turn real deployment experience into requirements that are clear enough to verify consistently.

Source: Joseph Jones · kubernetes.io

Comments

Sign in to comment.
No comments yet. Be the first to comment.

Related Articles

Announcing the Workload Conformance Working Group