Workflow

View as Markdown

Workflow manages a single certification category run. It is created by the Certification controller — one per category — and is not typically created directly by users.

Status fields

FieldTypeDescription
conditions[]ConditionExclusive set: InProgress, Succeeded, Failed. Independent (additive): ValidationFailed — aggregates Job validation failures; set True alongside Failed whenever any Job failed performance threshold validation, even when hardware failures determine the Failed reason. When the threshold miss is the only cause, Failed carries reason JobValidationFailed
namespacestringResolved namespace where Jobs and dependencies are created
succeededNodesRefTypedLocalObjectReferenceConfigMap reference for the succeeded-nodes list
failedNodesRefTypedLocalObjectReferenceConfigMap reference for the failed-nodes list
orchestration.completedIterationsintNumber of fully completed iterations
orchestration.currentIterationintThe iteration currently in progress (1-based)
orchestration.totalNodesintTotal nodes discovered from the target
orchestration.nodesPerJobintNodes per job (auto-detected from workload template)
orchestration.detectedGPUArchitecturestringDetected GPU architecture (e.g. gb200)
orchestration.detectedPlatformstringDetected cloud platform (e.g. aws)
orchestration.appliedOverrides[]AppliedOverrideWhich spec overrides matched and were applied
orchestration.groups[]GroupStatusPer-group job status for the current iteration

Naming

Workflows are named <certificationName>-<domain>-<variant>.

Lifecycle

  1. Pulls WorkflowSpec from the catalog for the assigned {domain, variant}.
  2. Detects platform and GPU architecture; applies matching overrides.
  3. Creates a child Job.
  4. Manages iteration count; marks Succeeded or Failed when complete.