Runtime Options

Beyond the YAML configuration file, Conformance EE provides additional runtime options.

Interactive Mode (CLI)

The conformance-tester CLI is the recommended way to run Conformance EE. After downloading the binary via kubermatic-ee-downloader, launch it:

./conformance-tester

The interactive TUI guides you through all runtime decisions:

StepDescription
EnvironmentChoose between deploying locally or to an existing cluster
KubeconfigSelect a kubeconfig source (from KUBECONFIG env, default path, or custom file)
ProviderChoose the cloud provider to test
Kubernetes VersionsSelect which Kubernetes versions to include
OS DistributionsPick OS distributions and image sources
DatacentersChoose target datacenters
Cluster SettingsConfigure CNI, proxy mode, expose strategy, and other cluster modifiers
Machine SettingsConfigure CPU, memory, disk, and other machine modifiers
Review & DeployReview the generated test matrix and deploy

All selections made in the TUI are translated into a configuration file and deployed automatically.

Environment Variables

VariableDescription
CONFORMANCE_TESTER_CONFIG_FILEPath to the YAML configuration file (used in automated/in-cluster mode)
KUBECONFIGPath to the kubeconfig file (the TUI auto-detects this)
HIDE_PRESET_CREDENTIALSSet to true to mask credential values in the TUI

In-Cluster Flags

When running Conformance EE as a Kubernetes Job (see In-Cluster Deployment), additional flags can be passed in the Job spec’s args field to control test execution:

FlagDefaultDescription
--datacenters""Comma-separated list of datacenters to test
--kube-versions""Comma-separated Kubernetes versions to test
--skip-cluster-creationfalseSkip cluster creation; use existing clusters
--skip-cluster-deletionfalseKeep clusters after tests (useful for debugging)
--update-clustersfalseUpgrade existing clusters before running tests
--verbose-logsfalseShow additional log output from test code

Examples

Test specific datacenters:

args:
  - --datacenters=dc-1,dc-2

Keep clusters alive for debugging:

args:
  - --skip-cluster-deletion

Reuse existing clusters from a previous run:

args:
  - --skip-cluster-creation

Upgrade existing clusters before testing:

args:
  - --update-clusters
  - --skip-cluster-creation