feat/configuraion-api-changes

docs/api-reference/operator-api.md

@@ -608,9 +608,7 @@ _Appears in:_
 
 _Underlying type:_ _string_
 
-TopologyDomain represents a predefined topology level in the hierarchy.
-Topology ordering (broadest to narrowest):
-Region > Zone > DataCenter > Block > Rack > Host > Numa
+TopologyDomain represents a level in the cluster topology hierarchy.
 
 
 
@@ -634,16 +632,19 @@ _Appears in:_
 
 
 TopologyLevel defines a single level in the topology hierarchy.
+Maps a platform-agnostic domain to a platform-specific node label key,
+allowing workload operators a consistent way to reference topology levels when defining TopologyConstraint's.
 
 
 
 _Appears in:_
+- [ClusterTopologyConfiguration](#clustertopologyconfiguration)
 - [ClusterTopologySpec](#clustertopologyspec)
 
 | Field | Description | Default | Validation |
 | --- | --- | --- | --- |
-| `domain` _[TopologyDomain](#topologydomain)_ | Domain is the predefined level identifier used in TopologyConstraint references.<br />Must be one of: region, zone, datacenter, block, rack, host, numa |  | Enum: [region zone datacenter block rack host numa] <br />Required: \{\} <br /> |
-| `key` _string_ | Key is the node label key that identifies this topology domain.<br />Must be a valid Kubernetes label key (qualified name).<br />Examples: "topology.kubernetes.io/zone", "kubernetes.io/hostname" |  | MaxLength: 63 <br />MinLength: 1 <br />Required: \{\} <br /> |
+| `domain` _[TopologyDomain](#topologydomain)_ | Domain is a platform provider-agnostic level identifier.<br />Must be one of: region, zone, datacenter, block, rack, host, numa |  | Enum: [region zone datacenter block rack host numa] <br />Required: \{\} <br /> |
+| `key` _string_ | Key is the node label key that identifies this topology domain.<br />Must be a valid Kubernetes label key (qualified name).<br />Examples: "topology.kubernetes.io/zone", "kubernetes.io/hostname" |  | MaxLength: 63 <br />MinLength: 1 <br />Pattern: `^(([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9]/)?([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9]$` <br />Required: \{\} <br /> |
 
 
 
@@ -702,7 +703,7 @@ _Appears in:_
 | Field | Description | Default | Validation |
 | --- | --- | --- | --- |
 | `enabled` _boolean_ | Enabled indicates whether topology-aware scheduling is enabled. |  |  |
-| `name` _string_ | Name is the ClusterTopology resource name to use.<br />Defaults to "grove-topology" if not specified when topology is enabled. |  |  |
+| `levels` _[TopologyLevel](#topologylevel) array_ | Levels is an ordered list of topology levels from broadest to narrowest scope.<br />Used to create/update the ClusterTopology CR at operator startup. |  |  |
 
 
 #### ControllerConfiguration

docs/designs/topology.md

@@ -172,6 +172,7 @@ Domain TopologyDomain `json:"domain"`
 // +kubebuilder:validation:Required
 // +kubebuilder:validation:MinLength=1
 // +kubebuilder:validation:MaxLength=63
+// +kubebuilder:validation:Pattern=`^(([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9]/)?([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9]$`
 Key string `json:"key"`
 }
 

hack/ld-flags.sh

@@ -39,8 +39,7 @@ function build_ld_flags() {
 
   FLAGS="-X $PACKAGE_PATH/version.gitCommit=$(git rev-parse --verify HEAD)
          -X $PACKAGE_PATH/version.gitTreeState=$tree_state
-         -X $PACKAGE_PATH/version.buildDate=$build_date
-         -X $PACKAGE_PATH/version.programName=$PROGRAM_NAME"
+         -X $PACKAGE_PATH/version.buildDate=$build_date"
 
   # The k8s.component-base/version.gitVersion can not be set to the version of grove
   # due to the error: "emulation version 1.33 is not between [1.31, 0.1.0-dev]".

operator/Makefile

@@ -86,12 +86,6 @@ cover-html: test-cover
 	@go tool cover -html=coverage.out -o coverage.html
 	@echo "Coverage report generated at coverage.html"
 
-# Run envtest tests (requires envtest binaries)
-.PHONY: test-envtest
-test-envtest: $(SETUP_ENVTEST)
-	@echo "Running envtest with CRD validation..."
-	@KUBEBUILDER_ASSETS=$$($(SETUP_ENVTEST) use -p path) go test ./internal/webhook/admission/clustertopology/validation -v -run TestClusterTopologyCRDValidation
-
 # Run e2e tests
 .PHONY: test-e2e
 test-e2e:

operator/api/common/constants/constants.go

@@ -16,6 +16,15 @@
 
 package constants
 
+const (
+	// OperatorName is the name of the Grove operator.
+	OperatorName = "grove-operator"
+	// OperatorConfigGroupName is the name of the group for Grove operator configuration.
+	OperatorConfigGroupName = "operator.config.grove.io"
+	// OperatorGroupName is the name of the group for all Grove custom resources.
+	OperatorGroupName = "grove.io"
+)
+
 // Constants for finalizers.
 const (
 	// FinalizerPodCliqueSet is the finalizer for PodCliqueSet that is added to `.metadata.finalizers[]` slice. This will be placed on all PodCliqueSet resources
@@ -107,4 +116,6 @@ const (
 	KindPodClique = "PodClique"
 	// KindPodCliqueScalingGroup is the kind for a PodCliqueScalingGroup resource.
 	KindPodCliqueScalingGroup = "PodCliqueScalingGroup"
+	// KindClusterTopology is the kind for a ClusterTopology resource.
+	KindClusterTopology = "ClusterTopology"
 )

operator/api/config/v1alpha1/defaults.go

@@ -27,7 +27,6 @@ const (
 	defaultLeaderElectionResourceLock    = "leases"
 	defaultLeaderElectionResourceName    = "grove-operator-leader-election"
 	defaultWebhookServerTLSServerCertDir = "/etc/grove-operator/webhook-certs"
-	defaultTopologyName                  = "grove-topology"
 )
 
 // SetDefaults_ClientConnectionConfiguration sets defaults for the k8s client connection.
@@ -115,10 +114,3 @@ func SetDefaults_PodCliqueScalingGroupControllerConfiguration(obj *PodCliqueScal
 		obj.ConcurrentSyncs = ptr.To(1)
 	}
 }
-
-// SetDefaults_ClusterTopologyConfiguration sets defaults for the ClusterTopologyConfiguration.
-func SetDefaults_ClusterTopologyConfiguration(obj *ClusterTopologyConfiguration) {
-	if obj.Enabled && obj.Name == "" {
-		obj.Name = defaultTopologyName
-	}
-}

operator/api/config/v1alpha1/register.go

@@ -17,16 +17,15 @@
 package v1alpha1
 
 import (
+	"github.com/ai-dynamo/grove/operator/api/common/constants"
+
 	"k8s.io/apimachinery/pkg/runtime"
 	"k8s.io/apimachinery/pkg/runtime/schema"
 )
 
-// GroupName is the group name used in this package
-const GroupName = "operator.config.grove.io"
-
 var (
 	// SchemeGroupVersion is group version used to register these objects
-	SchemeGroupVersion = schema.GroupVersion{Group: GroupName, Version: "v1alpha1"}
+	SchemeGroupVersion = schema.GroupVersion{Group: constants.OperatorConfigGroupName, Version: "v1alpha1"}
 	// SchemeBuilder is used to add go types to the GroupVersionKind scheme
 	SchemeBuilder = runtime.NewSchemeBuilder(addKnownTypes, addDefaultingFuncs)
 	// AddToScheme is a reference to the Scheme Builder's AddToScheme function.

operator/api/config/v1alpha1/types.go

@@ -17,6 +17,8 @@
 package v1alpha1
 
 import (
+	corev1alpha1 "github.com/ai-dynamo/grove/operator/api/core/v1alpha1"
+
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 )
 
@@ -55,14 +57,14 @@ var (
 type OperatorConfiguration struct {
 	metav1.TypeMeta  `json:",inline"`
 	ClientConnection ClientConnectionConfiguration `json:"runtimeClientConnection"`
-	LeaderElection   LeaderElectionConfiguration
-	Server           ServerConfiguration          `json:"server"`
-	Debugging        *DebuggingConfiguration      `json:"debugging,omitempty"`
-	Controllers      ControllerConfiguration      `json:"controllers"`
-	LogLevel         LogLevel                     `json:"logLevel"`
-	LogFormat        LogFormat                    `json:"logFormat"`
-	Authorizer       AuthorizerConfig             `json:"authorizer"`
-	ClusterTopology  ClusterTopologyConfiguration `json:"clusterTopology"`
+	LeaderElection   LeaderElectionConfiguration   `json:"leaderElection"`
+	Server           ServerConfiguration           `json:"server"`
+	Debugging        *DebuggingConfiguration       `json:"debugging,omitempty"`
+	Controllers      ControllerConfiguration       `json:"controllers"`
+	LogLevel         LogLevel                      `json:"logLevel"`
+	LogFormat        LogFormat                     `json:"logFormat"`
+	Authorizer       AuthorizerConfig              `json:"authorizer"`
+	ClusterTopology  ClusterTopologyConfiguration  `json:"clusterTopology"`
 }
 
 // LeaderElectionConfiguration defines the configuration for the leader election.
@@ -193,8 +195,8 @@ type AuthorizerConfig struct {
 type ClusterTopologyConfiguration struct {
 	// Enabled indicates whether topology-aware scheduling is enabled.
 	Enabled bool `json:"enabled"`
-	// Name is the ClusterTopology resource name to use.
-	// Defaults to "grove-topology" if not specified when topology is enabled.
+	// Levels is an ordered list of topology levels from broadest to narrowest scope.
+	// Used to create/update the ClusterTopology CR at operator startup.
 	// +optional
-	Name string `json:"name,omitempty"`
+	Levels []corev1alpha1.TopologyLevel `json:"levels,omitempty"`
 }

operator/api/config/validation/validation.go

@@ -17,9 +17,12 @@
 package validation
 
 import (
+	"fmt"
+	"slices"
 	"strings"
 
 	configv1alpha1 "github.com/ai-dynamo/grove/operator/api/config/v1alpha1"
+	corev1alpha1 "github.com/ai-dynamo/grove/operator/api/core/v1alpha1"
 
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
 	"k8s.io/apimachinery/pkg/util/sets"
@@ -114,11 +117,51 @@ func mustBeGreaterThanZeroDuration(duration metav1.Duration, fldPath *field.Path
 }
 
 // validateClusterTopologyConfiguration validates the cluster topology configuration.
-// When cluster topology is enabled, it ensures the topology name is provided.
+// When cluster topology is enabled, it ensures the topology name and levels are provided,
+// and validates domain and key uniqueness.
 func validateClusterTopologyConfiguration(clusterTopologyCfg configv1alpha1.ClusterTopologyConfiguration, fldPath *field.Path) field.ErrorList {
 	allErrs := field.ErrorList{}
-	if clusterTopologyCfg.Enabled && len(strings.TrimSpace(clusterTopologyCfg.Name)) == 0 {
-		allErrs = append(allErrs, field.Required(fldPath.Child("name"), "clusterTopology name is required"))
+	if !clusterTopologyCfg.Enabled {
+		return allErrs
+	}
+	allErrs = validateClusterTopologyLevels(clusterTopologyCfg.Levels, fldPath.Child("levels"))
+	return allErrs
+}
+
+func validateClusterTopologyLevels(levels []corev1alpha1.TopologyLevel, fldPath *field.Path) field.ErrorList {
+	allErrs := field.ErrorList{}
+	if len(levels) == 0 {
+		allErrs = append(allErrs, field.Required(fldPath, "levels are required when topology is enabled"))
+	}
+	allErrs = append(allErrs, mustHaveSupportedTopologyDomains(levels, fldPath)...)
+	allErrs = append(allErrs, mustHaveUniqueTopologyLevels(levels, fldPath)...)
+	return allErrs
+}
+
+func mustHaveSupportedTopologyDomains(levels []corev1alpha1.TopologyLevel, fldPath *field.Path) field.ErrorList {
+	allErrs := field.ErrorList{}
+	supportedDomains := corev1alpha1.SupportedTopologyDomains()
+	for i, level := range levels {
+		if !slices.Contains(supportedDomains, level.Domain) {
+			allErrs = append(allErrs, field.Invalid(fldPath.Index(i).Child("domain"), level.Domain, fmt.Sprintf("must be one of %v", supportedDomains)))
+		}
+	}
+	return allErrs
+}
+
+func mustHaveUniqueTopologyLevels(levels []corev1alpha1.TopologyLevel, fldPath *field.Path) field.ErrorList {
+	allErrs := field.ErrorList{}
+	seenDomains := make(map[corev1alpha1.TopologyDomain]struct{})
+	seenKeys := make(map[string]struct{})
+	for i, level := range levels {
+		if _, exists := seenDomains[level.Domain]; exists {
+			allErrs = append(allErrs, field.Duplicate(fldPath.Index(i).Child("domain"), level.Domain))
+		}
+		if _, exists := seenKeys[level.Key]; exists {
+			allErrs = append(allErrs, field.Duplicate(fldPath.Index(i).Child("key"), level.Key))
+		}
+		seenDomains[level.Domain] = struct{}{}
+		seenKeys[level.Key] = struct{}{}
 	}
 	return allErrs
 }