Add Failure Domains for Worker nodes feature for Nutanix provider

config/crd/bases/anywhere.eks.amazonaws.com_nutanixdatacenterconfigs.yaml

@@ -115,6 +115,12 @@ spec:
                         - type
                         type: object
                       type: array
+                    workerMachineGroups:
+                      description: Worker Machine Groups holds the list of worker
+                        machine group names that will use this failure domain.
+                      items:
+                        type: string
+                      type: array
                   required:
                   - name
                   type: object

config/crd/bases/anywhere.eks.amazonaws.com_nutanixmachineconfigs.yaml

@@ -74,6 +74,28 @@ spec:
                 required:
                 - type
                 type: object
+              gpus:
+                description: List of GPU devices that should be added to the VMs.
+                items:
+                  description: NutanixGPUIdentifier holds VM GPU device configuration.
+                  properties:
+                    deviceID:
+                      description: deviceID is the device ID of the GPU device.
+                      format: int64
+                      type: integer
+                    name:
+                      description: vendorID is the vendor ID of the GPU device.
+                      type: string
+                    type:
+                      description: type is the type of the GPU device.
+                      enum:
+                      - deviceID
+                      - name
+                      type: string
+                  required:
+                  - type
+                  type: object
+                type: array
               image:
                 description: image is to identify the OS image uploaded to the Prism
                   Central (PC) The image identifier (uuid or name) can be obtained

go.mod

@@ -181,7 +181,7 @@ require (
 	go.opentelemetry.io/otel/trace v1.20.0 // indirect
 	go.uber.org/multierr v1.11.0 // indirect
 	golang.org/x/mod v0.14.0 // indirect
-	golang.org/x/sync v0.6.0 // indirect
+	golang.org/x/sync v0.6.0
 	golang.org/x/sys v0.18.0 // indirect
 	golang.org/x/term v0.18.0 // indirect
 	golang.org/x/time v0.5.0 // indirect

pkg/api/v1alpha1/nutanixdatacenterconfig_types.go

@@ -68,6 +68,10 @@ type NutanixDatacenterFailureDomain struct {
 	// Subnets holds the list of subnets identifiers cluster's network subnets.
 	// +kubebuilder:validation:Required
 	Subnets []NutanixResourceIdentifier `json:"subnets,omitempty"`
+
+	// Worker Machine Groups holds the list of worker machine group names that will use this failure domain.
+	// +optional
+	WorkerMachineGroups []string `json:"workerMachineGroups,omitempty"`
 }
 
 // NutanixDatacenterConfigStatus defines the observed state of NutanixDatacenterConfig.
@@ -165,7 +169,7 @@ func (in *NutanixDatacenterConfig) Validate() error {
 		}
 	}
 
-	if in.Spec.FailureDomains != nil && len(in.Spec.FailureDomains) != 0 {
+	if len(in.Spec.FailureDomains) != 0 {
 		dccName := in.Namespace + "/" + in.Name
 		validateClusterResourceIdentifier := createValidateNutanixResourceFunc("NutanixDatacenterConfig.Spec.FailureDomains.Cluster", "cluster", dccName)
 		validateSubnetResourceIdentifier := createValidateNutanixResourceFunc("NutanixDatacenterConfig.Spec.FailureDomains.Subnets", "subnet", dccName)

pkg/providers/nutanix/config/md-template.yaml

@@ -1,3 +1,106 @@
+{{- if $.failureDomains -}}{{ range $fd := $.failureDomains -}}
+apiVersion: cluster.x-k8s.io/v1beta1
+kind: MachineDeployment
+metadata:
+  labels:
+    cluster.x-k8s.io/cluster-name: "{{$.clusterName}}"
+  name: "{{$.workerNodeGroupName}}-{{$fd.Name}}"
+  namespace: "{{$.eksaSystemNamespace}}"
+{{- if $.autoscalingConfig }}
+  annotations:
+    cluster.x-k8s.io/cluster-api-autoscaler-node-group-min-size: "{{ $.autoscalingConfig.MinCount }}"
+    cluster.x-k8s.io/cluster-api-autoscaler-node-group-max-size: "{{ $.autoscalingConfig.MaxCount }}"
+{{- end }}
+spec:
+  clusterName: "{{$.clusterName}}"
+{{- if not $.autoscalingConfig }}
+  replicas: {{ index $.failureDomainsReplicas $fd.Name }}
+{{- end }}
+  selector:
+    matchLabels: {}
+  template:
+    metadata:
+      labels:
+        cluster.x-k8s.io/cluster-name: "{{$.clusterName}}"
+    spec:
+      failureDomain: "{{$fd.Name}}"
+      bootstrap:
+        configRef:
+          apiVersion: bootstrap.cluster.x-k8s.io/v1beta1
+          kind: KubeadmConfigTemplate
+          name: "{{$.workloadkubeadmconfigTemplateName}}"
+      clusterName: "{{$.clusterName}}"
+      infrastructureRef:
+        apiVersion: infrastructure.cluster.x-k8s.io/v1beta1
+        kind: NutanixMachineTemplate
+        name: "{{$.workloadTemplateName}}-{{$fd.Name}}"
+      version: "{{$.kubernetesVersion}}"
+{{- if $.upgradeRolloutStrategy }}
+  strategy:
+    rollingUpdate:
+      maxSurge: {{$.maxSurge}}
+      maxUnavailable: {{$.maxUnavailable}}
+{{- end }}
+---
+apiVersion: infrastructure.cluster.x-k8s.io/v1beta1
+kind: NutanixMachineTemplate
+metadata:
+  name: "{{$.workloadTemplateName}}-{{$fd.Name}}"
+  namespace: "{{$.eksaSystemNamespace}}"
+spec:
+  template:
+    spec:
+      providerID: "nutanix://{{$.clusterName}}-m1"
+      vcpusPerSocket: {{$.vcpusPerSocket}}
+      vcpuSockets: {{$.vcpuSockets}}
+      memorySize: {{$.memorySize}}
+      systemDiskSize: {{$.systemDiskSize}}
+      image:
+{{- if (eq $.imageIDType "name") }}
+        type: name
+        name: "{{$.imageName}}"
+{{ else if (eq $.imageIDType "uuid") }}
+        type: uuid
+        uuid: "{{$.imageUUID}}"
+{{ end }}
+      cluster:
+{{- if (eq $fd.Cluster.Type "name") }}
+        type: name
+        name: "{{ $fd.Cluster.Name }}"
+{{- else if (eq $fd.Cluster.Type "uuid") }}
+        type: uuid
+        uuid: "{{ $fd.Cluster.UUID }}"
+{{ end }}
+      subnet:
+{{- range $subnet := $fd.Subnets }}
+{{- if (eq $subnet.Type "name") }}
+        - type: name
+          name: "{{ $subnet.Name }}"
+{{- else if (eq $subnet.Type "uuid") }}
+        - type: uuid
+          uuid: "{{ $subnet.UUID }}"
+{{- end }}
+{{- end }}
+{{- if $.projectIDType}}
+      project:
+{{- if (eq $.projectIDType "name") }}
+        type: name
+        name: "{{$.projectName}}"
+{{- else if (eq $.projectIDType "uuid") }}
+        type: uuid
+        uuid: "{{$.projectUUID}}"
+{{ end }}
+{{ end }}
+{{- if $.additionalCategories}}
+      additionalCategories:
+{{- range $.additionalCategories}}
+        - key:   "{{ $.Key }}"
+          value: "{{ $.Value }}"
+{{- end }}
+{{- end }}
+---
+{{ end -}}
+{{- else -}}
 apiVersion: cluster.x-k8s.io/v1beta1
 kind: MachineDeployment
 metadata:
@@ -107,6 +210,7 @@ spec:
 {{- end }}
 {{- end }}
 ---
+{{ end -}}
 apiVersion: bootstrap.cluster.x-k8s.io/v1beta1
 kind: KubeadmConfigTemplate
 metadata:

pkg/providers/nutanix/provider_test.go

@@ -312,13 +312,13 @@ func TestNutanixProviderSetupAndValidateCreate(t *testing.T) {
 			name:            "cluster config with unsupported upgrade strategy configuration for cp",
 			clusterConfFile: "testdata/cluster_nutanix_with_upgrade_strategy_cp.yaml",
 			expectErr:       true,
-			expectErrStr:    "failed setup and validations: Upgrade rollout strategy customization is not supported for nutanix provider",
+			expectErrStr:    "failed setup and validations: upgrade rollout strategy customization is not supported for nutanix provider",
 		},
 		{
 			name:            "cluster config with unsupported upgrade strategy configuration for md",
 			clusterConfFile: "testdata/cluster_nutanix_with_upgrade_strategy_md.yaml",
 			expectErr:       true,
-			expectErrStr:    "failed setup and validations: Upgrade rollout strategy customization is not supported for nutanix provider",
+			expectErrStr:    "failed setup and validations: upgrade rollout strategy customization is not supported for nutanix provider",
 		},
 	}
 
@@ -508,13 +508,13 @@ func TestNutanixProviderSetupAndValidateDeleteCluster(t *testing.T) {
 			name:            "cluster config with unsupported upgrade strategy configuration for cp",
 			clusterConfFile: "testdata/cluster_nutanix_with_upgrade_strategy_cp.yaml",
 			expectErr:       true,
-			expectErrStr:    "failed setup and validations: Upgrade rollout strategy customization is not supported for nutanix provider",
+			expectErrStr:    "failed setup and validations: upgrade rollout strategy customization is not supported for nutanix provider",
 		},
 		{
 			name:            "cluster config with unsupported upgrade strategy configuration for md",
 			clusterConfFile: "testdata/cluster_nutanix_with_upgrade_strategy_md.yaml",
 			expectErr:       true,
-			expectErrStr:    "failed setup and validations: Upgrade rollout strategy customization is not supported for nutanix provider",
+			expectErrStr:    "failed setup and validations: upgrade rollout strategy customization is not supported for nutanix provider",
 		},
 	}
 
@@ -560,13 +560,13 @@ func TestNutanixProviderSetupAndValidateUpgradeCluster(t *testing.T) {
 			name:            "cluster config with unsupported upgrade strategy configuration for cp",
 			clusterConfFile: "testdata/cluster_nutanix_with_upgrade_strategy_cp.yaml",
 			expectErr:       true,
-			expectErrStr:    "failed setup and validations: Upgrade rollout strategy customization is not supported for nutanix provider",
+			expectErrStr:    "failed setup and validations: upgrade rollout strategy customization is not supported for nutanix provider",
 		},
 		{
 			name:            "cluster config with unsupported upgrade strategy configuration for md",
 			clusterConfFile: "testdata/cluster_nutanix_with_upgrade_strategy_md.yaml",
 			expectErr:       true,
-			expectErrStr:    "failed setup and validations: Upgrade rollout strategy customization is not supported for nutanix provider",
+			expectErrStr:    "failed setup and validations: upgrade rollout strategy customization is not supported for nutanix provider",
 		},
 	}
 

pkg/providers/nutanix/template.go

@@ -346,10 +346,53 @@ func buildTemplateMapCP(
 	return values, nil
 }
 
+func calcFailureDomainReplicas(workerNodeGroupConfiguration v1alpha1.WorkerNodeGroupConfiguration, failureDomains []v1alpha1.NutanixDatacenterFailureDomain) map[string]int {
+	replicasPerFailureDomain := make(map[string]int)
+	failureDomainCount := len(failureDomains)
+
+	if workerNodeGroupConfiguration.AutoScalingConfiguration != nil {
+		return replicasPerFailureDomain
+	}
+
+	if failureDomainCount == 0 {
+		return replicasPerFailureDomain
+	}
+
+	workerNodeGroupCount := failureDomainCount
+	if workerNodeGroupConfiguration.Count != nil {
+		workerNodeGroupCount = int(*workerNodeGroupConfiguration.Count)
+	}
+
+	minCount := int(workerNodeGroupCount / failureDomainCount)
+
+	for i := 0; i < len(failureDomains); i++ {
+		replicasPerFailureDomain[failureDomains[i].Name] = minCount
+	}
+	replicasPerFailureDomain[failureDomains[0].Name] = workerNodeGroupCount - (failureDomainCount-1)*minCount
+
+	return replicasPerFailureDomain
+}
+
+func getFailureDomainsForWorkerNodeGroup(allFailureDomains []v1alpha1.NutanixDatacenterFailureDomain, workerNodeGroupConfigurationName string) []v1alpha1.NutanixDatacenterFailureDomain {
+	result := make([]v1alpha1.NutanixDatacenterFailureDomain, 0)
+	for _, fd := range allFailureDomains {
+		for _, workerMachineGroup := range fd.WorkerMachineGroups {
+			if workerMachineGroup == workerNodeGroupConfigurationName {
+				result = append(result, fd)
+			}
+		}
+	}
+
+	return result
+}
+
 func buildTemplateMapMD(clusterSpec *cluster.Spec, workerNodeGroupMachineSpec v1alpha1.NutanixMachineConfigSpec, workerNodeGroupConfiguration v1alpha1.WorkerNodeGroupConfiguration) (map[string]interface{}, error) {
 	versionsBundle := clusterSpec.WorkerNodeGroupVersionsBundle(workerNodeGroupConfiguration)
 	format := "cloud-config"
 
+	failureDomainsForWorkerNodeGroup := getFailureDomainsForWorkerNodeGroup(clusterSpec.NutanixDatacenter.Spec.FailureDomains, workerNodeGroupConfiguration.Name)
+	replicasPerFailureDomain := calcFailureDomainReplicas(workerNodeGroupConfiguration, failureDomainsForWorkerNodeGroup)
+
 	values := map[string]interface{}{
 		"clusterName":            clusterSpec.Cluster.Name,
 		"eksaSystemNamespace":    constants.EksaSystemNamespace,
@@ -374,6 +417,8 @@ func buildTemplateMapMD(clusterSpec *cluster.Spec, workerNodeGroupMachineSpec v1
 		"subnetUUID":             workerNodeGroupMachineSpec.Subnet.UUID,
 		"workerNodeGroupName":    fmt.Sprintf("%s-%s", clusterSpec.Cluster.Name, workerNodeGroupConfiguration.Name),
 		"workerNodeGroupTaints":  workerNodeGroupConfiguration.Taints,
+		"failureDomains":         failureDomainsForWorkerNodeGroup,
+		"failureDomainsReplicas": replicasPerFailureDomain,
 	}
 
 	if clusterSpec.Cluster.Spec.RegistryMirrorConfiguration != nil {