Fix Volcano instaillation guide/ Add gang scheduling configuration in /config dir / Add unit test

config/manager/controller_manager_config.yaml

@@ -4,3 +4,6 @@ leaderElection:
   leaderElect: true
 internalCertManagement:
   enable: true
+# If you want to enable gang scheduling, you need to uncomment the following section(take volcano as an example)
+# gangSchedulingManagement:
+#   schedulerProvider: volcano

config/rbac/kustomization.yaml

@@ -14,4 +14,6 @@ resources:
 - auth_proxy_role.yaml
 - auth_proxy_role_binding.yaml
 - auth_proxy_client_clusterrole.yaml
-- auth_proxy_client_binding.yaml
+- auth_proxy_client_binding.yaml
+# If you want to enable gang scheduling with volcano, you need to uncomment the following line
+# - volcano_rbac.yaml

config/rbac/volcano_rbac.yaml

@@ -0,0 +1,30 @@
+# This file provides RBAC permissions for volcano gang scheduling.
+# To enable volcano gang scheduling, uncomment this file in kustomization.yaml
+# and ensure gangSchedulingManagement.schedulerProvider is set to "volcano" in controller configuration.
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  name: lws-volcano-role
+rules:
+- apiGroups:
+  - scheduling.volcano.sh
+  resources:
+  - podgroups
+  verbs:
+  - create
+  - get
+  - list
+  - watch
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  name: lws-volcano-rolebinding
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: lws-volcano-role
+subjects:
+- kind: ServiceAccount
+  name: lws-controller-manager
+  namespace: system

docs/examples/sample/gang-scheduling/README.md

@@ -32,7 +32,22 @@ To enable gang scheduling, you must enable the feature flag and specify `volcano
 
   Refer to the [install-by-kubectl](https://lws.sigs.k8s.io/docs/installation/#install-by-kubectl). Gang scheduling is **disabled by default**. To enable gang scheduling capabilities, you need to:
 
-  1. **Update the configuration ConfigMap** to enable gang scheduling settings:
+  1. **Update RBAC permissions for the lws-controller-manager**
+     Run `kubectl edit clusterrole lws-manager-role` and add the following rule to the `rules` list. This grants the LWS controller manager the necessary permissions to manage Volcano PodGroups.
+     ```yaml
+     # Add this block to the 'rules' section
+     - apiGroups:
+       - scheduling.volcano.sh
+       resources:
+       - podgroups
+       verbs:
+       - create
+       - get
+       - list
+       - watch
+     ```
+
+  2. **Update the configuration ConfigMap** to enable gang scheduling settings:
      ```yaml
      apiVersion: v1
      kind: ConfigMap
@@ -52,7 +67,7 @@ To enable gang scheduling, you must enable the feature flag and specify `volcano
            schedulerProvider: volcano
      ```
 
-  2. **Restart the lws-controller-manager** to apply the new configuration:
+  3. **Restart the lws-controller-manager** to apply the new configuration:
      ```sh
      kubectl rollout restart deployment/lws-controller-manager -n lws-system
      ```

pkg/utils/utils_test.go

@@ -20,6 +20,11 @@ import (
 	"testing"
 
 	"github.com/google/go-cmp/cmp"
+	corev1 "k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/api/resource"
+	"k8s.io/utils/ptr"
+
+	leaderworkerset "sigs.k8s.io/lws/api/leaderworkerset/v1"
 )
 
 func Test_SortByIndex(t *testing.T) {
@@ -63,3 +68,133 @@ func Test_SortByIndex(t *testing.T) {
 		})
 	}
 }
+
+func TestCalculatePGMinResources(t *testing.T) {
+	testCases := []struct {
+		name string
+		lws  *leaderworkerset.LeaderWorkerSet
+		want corev1.ResourceList
+	}{
+		{
+			name: "leader and worker with different resources",
+			lws: &leaderworkerset.LeaderWorkerSet{
+				Spec: leaderworkerset.LeaderWorkerSetSpec{
+					LeaderWorkerTemplate: leaderworkerset.LeaderWorkerTemplate{
+						Size: ptr.To[int32](3),
+						LeaderTemplate: &corev1.PodTemplateSpec{
+							Spec: corev1.PodSpec{
+								Containers: []corev1.Container{
+									{
+										Resources: corev1.ResourceRequirements{
+											Requests: corev1.ResourceList{
+												corev1.ResourceCPU:    resource.MustParse("1"),
+												corev1.ResourceMemory: resource.MustParse("1Gi"),
+											},
+										},
+									},
+								},
+							},
+						},
+						WorkerTemplate: corev1.PodTemplateSpec{
+							Spec: corev1.PodSpec{
+								Containers: []corev1.Container{
+									{
+										Resources: corev1.ResourceRequirements{
+											Requests: corev1.ResourceList{
+												corev1.ResourceCPU:    resource.MustParse("2"),
+												corev1.ResourceMemory: resource.MustParse("2Gi"),
+											},
+										},
+									},
+								},
+							},
+						},
+					},
+				},
+			},
+			want: corev1.ResourceList{
+				corev1.ResourceCPU:    resource.MustParse("5"),   // 1 (leader) + 2 * 2 (workers)
+				corev1.ResourceMemory: resource.MustParse("5Gi"), // 1 (leader) + 2 * 2 (workers)
+			},
+		},
+		{
+			name: "only worker template specified",
+			lws: &leaderworkerset.LeaderWorkerSet{
+				Spec: leaderworkerset.LeaderWorkerSetSpec{
+					LeaderWorkerTemplate: leaderworkerset.LeaderWorkerTemplate{
+						Size: ptr.To[int32](3),
+						WorkerTemplate: corev1.PodTemplateSpec{
+							Spec: corev1.PodSpec{
+								Containers: []corev1.Container{
+									{
+										Resources: corev1.ResourceRequirements{
+											Requests: corev1.ResourceList{
+												corev1.ResourceCPU:    resource.MustParse("2"),
+												corev1.ResourceMemory: resource.MustParse("2Gi"),
+											},
+										},
+									},
+								},
+							},
+						},
+					},
+				},
+			},
+			want: corev1.ResourceList{
+				corev1.ResourceCPU:    resource.MustParse("6"),   // 2 (leader) + 2 * 2 (workers)
+				corev1.ResourceMemory: resource.MustParse("6Gi"), // 2 (leader) + 2 * 2 (workers)
+			},
+		},
+		{
+			name: "size is 1",
+			lws: &leaderworkerset.LeaderWorkerSet{
+				Spec: leaderworkerset.LeaderWorkerSetSpec{
+					LeaderWorkerTemplate: leaderworkerset.LeaderWorkerTemplate{
+						Size: ptr.To[int32](1),
+						LeaderTemplate: &corev1.PodTemplateSpec{
+							Spec: corev1.PodSpec{
+								Containers: []corev1.Container{
+									{
+										Resources: corev1.ResourceRequirements{
+											Requests: corev1.ResourceList{
+												corev1.ResourceCPU: resource.MustParse("1"),
+											},
+										},
+									},
+								},
+							},
+						},
+					},
+				},
+			},
+			want: corev1.ResourceList{
+				corev1.ResourceCPU: resource.MustParse("1"),
+			},
+		},
+		{
+			name: "no resource requests",
+			lws: &leaderworkerset.LeaderWorkerSet{
+				Spec: leaderworkerset.LeaderWorkerSetSpec{
+					LeaderWorkerTemplate: leaderworkerset.LeaderWorkerTemplate{
+						Size: ptr.To[int32](3),
+						WorkerTemplate: corev1.PodTemplateSpec{
+							Spec: corev1.PodSpec{
+								Containers: []corev1.Container{{}},
+							},
+						},
+					},
+				},
+			},
+			want: corev1.ResourceList{},
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			got := CalculatePGMinResources(tc.lws)
+			if diff := cmp.Diff(tc.want, got); diff != "" {
+				t.Errorf("CalculatePGMinResources() mismatch (-want +got):\n%s", diff)
+			}
+		})
+	}
+}