fix: populate Status.Selector in CacheEngine for worker pod discovery

.github/scripts/gha-e2e.sh

@@ -104,9 +104,14 @@
     set -e
     bash test/gha-e2e/curvine/test.sh
 }
+function cache_selector_e2e() {
+    set -e
+    bash test/gha-e2e/cacheruntime-selector/test.sh
+}
 
 check_control_plane_status
 alluxio_e2e
 jindo_e2e
 juicefs_e2e
 curvine_e2e
+cache_selector_e2e

api/v1alpha1/cacheruntime_types.go

@@ -179,6 +179,7 @@ type CacheRuntimeSpec struct {
 
 // +kubebuilder:object:root=true
 // +kubebuilder:subresource:status
+// +kubebuilder:subresource:scale:specpath=.spec.worker.replicas,statuspath=.status.worker.currentReplicas,selectorpath=.status.selector
 // +kubebuilder:printcolumn:name="Age",type="date",JSONPath=".metadata.creationTimestamp",priority=0
 // +kubebuilder:printcolumn:name="Status",type="string",JSONPath=".status.phase",priority=0
 // +kubebuilder:resource:scope=Namespaced

config/crd/bases/data.fluid.io_cacheruntimes.yaml

@@ -1656,4 +1656,8 @@ spec:
     served: true
     storage: true
     subresources:
+      scale:
+        labelSelectorPath: .status.selector
+        specReplicasPath: .spec.worker.replicas
+        statusReplicasPath: .status.worker.currentReplicas
       status: {}

pkg/ddc/cache/engine/master.go

@@ -84,8 +84,7 @@ func (e *CacheEngine) setupMasterInternal(masterValue *common.CacheRuntimeCompon
 		runtimeToUpdate := runtime.DeepCopy()
 		runtimeToUpdate.Status.Master = masterStatus
 
-		// TODO(cache runtime): figure out how to use this selector
-		// runtimeToUpdate.Status.Selector = e.getWorkerSelectors()
+		runtimeToUpdate.Status.Selector = e.getWorkerSelectors()
 
 		if len(runtimeToUpdate.Status.Conditions) == 0 {
 			runtimeToUpdate.Status.Conditions = []datav1alpha1.RuntimeCondition{}

pkg/ddc/cache/engine/status.go

@@ -149,7 +149,7 @@ func (e *CacheEngine) CheckAndUpdateRuntimeStatus(value *common.CacheRuntimeStat
 			runtimeToUpdate.Status.SetupDuration = utils.CalculateDuration(runtimeToUpdate.CreationTimestamp.Time, time.Now())
 		}
 
-		// TODO(cache runtime): set the CacheRuntime Status left fields:  Selector
+		runtimeToUpdate.Status.Selector = e.getWorkerSelectors()
 		runtimeToUpdate.Status.ValueFile = common.GetCacheRuntimeConfigConfigMapName(e.name)
 
 		if !reflect.DeepEqual(runtime.Status, runtimeToUpdate.Status) {

pkg/ddc/cache/engine/util.go

@@ -23,6 +23,7 @@ import (
 
 	"github.com/fluid-cloudnative/fluid/pkg/common"
 	"github.com/fluid-cloudnative/fluid/pkg/utils"
+	"k8s.io/apimachinery/pkg/labels"
 	"k8s.io/apimachinery/pkg/util/validation"
 )
 
@@ -129,3 +130,15 @@ func GetEmptyDirTieredStoreMountPath(levelIndex int) string {
 func getTieredStoreMountPath(levelIndex int, pathIndex int, mediumType string) string {
 	return fmt.Sprintf("/etc/fluid/mount/tiered-store/level-%d-index-%d-%s", levelIndex, pathIndex, mediumType)
 }
+
+// getWorkerSelectors returns the label selector string for CacheRuntime worker pods.
+// This is used to populate Status.Selector so that HPA and other tooling can
+// discover worker pods, matching the pattern used by other runtimes (JindoCache,
+// JuiceFS, Vineyard, EFC).
+func (e *CacheEngine) getWorkerSelectors() string {
+	workerName := common.GetCacheComponentName(e.name, common.ComponentTypeWorker)
+	return labels.SelectorFromSet(labels.Set{
+		common.LabelCacheRuntimeName:          e.name,
+		common.LabelCacheRuntimeComponentName: workerName,
+	}).String()
+}

pkg/ddc/cache/engine/util_test.go

@@ -19,6 +19,8 @@ package engine
 import (
 	"strings"
 
+	"github.com/fluid-cloudnative/fluid/pkg/common"
+	"github.com/fluid-cloudnative/fluid/pkg/utils/fake"
 	. "github.com/onsi/ginkgo/v2"
 	. "github.com/onsi/gomega"
 	"k8s.io/apimachinery/pkg/util/validation"
@@ -194,3 +196,51 @@ var _ = Describe("getSecretVolumeName Tests", Label("pkg.ddc.cache.engine.util_t
 		})
 	})
 })
+
+var _ = Describe("getWorkerSelectors Tests", Label("pkg.ddc.cache.engine.util_test.go"), func() {
+	Describe("getWorkerSelectors", func() {
+		var engine *CacheEngine
+
+		BeforeEach(func() {
+			engine = &CacheEngine{
+				name:      "test-runtime",
+				namespace: "default",
+				Log:       fake.NullLogger(),
+			}
+		})
+
+		It("returns a non-empty selector string", func() {
+			selector := engine.getWorkerSelectors()
+			Expect(selector).NotTo(BeEmpty())
+		})
+
+		It("selector contains the runtime name label", func() {
+			selector := engine.getWorkerSelectors()
+			Expect(selector).To(ContainSubstring(common.LabelCacheRuntimeName))
+			Expect(selector).To(ContainSubstring("test-runtime"))
+		})
+
+		It("selector contains the worker component name label", func() {
+			workerName := common.GetCacheComponentName("test-runtime", common.ComponentTypeWorker)
+			selector := engine.getWorkerSelectors()
+			Expect(selector).To(ContainSubstring(common.LabelCacheRuntimeComponentName))
+			Expect(selector).To(ContainSubstring(workerName))
+		})
+
+		It("returns different selectors for different runtime names", func() {
+			engine1 := &CacheEngine{name: "runtime-a", namespace: "default", Log: fake.NullLogger()}
+			engine2 := &CacheEngine{name: "runtime-b", namespace: "default", Log: fake.NullLogger()}
+			Expect(engine1.getWorkerSelectors()).NotTo(Equal(engine2.getWorkerSelectors()))
+		})
+
+		It("returns the exact expected selector string", func() {
+			// Pin the selector string so any future change to label keys, ordering,
+			// or escaping is caught immediately.
+			engine := &CacheEngine{name: "test-runtime", namespace: "default", Log: fake.NullLogger()}
+			workerName := common.GetCacheComponentName("test-runtime", common.ComponentTypeWorker)
+			expected := common.LabelCacheRuntimeComponentName + "=" + workerName +
+				"," + common.LabelCacheRuntimeName + "=test-runtime"
+			Expect(engine.getWorkerSelectors()).To(Equal(expected))
+		})
+	})
+})

test/gha-e2e/cacheruntime-selector/cacheruntime.yaml

@@ -0,0 +1,24 @@
+apiVersion: data.fluid.io/v1alpha1
+kind: CacheRuntime
+metadata:
+  name: selector-demo
+spec:
+  runtimeClassName: selector-demo
+  master:
+    # 如何区分 master 和 journal 的配置，两个在一个进程中（前缀，交给Curvine自行处理）
+    options:  # https://curvineio.github.io/zh-cn/docs/Deploy/Deploy-Curvine-Cluster/Distributed-Mode/conf#master%E9%85%8D%E7%BD%AE%E9%A1%B9
+      key1: master-value1
+    replicas: 1
+  worker:
+    options: # https://curvineio.github.io/zh-cn/docs/Deploy/Deploy-Curvine-Cluster/Distributed-Mode/conf#worker%E9%85%8D%E7%BD%AE%E9%A1%B9
+      key1: worker-value1
+    replicas: 1
+    tieredStore:
+      levels: #worker缓存配置
+        - low: "0.5"
+          high: "0.8"
+          emptyDir:
+            quota: 1Gi
+  client:
+    options:
+      key1: value1

test/gha-e2e/cacheruntime-selector/cacheruntimeclass.yaml

@@ -0,0 +1,210 @@
+apiVersion: data.fluid.io/v1alpha1
+kind: CacheRuntimeClass
+metadata:
+  name: selector-demo
+fileSystemType: curvinefs
+extraResources:
+  configMaps:
+    - name: curvine-config
+      data:
+        # TOML格式配置模板， 使用 hairyhenderson/gomplate 镜像要求的模板格式
+        cluster.toml: |
+          # master configuration
+          [master]
+          meta_dir = "testing/meta"
+
+          # masta ha raft configuration.
+          [journal]
+          journal_addrs = [
+            {{ range $index := seq 0 (sub (ds "config").master.replicas 1) }}
+            {{- if $index }},{{ end }}
+            {id = {{ add $index 1 }}, hostname = "{{(ds "config").master.name}}-{{$index}}.{{(ds "config").master.service.name}}", port = 8996}
+            {{- end }}
+          ]
+          journal_dir = "testing/journal"
+
+          # Worker configuration, translate Gi to GB
+          [worker]
+          dir_reserved = "0"
+          data_dir = [
+          {{- range $index, $level := (ds "config").worker.tieredStoreLevels }}
+          {{- range $pathIndex, $mountPath := $level.mountPaths }}
+          {{- if or $index $pathIndex }},{{ end }}
+          "[{{$level.mediumType}}:{{ index $level.quotas $pathIndex | strings.ReplaceAll "i" "B"}}]{{ $mountPath }}"
+          {{- end }}
+          {{- end }}
+          ]
+dataOperationSpecs:
+  - name: DataLoad
+    command:
+      - "/bin/bash"
+      - "-c"
+    args:
+      # Actually, the cache runtime image should use $(FLUID_RUNTIME_CONFIG_PATH) to generate the config file, and
+      # use $(FLUID_DATALOAD_DATA_PATH) to execute data load.
+      - |
+        # currently we have no customized image supporting dataload for curvine, so we write the curvine.toml with fixed journal address for test case.
+        echo -e '[journal]\njournal_addrs = [\n{id=1, hostname="selector-demo-master-0.svc-selector-demo-master"}\n]' > /etc/curvine.toml
+
+        IFS=: read -ra paths <<< "$FLUID_DATALOAD_DATA_PATH"
+        for p in "${paths[@]}"; do
+          /app/curvine/bin/cv load "$p" --watch --conf /etc/curvine.toml || {
+            echo "Error: load $p failed."
+            exit 1
+          }
+        done
+topology:
+  master:
+    service: #需要为master创建Headless Service
+      headless: { }
+    dependencies:
+      extraResources:
+        # 使用 extraResources 时，需要定义其挂载路径
+        configMaps:
+          - name: curvine-config
+            mountPath: "/templates"
+    executionEntries:
+      mountUFS:
+        command: 
+        - bash
+        - "-c"
+        - "/etc/curvine/mount/mountUfs.sh"
+        timeout: 30
+    template:
+      spec:
+        restartPolicy: Always
+        # 根据 runtime 生成的配置文件
+        initContainers:
+          - name: init-curvine
+            image: hairyhenderson/gomplate:alpine
+            # 挂载共享卷到容器内路径
+            volumeMounts:
+              - name: shared-config-volume
+                mountPath: /etc/curvine  # 配置文件存放目录
+            command: [ "gomplate" ]
+            args: [ "-d", "config=$(FLUID_RUNTIME_CONFIG_PATH)", "-f", "/templates/cluster.toml", "-o", "/etc/curvine/curvine.toml" ]
+        containers:
+          - name: master
+            image: "curvine/curvine:latest"
+            command:
+              - /entrypoint.sh
+            args:
+              - master
+              - start
+            env:
+              # /entrypoint.sh 不支持参数指定配置文件，支撑环境变量配置，默认/app/curvine/conf/curvine-cluster.toml
+              - name: CURVINE_CONF_FILE
+                value: "/etc/curvine/curvine.toml"
+              - name: POD_NAME
+                valueFrom:
+                  fieldRef:
+                    fieldPath: metadata.name
+              # curvine checks master hostname which should be one of journal_addrs
+              - name: CURVINE_MASTER_HOSTNAME
+                value: "$(POD_NAME).$(FLUID_RUNTIME_COMPONENT_SVC_NAME)"
+            volumeMounts:
+              - name: shared-config-volume
+                mountPath: /etc/curvine  # 配置文件存放目录
+              - name: curvine-mount-volume
+                mountPath: /etc/curvine/mount  # 配置文件存放目录
+            imagePullPolicy: IfNotPresent
+        volumes:
+        # emptyDir 共享存储（init容器和主容器互通）
+        - name: shared-config-volume
+          emptyDir: { }
+        - name: curvine-mount-volume
+          configMap:
+            name: curvine-mount
+            defaultMode: 0755
+  worker:
+    service:
+      headless: { } #需要为worker创建Headless Service
+    dependencies:
+      extraResources:
+        # 使用 extraResources 时，需要定义其挂载路径
+        configMaps:
+          - name: curvine-config
+            mountPath: "/templates"
+    template:
+      spec:
+        restartPolicy: Always
+        # 根据 runtime 生成的配置文件
+        initContainers:
+          - name: init-curvine
+            image: hairyhenderson/gomplate:alpine
+            # 挂载共享卷到容器内路径
+            volumeMounts:
+              - name: shared-config-volume
+                mountPath: /etc/curvine  # 配置文件存放目录
+            command: [ "gomplate" ]
+            args: [ "-d", "config=$(FLUID_RUNTIME_CONFIG_PATH)", "-f", "/templates/cluster.toml", "-o", "/etc/curvine/curvine.toml" ]
+        containers:
+          - name: worker
+            image: "curvine/curvine:latest"
+            command:
+              - /entrypoint.sh
+            args:
+              - worker
+              - start
+            env:
+              # /entrypoint.sh 不支持参数指定配置文件，支撑环境变量配置，默认/app/curvine/conf/curvine-cluster.toml
+              - name: CURVINE_CONF_FILE
+                value: "/etc/curvine/curvine.toml"
+              - name: POD_NAME
+                valueFrom:
+                  fieldRef:
+                    fieldPath: metadata.name
+              - name: CURVINE_WORKER_HOSTNAME
+                value: "$(POD_NAME).$(FLUID_RUNTIME_COMPONENT_SVC_NAME)"
+            volumeMounts:
+              - name: shared-config-volume
+                mountPath: /etc/curvine  # 配置文件存放目录
+            imagePullPolicy: IfNotPresent
+        volumes:
+          # emptyDir 共享存储（init容器和主容器互通）
+          - name: shared-config-volume
+            emptyDir: { }
+  client:
+    dependencies:
+      # Uncomment to enable secret mount for client (e.g., for JuiceFS FUSE pods)
+      # secretMount:
+      #   enabled: true
+      extraResources:
+        # 使用 extraResources 时，需要定义其挂载路径
+        configMaps:
+          - name: curvine-config
+            mountPath: "/templates"
+    template:
+      spec:
+        restartPolicy: Always
+        # 根据 runtime 生成的配置文件
+        initContainers:
+          - name: init-curvine
+            image: hairyhenderson/gomplate:alpine
+            # 挂载共享卷到容器内路径
+            volumeMounts:
+              - name: shared-config-volume
+                mountPath: /etc/curvine  # 配置文件存放目录
+            command: [ "gomplate" ]
+            args: [ "-d", "config=$(FLUID_RUNTIME_CONFIG_PATH)", "-f", "/templates/cluster.toml", "-o", "/etc/curvine/curvine.toml" ]
+        containers:
+          - name: client
+            image: "curvine/curvine:latest"
+            securityContext: #通常client需要配置privileged，用于操作fuse设备
+              privileged: true
+              runAsUser: 0
+            command:
+              - /app/curvine/lib/curvine-fuse
+            args:
+              - "--mnt-path"
+              - "$(FLUID_RUNTIME_MOUNT_PATH)"
+              - "--conf"
+              - "/etc/curvine/curvine.toml"
+            volumeMounts:
+              - name: shared-config-volume
+                mountPath: /etc/curvine  # 配置文件存放目录
+            imagePullPolicy: IfNotPresent
+        volumes:
+          # emptyDir 共享存储（init容器和主容器互通）
+          - name: shared-config-volume
+            emptyDir: { }