bugfix: fix juicefsruntime stucks with "worker not ready" (#5103)

* bugfix: fix juicefsruntime stucks with "worker not ready" and no worker pod created

Signed-off-by: trafalgarzzz <trafalgarz@outlook.com>

* fix unit tests

Signed-off-by: trafalgarzzz <trafalgarz@outlook.com>

---------

Signed-off-by: trafalgarzzz <trafalgarz@outlook.com>

Author: TrafalgarZZZ

pkg/ddc/alluxio/deprecated_label_test.go

@@ -53,7 +53,7 @@ func getTestAlluxioEngine(client client.Client, name string, namespace string) *
 //   - name (string): The name of the resource.
 //   - suffix (string): The suffix of the label, which is not used in this test case.
 //
-// This test checks various combinations of `namespace` and `name`, 
+// This test checks various combinations of `namespace` and `name`,
 // and validates whether the generated label name matches the expected output, ensuring the function logic is correct.
 func TestAlluxioEngine_GetDeprecatedCommonLabelname(t *testing.T) {
 	testCases := []struct {

pkg/ddc/juicefs/master.go

@@ -22,19 +22,20 @@ import (
 
 	"github.com/fluid-cloudnative/fluid/pkg/utils/kubeclient"
 
-	corev1 "k8s.io/api/core/v1"
 	apierrs "k8s.io/apimachinery/pkg/api/errors"
 	"k8s.io/client-go/util/retry"
 
 	datav1alpha1 "github.com/fluid-cloudnative/fluid/api/v1alpha1"
-	"github.com/fluid-cloudnative/fluid/pkg/utils"
 )
 
 func (j JuiceFSEngine) CheckMasterReady() (ready bool, err error) {
 	// JuiceFS Runtime has no master role
 	return true, nil
 }
 
+// ShouldSetupMaster checks if a further call of func `SetupMaster` is needed.
+// JuiceFS Runtime has no master role, so the function check runtime.status.WorkerPhase
+// to know if juicefs is installed and set up.
 func (j JuiceFSEngine) ShouldSetupMaster() (should bool, err error) {
 	runtime, err := j.getRuntime()
 	if err != nil {
@@ -50,6 +51,9 @@ func (j JuiceFSEngine) ShouldSetupMaster() (should bool, err error) {
 	return
 }
 
+// SetupMaster installs juicefs components into the cluster.
+// JuiceFS Runtime has no master role, implementing func `SetupMaster` here
+// is just for a same lifecycle as other runtimes (other runtimes may have master component)
 func (j JuiceFSEngine) SetupMaster() (err error) {
 	workerName := j.getWorkerName()
 
@@ -58,7 +62,7 @@ func (j JuiceFSEngine) SetupMaster() (err error) {
 	if err != nil && apierrs.IsNotFound(err) {
 		//1. Is not found error
 		j.Log.V(1).Info("SetupMaster", "worker", workerName)
-		return j.setupMasterInternal()
+		return j.installJuiceFS()
 	} else if err != nil {
 		//2. Other errors
 		return
@@ -75,26 +79,10 @@ func (j JuiceFSEngine) SetupMaster() (err error) {
 		}
 		runtimeToUpdate := runtime.DeepCopy()
 
-		runtimeToUpdate.Status.WorkerPhase = datav1alpha1.RuntimePhaseNotReady
-		replicas := runtimeToUpdate.Spec.Worker.Replicas
-		if replicas == 0 {
-			replicas = 1
-		}
-
 		// Init selector for worker
 		runtimeToUpdate.Status.Selector = j.getWorkerSelectors()
-		runtimeToUpdate.Status.DesiredWorkerNumberScheduled = replicas
 		runtimeToUpdate.Status.ValueFileConfigmap = j.getHelmValuesConfigMapName()
 
-		if len(runtimeToUpdate.Status.Conditions) == 0 {
-			runtimeToUpdate.Status.Conditions = []datav1alpha1.RuntimeCondition{}
-		}
-		cond := utils.NewRuntimeCondition(datav1alpha1.RuntimeWorkersInitialized, datav1alpha1.RuntimeWorkersInitializedReason,
-			"The worker is initialized.", corev1.ConditionTrue)
-		runtimeToUpdate.Status.Conditions =
-			utils.UpdateRuntimeCondition(runtimeToUpdate.Status.Conditions,
-				cond)
-
 		if !reflect.DeepEqual(runtime.Status, runtimeToUpdate.Status) {
 			return j.Client.Status().Update(context.TODO(), runtimeToUpdate)
 		}

pkg/ddc/juicefs/master_internal.go

@@ -29,8 +29,7 @@ import (
 	"github.com/fluid-cloudnative/fluid/pkg/utils/kubeclient"
 )
 
-// setup fuse
-func (j *JuiceFSEngine) setupMasterInternal() (err error) {
+func (j *JuiceFSEngine) installJuiceFS() (err error) {
 	var (
 		chartName = utils.GetChartsDirectory() + "/" + common.JuiceFSChart
 	)

pkg/ddc/juicefs/master_internal_test.go

@@ -153,7 +153,7 @@ func TestSetupMasterInternal(t *testing.T) {
 	if err != nil {
 		t.Fatal(err.Error())
 	}
-	err = engine.setupMasterInternal()
+	err = engine.installJuiceFS()
 	if err != nil {
 		t.Errorf("fail to exec check helm release: %v", err)
 	}
@@ -164,7 +164,7 @@ func TestSetupMasterInternal(t *testing.T) {
 	if err != nil {
 		t.Fatal(err.Error())
 	}
-	err = engine.setupMasterInternal()
+	err = engine.installJuiceFS()
 	if err == nil {
 		t.Errorf("fail to catch the error: %v", err)
 	}
@@ -181,7 +181,7 @@ func TestSetupMasterInternal(t *testing.T) {
 	if err != nil {
 		t.Fatal(err.Error())
 	}
-	err = engine.setupMasterInternal()
+	err = engine.installJuiceFS()
 	if err == nil {
 		t.Errorf("fail to catch the error")
 	}
@@ -192,7 +192,7 @@ func TestSetupMasterInternal(t *testing.T) {
 	if err != nil {
 		t.Fatal(err.Error())
 	}
-	err = engine.setupMasterInternal()
+	err = engine.installJuiceFS()
 	fmt.Println(err)
 	if err != nil {
 		t.Errorf("fail to install release")

pkg/ddc/juicefs/master_test.go

@@ -147,15 +147,6 @@ func TestJuiceFSEngine_SetupMaster(t *testing.T) {
 			t.Errorf("fail to exec the func with error %v", err)
 			return
 		}
-		juicefsruntime, err := test.engine.getRuntime()
-		if err != nil {
-			t.Errorf("fail to get the runtime")
-			return
-		}
-		if juicefsruntime.Status.WorkerPhase == datav1alpha1.RuntimePhaseNone {
-			t.Errorf("fail to update the runtime")
-			return
-		}
 	}
 }