Merge branch 'main' into users/lingyun/instanceHistory

api/constants/constants.go

@@ -18,7 +18,8 @@ var (
 )
 
 const (
-	FullGroupName = "symphony"
+	FullGroupName       = "symphony"
+	TargetRuntimePrefix = "target-runtime"
 
 	// system annotations, reserved and should not be modified by client.
 	AzureCorrelationIdKey        = "management.azure.com/correlationId"
@@ -28,6 +29,7 @@ const (
 	AzureResourceIdKey           = "management.azure.com/resourceId"
 	AzureSystemDataKey           = "management.azure.com/systemData"
 	AzureTenantIdKey             = "management.azure.com/tenantId" // Not used
+	GuidKey                      = "Guid"
 	RunningAzureCorrelationIdKey = "management.azure.com/runningCorrelationId"
 	SummaryJobIdKey              = "SummaryJobIdKey"
 	OperationStartTimeKeyPostfix = FullGroupName + "/started-at" // instance/target

api/pkg/apis/v1alpha1/managers/jobs/jobs-manager.go

@@ -12,6 +12,7 @@ import (
 	"fmt"
 	"time"
 
+	"github.com/eclipse-symphony/symphony/api/constants"
 	"github.com/eclipse-symphony/symphony/api/pkg/apis/v1alpha1/model"
 	"github.com/eclipse-symphony/symphony/api/pkg/apis/v1alpha1/utils"
 	api_utils "github.com/eclipse-symphony/symphony/api/pkg/apis/v1alpha1/utils"
@@ -328,7 +329,7 @@ func (s *JobsManager) DelayOrSkipJob(ctx context.Context, namespace string, obje
 
 	key := "h_" + job.Id
 	if objectType == "target" {
-		key = fmt.Sprintf("h_%s-%s", "target-runtime", job.Id)
+		key = fmt.Sprintf("h_%s-%s", constants.TargetRuntimePrefix, job.Id)
 	}
 	//check if a manager is working on the job
 	var entry states.StateEntry

api/pkg/apis/v1alpha1/managers/solution/metrics/metrics.go

@@ -40,6 +40,11 @@ func New() (*Metrics, error) {
 
 // Close closes all metrics.
 func (m *Metrics) Close() {
+	if m == nil {
+		return
+	}
+
+	m.apiComponentCount.Close()
 }
 
 // ApiComponentCount gets the total count of components for an API operation.

api/pkg/apis/v1alpha1/managers/solution/solution-manager.go

@@ -171,7 +171,7 @@ func (s *SolutionManager) getPreviousState(ctx context.Context, instance string,
 	log.InfofCtx(ctx, " M (Solution): failed to get previous state for instance %s in namespace %s: %+v", instance, namespace, err)
 	return nil
 }
-func (s *SolutionManager) GetSummary(ctx context.Context, key string, namespace string) (model.SummaryResult, error) {
+func (s *SolutionManager) GetSummary(ctx context.Context, summaryId string, name string, namespace string) (model.SummaryResult, error) {
 	// lock.Lock()
 	// defer lock.Unlock()
 
@@ -182,28 +182,41 @@ func (s *SolutionManager) GetSummary(ctx context.Context, key string, namespace
 	defer observ_utils.CloseSpanWithError(span, &err)
 	defer observ_utils.EmitUserDiagnosticsLogs(ctx, &err)
 
-	log.InfofCtx(ctx, " M (Solution): get summary, key: %s, namespace: %s", key, namespace)
+	log.InfofCtx(ctx, " M (Solution): get summary, name: %s, summaryId: %s, namespace: %s", name, summaryId, namespace)
 
 	var state states.StateEntry
 	state, err = s.StateProvider.Get(ctx, states.GetRequest{
-		ID: fmt.Sprintf("%s-%s", "summary", key),
+		ID: fmt.Sprintf("%s-%s", "summary", summaryId),
 		Metadata: map[string]interface{}{
 			"namespace": namespace,
 			"group":     model.SolutionGroup,
 			"version":   "v1",
 			"resource":  Summary,
 		},
 	})
+	if err != nil && api_utils.IsNotFound(err) {
+		// if get summary by guid not found, try to get the summary by name
+		log.InfofCtx(ctx, " M (Solution): failed to get deployment summary[%s] by summaryId, error: %+v. Try to get summary by name", summaryId, err)
+		state, err = s.StateProvider.Get(ctx, states.GetRequest{
+			ID: fmt.Sprintf("%s-%s", "summary", name),
+			Metadata: map[string]interface{}{
+				"namespace": namespace,
+				"group":     model.SolutionGroup,
+				"version":   "v1",
+				"resource":  Summary,
+			},
+		})
+	}
 	if err != nil {
-		log.ErrorfCtx(ctx, " M (Solution): failed to get deployment summary[%s]: %+v", key, err)
+		log.ErrorfCtx(ctx, " M (Solution): failed to get deployment summary[%s]: %+v", summaryId, err)
 		return model.SummaryResult{}, err
 	}
 
 	var result model.SummaryResult
 	jData, _ := json.Marshal(state.Body)
 	err = json.Unmarshal(jData, &result)
 	if err != nil {
-		log.ErrorfCtx(ctx, " M (Solution): failed to deserailze deployment summary[%s]: %+v", key, err)
+		log.ErrorfCtx(ctx, " M (Solution): failed to deserailze deployment summary[%s]: %+v", summaryId, err)
 		return model.SummaryResult{}, err
 	}
 
@@ -324,8 +337,13 @@ func (s *SolutionManager) Reconcile(ctx context.Context, deployment model.Deploy
 		deploymentType = DeploymentType_Delete
 	}
 	summary.IsRemoval = remove
+	summaryId := deployment.Instance.ObjectMeta.GetSummaryId()
+	if summaryId == "" {
+		log.ErrorfCtx(ctx, " M (Solution): object GUID is null: %+v", err)
+		return summary, err
+	}
 
-	err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, deployment.Generation, deployment.Hash, summary, namespace)
+	err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, summaryId, deployment.Generation, deployment.Hash, summary, namespace)
 	if err != nil {
 		log.ErrorfCtx(ctx, " M (Solution): failed to save summary progress: %+v", err)
 		return summary, err
@@ -336,7 +354,7 @@ func (s *SolutionManager) Reconcile(ctx context.Context, deployment model.Deploy
 			if deployment.IsDryRun {
 				summary.SuccessCount = 0
 			}
-			s.concludeSummary(ctx, deployment.Instance.ObjectMeta.Name, deployment.Generation, deployment.Hash, summary, namespace)
+			s.concludeSummary(ctx, deployment.Instance.ObjectMeta.Name, summaryId, deployment.Generation, deployment.Hash, summary, namespace)
 		} else {
 			log.ErrorfCtx(ctx, " M (Solution): panic happens: %v", debug.Stack())
 			panic(r)
@@ -424,7 +442,7 @@ func (s *SolutionManager) Reconcile(ctx context.Context, deployment model.Deploy
 		summary.PlannedDeployment += len(step.Components)
 	}
 	summary.CurrentDeployed = 0
-	err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, deployment.Generation, deployment.Hash, summary, namespace)
+	err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, summaryId, deployment.Generation, deployment.Hash, summary, namespace)
 	if err != nil {
 		log.ErrorfCtx(ctx, " M (Solution): failed to save summary progress: %+v", err)
 		return summary, err
@@ -522,7 +540,7 @@ func (s *SolutionManager) Reconcile(ctx context.Context, deployment model.Deploy
 				targetResult[step.Target] = 1
 				summary.AllAssignedDeployed = plannedCount == planSuccessCount
 				summary.UpdateTargetResult(step.Target, model.TargetResultSpec{Status: "OK", Message: "", ComponentResults: componentResults})
-				err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, deployment.Generation, deployment.Hash, summary, namespace)
+				err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, summaryId, deployment.Generation, deployment.Hash, summary, namespace)
 				if err != nil {
 					log.ErrorfCtx(ctx, " M (Solution): failed to save summary progress: %+v", err)
 					return summary, err
@@ -559,7 +577,7 @@ func (s *SolutionManager) Reconcile(ctx context.Context, deployment model.Deploy
 		}
 		planSuccessCount++
 		summary.CurrentDeployed += len(step.Components)
-		err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, deployment.Generation, deployment.Hash, summary, namespace)
+		err = s.saveSummaryProgress(ctx, deployment.Instance.ObjectMeta.Name, summaryId, deployment.Generation, deployment.Hash, summary, namespace)
 		if err != nil {
 			log.ErrorfCtx(ctx, " M (Solution): failed to save summary progress: %+v", err)
 			return summary, err
@@ -636,11 +654,11 @@ func (s *SolutionManager) getTargetStateForStep(step model.DeploymentStep, deplo
 	return targetSpec
 }
 
-func (s *SolutionManager) saveSummary(ctx context.Context, objectName string, generation string, hash string, summary model.SummarySpec, state model.SummaryState, namespace string) error {
+func (s *SolutionManager) saveSummary(ctx context.Context, objectName string, summaryId string, generation string, hash string, summary model.SummarySpec, state model.SummaryState, namespace string) error {
 	// TODO: delete this state when time expires. This should probably be invoked by the vendor (via GetSummary method, for instance)
-	log.DebugfCtx(ctx, " M (Solution): saving summary, objectName: %s, state: %s, namespace: %s, jobid: %s, hash %s, targetCount %d, successCount %d",
-		objectName, state, namespace, summary.JobID, hash, summary.TargetCount, summary.SuccessCount)
-	oldSummary, err := s.GetSummary(ctx, objectName, namespace)
+	log.DebugfCtx(ctx, " M (Solution): saving summary, objectName: %s, summaryId: %s, state: %v, namespace: %s, jobid: %s, hash %s, targetCount %d, successCount %d",
+		objectName, summaryId, state, namespace, summary.JobID, hash, summary.TargetCount, summary.SuccessCount)
+	oldSummary, err := s.GetSummary(ctx, summaryId, objectName, namespace)
 	if err != nil && !v1alpha2.IsNotFound(err) {
 		log.ErrorfCtx(ctx, " M (Solution): failed to get previous summary: %+v", err)
 		return err
@@ -664,7 +682,7 @@ func (s *SolutionManager) saveSummary(ctx context.Context, objectName string, ge
 	}
 	_, err = s.StateProvider.Upsert(ctx, states.UpsertRequest{
 		Value: states.StateEntry{
-			ID: fmt.Sprintf("%s-%s", "summary", objectName),
+			ID: fmt.Sprintf("%s-%s", "summary", summaryId),
 			Body: model.SummaryResult{
 				Summary:        summary,
 				Generation:     generation,
@@ -683,12 +701,12 @@ func (s *SolutionManager) saveSummary(ctx context.Context, objectName string, ge
 	return err
 }
 
-func (s *SolutionManager) saveSummaryProgress(ctx context.Context, objectName string, generation string, hash string, summary model.SummarySpec, namespace string) error {
-	return s.saveSummary(ctx, objectName, generation, hash, summary, model.SummaryStateRunning, namespace)
+func (s *SolutionManager) saveSummaryProgress(ctx context.Context, objectName string, summaryId string, generation string, hash string, summary model.SummarySpec, namespace string) error {
+	return s.saveSummary(ctx, objectName, summaryId, generation, hash, summary, model.SummaryStateRunning, namespace)
 }
 
-func (s *SolutionManager) concludeSummary(ctx context.Context, objectName string, generation string, hash string, summary model.SummarySpec, namespace string) error {
-	return s.saveSummary(ctx, objectName, generation, hash, summary, model.SummaryStateDone, namespace)
+func (s *SolutionManager) concludeSummary(ctx context.Context, objectName string, summaryId string, generation string, hash string, summary model.SummarySpec, namespace string) error {
+	return s.saveSummary(ctx, objectName, summaryId, generation, hash, summary, model.SummaryStateDone, namespace)
 }
 
 func (s *SolutionManager) canSkipStep(ctx context.Context, step model.DeploymentStep, target string, provider tgt.ITargetProvider, previousComponents []model.ComponentSpec, currentState model.DeploymentState) bool {

api/pkg/apis/v1alpha1/managers/solution/solution-manager_test.go

@@ -313,8 +313,12 @@ func TestSortByDepedenciesAllSelfReferences(t *testing.T) {
 }
 func TestMockGet(t *testing.T) {
 	id := uuid.New().String()
+	name := "testInstance"
 	deployment := model.DeploymentSpec{
 		Instance: model.InstanceState{
+			ObjectMeta: model.ObjectMeta{
+				Name: name,
+			},
 			Spec: &model.InstanceSpec{},
 		},
 		Solution: model.SolutionState{
@@ -354,6 +358,8 @@ func TestMockGet(t *testing.T) {
 			},
 		},
 	}
+	guid := uuid.New().String()
+	deployment.Instance.ObjectMeta.SetGuid(guid)
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{})
 	stateProvider := &memorystate.MemoryStateProvider{}
@@ -372,7 +378,8 @@ func TestMockGet(t *testing.T) {
 	assert.Equal(t, 0, len(components))
 	assert.Equal(t, 0, len(state.TargetComponent))
 
-	_, err = manager.GetSummary(context.Background(), "", "default")
+	summaryKey := deployment.Instance.ObjectMeta.GetSummaryId()
+	_, err = manager.GetSummary(context.Background(), summaryKey, name, "default")
 	assert.NotNil(t, err)
 
 	_, err = manager.Reconcile(context.Background(), deployment, false, "default", "")
@@ -387,17 +394,17 @@ func TestMockGet(t *testing.T) {
 	assert.Equal(t, "mock", state.TargetComponent["a::T1"])
 	assert.Equal(t, "mock", state.TargetComponent["b::T1"])
 
-	_, err = manager.GetSummary(context.Background(), "", "default")
+	_, err = manager.GetSummary(context.Background(), summaryKey, name, "default")
 	assert.Nil(t, err)
 
 	// Test reconcile idempotency
 	_, err = manager.Reconcile(context.Background(), deployment, false, "default", "")
 	assert.Nil(t, err)
 
 	// Test summary deletion
-	err = manager.DeleteSummary(context.Background(), "", "default")
+	err = manager.DeleteSummary(context.Background(), summaryKey, "default")
 	assert.Nil(t, err)
-	_, err = manager.GetSummary(context.Background(), "", "default")
+	_, err = manager.GetSummary(context.Background(), summaryKey, name, "default")
 	assert.NotNil(t, err)
 }
 func TestMockGetTwoTargets(t *testing.T) {
@@ -464,6 +471,7 @@ func TestMockGetTwoTargets(t *testing.T) {
 			},
 		},
 	}
+	deployment.Instance.ObjectMeta.SetGuid(uuid.New().String())
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{ID: id})
 	stateProvider := &memorystate.MemoryStateProvider{}
@@ -560,6 +568,7 @@ func TestMockGetTwoTargetsTwoProviders(t *testing.T) {
 			},
 		},
 	}
+	deployment.Instance.ObjectMeta.SetGuid(uuid.New().String())
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{ID: id})
 	stateProvider := &memorystate.MemoryStateProvider{}
@@ -636,6 +645,7 @@ func TestMockApply(t *testing.T) {
 			},
 		},
 	}
+	deployment.Instance.ObjectMeta.SetGuid(uuid.New().String())
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{ID: id})
 	stateProvider := &memorystate.MemoryStateProvider{}
@@ -698,6 +708,7 @@ func TestMockApplyMultiRoles(t *testing.T) {
 			},
 		},
 	}
+	deployment.Instance.ObjectMeta.SetGuid(uuid.New().String())
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider2 := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{ID: id1})
@@ -759,6 +770,7 @@ func TestMockApplyWithUpdateAndRemove(t *testing.T) {
 			},
 		},
 	}
+	deployment.Instance.ObjectMeta.SetGuid(uuid.New().String())
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{ID: id})
 	stateProvider := &memorystate.MemoryStateProvider{}
@@ -812,6 +824,7 @@ func TestMockApplyWithError(t *testing.T) {
 			},
 		},
 	}
+	deployment.Instance.ObjectMeta.SetGuid(uuid.New().String())
 	targetProvider := &mock.MockTargetProvider{}
 	targetProvider.Init(mock.MockTargetProviderConfig{ID: id})
 	stateProvider := &memorystate.MemoryStateProvider{}

api/pkg/apis/v1alpha1/model/objectmeta.go

@@ -89,6 +89,10 @@ func (c ObjectMeta) DeepEquals(other IDeepEquals) (bool, error) {
 		return false, errors.New("parameter is not a ObjectMeta type")
 	}
 
+	if c.GetGuid() != otherC.GetGuid() {
+		return false, nil
+	}
+
 	if c.Name != otherC.Name {
 		return false, nil
 	}
@@ -150,3 +154,24 @@ func (c *ObjectMeta) PreserveSystemMetadata(metadata ObjectMeta) {
 		}
 	}
 }
+
+func (c *ObjectMeta) GetSummaryId() string {
+	if c.Annotations == nil || c.Annotations[constants.GuidKey] == "" {
+		return c.Name
+	}
+	return fmt.Sprintf("%s-%s", c.Name, c.Annotations[constants.GuidKey])
+}
+
+func (c *ObjectMeta) GetGuid() string {
+	if c.Annotations == nil {
+		return ""
+	}
+	return c.Annotations[constants.GuidKey]
+}
+
+func (c *ObjectMeta) SetGuid(guid string) {
+	if c.Annotations == nil {
+		c.Annotations = make(map[string]string)
+	}
+	c.Annotations[constants.GuidKey] = guid
+}

api/pkg/apis/v1alpha1/providers/metrics/metrics.go

@@ -83,6 +83,7 @@ func (m *Metrics) Close() {
 		return
 	}
 
+	m.providerOperationLatency.Close()
 	m.providerOperationErrors.Close()
 }