fix project metadata request (#3893)

orouz · mergify[bot] · commit 1cb9842b57a2 · 2026-02-11T16:12:02.000Z
(cherry picked from commit 10e1725) # Conflicts: # go.mod
diff --git a/go.mod b/go.mod
@@ -532,13 +532,23 @@ require (
 	go.opentelemetry.io/otel/sdk v1.38.0
 	go.opentelemetry.io/otel/trace v1.39.0
 	go.uber.org/multierr v1.11.0 // indirect
+<<<<<<< HEAD
 	golang.org/x/crypto v0.45.0 // indirect
 	golang.org/x/mod v0.29.0 // indirect
 	golang.org/x/net v0.47.0 // indirect
 	golang.org/x/sync v0.18.0 // indirect
 	golang.org/x/sys v0.38.0 // indirect
 	golang.org/x/term v0.37.0 // indirect
 	golang.org/x/text v0.31.0 // indirect
+=======
+	golang.org/x/crypto v0.47.0 // indirect
+	golang.org/x/mod v0.31.0 // indirect
+	golang.org/x/net v0.49.0 // indirect
+	golang.org/x/sync v0.19.0
+	golang.org/x/sys v0.40.0 // indirect
+	golang.org/x/term v0.39.0 // indirect
+	golang.org/x/text v0.33.0 // indirect
+>>>>>>> 10e17250 (fix project metadata request (#3893))
 	golang.org/x/time v0.14.0
 	golang.org/x/xerrors v0.0.0-20240903120638-7835f813f4da // indirect
 	google.golang.org/genproto v0.0.0-20250922171735-9219d122eba9 // indirect
diff --git a/internal/resources/fetching/fetchers/gcp/assets_fetcher.go b/internal/resources/fetching/fetchers/gcp/assets_fetcher.go
@@ -81,7 +81,6 @@ func NewGcpAssetsFetcher(_ context.Context, log *clog.Logger, ch chan fetching.R
 func (f *GcpAssetsFetcher) Fetch(ctx context.Context, cycleMetadata cycle.Metadata) error {
 	f.log.Info("GcpAssetsFetcher.Fetch start")
 	defer f.log.Info("GcpAssetsFetcher.Fetch done")
-	defer f.provider.Clear()
 
 	resultsCh := make(chan *inventory.ExtendedGcpAsset)
 	go f.provider.ListAssetTypes(ctx, lo.Keys(reversedGcpAssetTypes), resultsCh)
diff --git a/internal/resources/fetching/fetchers/gcp/assets_fetcher_test.go b/internal/resources/fetching/fetchers/gcp/assets_fetcher_test.go
@@ -69,7 +69,6 @@ func (s *GcpAssetsFetcherTestSuite) TestFetcher_Fetch() {
 			AssetType: "compute.googleapis.com/Instance",
 		},
 	}
-	mockInventoryService.EXPECT().Clear()
 	mockInventoryService.On("ListAssetTypes", mock.Anything, mock.Anything, mock.Anything).
 		Run(func(args mock.Arguments) {
 			ch := args.Get(2).(chan<- *inventory.ExtendedGcpAsset)
diff --git a/internal/resources/fetching/fetchers/gcp/log_sink_fetcher.go b/internal/resources/fetching/fetchers/gcp/log_sink_fetcher.go
@@ -57,7 +57,6 @@ func NewGcpLogSinkFetcher(_ context.Context, log *clog.Logger, ch chan fetching.
 func (f *GcpLogSinkFetcher) Fetch(ctx context.Context, cycleMetadata cycle.Metadata) error {
 	f.log.Info("Starting GcpLogSinkFetcher.Fetch")
 	defer f.log.Info("GcpLogSinkFetcher.Fetch done")
-	defer f.provider.Clear()
 
 	resultsCh := make(chan *inventory.ProjectAssets)
 	go f.provider.ListProjectAssets(ctx, []string{inventory.LogSinkAssetType}, resultsCh)
diff --git a/internal/resources/fetching/fetchers/gcp/log_sink_fetcher_test.go b/internal/resources/fetching/fetchers/gcp/log_sink_fetcher_test.go
@@ -70,7 +70,6 @@ func (s *GcpLogSinkFetcherTestSuite) TestLogSinkFetcher_Fetch_Success() {
 			AccountId: "1",
 		},
 	}
-	mockInventoryService.EXPECT().Clear()
 	mockInventoryService.On("ListProjectAssets", mock.Anything, []string{inventory.LogSinkAssetType}, mock.Anything).
 		Run(func(args mock.Arguments) {
 			ch := args.Get(2).(chan<- *inventory.ProjectAssets)
diff --git a/internal/resources/fetching/fetchers/gcp/monitoring_fetcher.go b/internal/resources/fetching/fetchers/gcp/monitoring_fetcher.go
@@ -52,7 +52,6 @@ func NewGcpMonitoringFetcher(_ context.Context, log *clog.Logger, ch chan fetchi
 func (f *GcpMonitoringFetcher) Fetch(ctx context.Context, cycleMetadata cycle.Metadata) error {
 	f.log.Info("Starting GcpMonitoringFetcher.Fetch")
 	defer f.log.Info("GcpMonitoringFetcher.Fetch done")
-	defer f.provider.Clear()
 
 	resultsCh := make(chan *inventory.MonitoringAsset)
 	go f.provider.ListMonitoringAssets(ctx, resultsCh)
diff --git a/internal/resources/fetching/fetchers/gcp/monitoring_fetcher_test.go b/internal/resources/fetching/fetchers/gcp/monitoring_fetcher_test.go
@@ -75,7 +75,6 @@ func (s *GcpMonitoringFetcherTestSuite) TestMonitoringFetcher_Fetch_Success() {
 		},
 	}
 
-	mockInventoryService.EXPECT().Clear()
 	mockInventoryService.On("ListMonitoringAssets", mock.Anything, mock.Anything).
 		Run(func(args mock.Arguments) {
 			ch := args.Get(1).(chan<- *inventory.MonitoringAsset)
diff --git a/internal/resources/fetching/fetchers/gcp/networks_fetcher.go b/internal/resources/fetching/fetchers/gcp/networks_fetcher.go
@@ -51,7 +51,6 @@ func NewGcpNetworksFetcher(_ context.Context, log *clog.Logger, ch chan fetching
 func (f *GcpNetworksFetcher) Fetch(ctx context.Context, cycleMetadata cycle.Metadata) error {
 	f.log.Info("Starting GcpNetworksFetcher.Fetch")
 	defer f.log.Info("GcpNetworksFetcher.Fetch done")
-	defer f.provider.Clear()
 
 	resultsCh := make(chan *inventory.ExtendedGcpAsset)
 	go f.provider.ListNetworkAssets(ctx, resultsCh)
diff --git a/internal/resources/fetching/fetchers/gcp/networks_fetcher_test.go b/internal/resources/fetching/fetchers/gcp/networks_fetcher_test.go
@@ -67,7 +67,6 @@ func (s *GcpNetworksFetcherTestSuite) TestNetworksFetcher_Fetch_Success() {
 			AccountId: "1",
 		},
 	}
-	mockInventoryService.EXPECT().Clear()
 	mockInventoryService.On("ListNetworkAssets", mock.Anything, mock.Anything).
 		Run(func(args mock.Arguments) {
 			ch := args.Get(1).(chan<- *inventory.ExtendedGcpAsset)
diff --git a/internal/resources/fetching/fetchers/gcp/policies_fetcher.go b/internal/resources/fetching/fetchers/gcp/policies_fetcher.go
@@ -52,7 +52,6 @@ func NewGcpPoliciesFetcher(_ context.Context, log *clog.Logger, ch chan fetching
 func (f *GcpPoliciesFetcher) Fetch(ctx context.Context, cycleMetadata cycle.Metadata) error {
 	f.log.Info("Starting GcpPoliciesFetcher.Fetch")
 	defer f.log.Info("GcpPoliciesFetcher.Fetch done")
-	defer f.provider.Clear()
 
 	resultsCh := make(chan *inventory.ProjectPoliciesAsset)
 	go f.provider.ListProjectsAncestorsPolicies(ctx, resultsCh)
diff --git a/internal/resources/fetching/fetchers/gcp/policies_fetcher_test.go b/internal/resources/fetching/fetchers/gcp/policies_fetcher_test.go
@@ -70,7 +70,6 @@ func (s *GcpPoliciesFetcherTestSuite) TestPoliciesFetcher_Fetch_Success() {
 		},
 	}
 
-	mockInventoryService.EXPECT().Clear()
 	mockInventoryService.On("ListProjectsAncestorsPolicies", mock.Anything, mock.Anything).
 		Run(func(args mock.Arguments) {
 			ch := args.Get(1).(chan<- *inventory.ProjectPoliciesAsset)
diff --git a/internal/resources/fetching/fetchers/gcp/service_usage_fetcher.go b/internal/resources/fetching/fetchers/gcp/service_usage_fetcher.go
@@ -57,7 +57,6 @@ func NewGcpServiceUsageFetcher(_ context.Context, log *clog.Logger, ch chan fetc
 func (f *GcpServiceUsageFetcher) Fetch(ctx context.Context, cycleMetadata cycle.Metadata) error {
 	f.log.Info("Starting GcpServiceUsageFetcher.Fetch")
 	defer f.log.Info("GcpServiceUsageFetcher.Fetch done")
-	defer f.provider.Clear()
 
 	resultsCh := make(chan *inventory.ProjectAssets)
 	go f.provider.ListProjectAssets(ctx, []string{inventory.ServiceUsageAssetType}, resultsCh)
diff --git a/internal/resources/fetching/fetchers/gcp/service_usage_fetcher_test.go b/internal/resources/fetching/fetchers/gcp/service_usage_fetcher_test.go
@@ -69,7 +69,6 @@ func (s *GcpServiceUsageFetcherTestSuite) TestServiceUsageFetcher_Fetch_Success(
 		},
 	}
 
-	mockInventoryService.EXPECT().Clear()
 	mockInventoryService.On("ListProjectAssets", mock.Anything, []string{inventory.ServiceUsageAssetType}, mock.Anything).
 		Run(func(args mock.Arguments) {
 			ch := args.Get(2).(chan<- *inventory.ProjectAssets)
diff --git a/internal/resources/providers/gcplib/inventory/resource_manager.go b/internal/resources/providers/gcplib/inventory/resource_manager.go
@@ -22,8 +22,11 @@ import (
 	"fmt"
 	"strings"
 	"sync"
+	"time"
 
 	"cloud.google.com/go/asset/apiv1/assetpb"
+	"golang.org/x/sync/singleflight"
+	"golang.org/x/time/rate"
 	"google.golang.org/api/cloudresourcemanager/v3"
 	"google.golang.org/api/option"
 
@@ -36,6 +39,8 @@ type ResourceManagerWrapper struct {
 	log                        *clog.Logger
 	config                     auth.GcpFactoryConfig
 	accountMetadataCache       sync.Map
+	metadataGroup              singleflight.Group
+	projectsRateLimiter        *rate.Limiter
 	getProjectDisplayName      func(ctx context.Context, parent string) string
 	getOrganizationDisplayName func(ctx context.Context, parent string) string
 }
@@ -50,36 +55,43 @@ func NewResourceManagerWrapper(ctx context.Context, log *clog.Logger, gcpConfig
 
 	var orgName string
 	var orgNameMu sync.Mutex
-	return &ResourceManagerWrapper{
+
+	wrapper := &ResourceManagerWrapper{
 		log:                  log,
 		config:               gcpConfig,
 		accountMetadataCache: sync.Map{},
-		getProjectDisplayName: func(ctx context.Context, parent string) string {
-			prj, err := crmService.Projects.Get(parent).Context(ctx).Do()
-			if err != nil {
-				log.Errorf("error fetching GCP Project: %s, error: %s", parent, err)
-				return ""
-			}
-			return prj.DisplayName
-		},
-		getOrganizationDisplayName: func(ctx context.Context, parent string) string {
-			orgNameMu.Lock()
-			defer orgNameMu.Unlock()
-
-			if orgName != "" {
-				return orgName
-			}
-
-			// parent is the ID of the organization, and is the same for every call, so we cache it
-			org, err := crmService.Organizations.Get(parent).Context(ctx).Do()
-			if err != nil {
-				log.Errorf("error fetching GCP Org: %s, error: %s", parent, err)
-				return ""
-			}
-			orgName = org.DisplayName
+		projectsRateLimiter:  rate.NewLimiter(rate.Every(time.Minute/600), 1),
+	}
+	wrapper.getProjectDisplayName = func(ctx context.Context, parent string) string {
+		if err := wrapper.projectsRateLimiter.Wait(ctx); err != nil {
+			log.Errorf("rate limiter error for GCP Project: %s, error: %s", parent, err)
+			return ""
+		}
+		prj, err := crmService.Projects.Get(parent).Context(ctx).Do()
+		if err != nil {
+			log.Errorf("error fetching GCP Project: %s, error: %s", parent, err)
+			return ""
+		}
+		return prj.DisplayName
+	}
+	wrapper.getOrganizationDisplayName = func(ctx context.Context, parent string) string {
+		orgNameMu.Lock()
+		defer orgNameMu.Unlock()
+
+		if orgName != "" {
 			return orgName
-		},
-	}, nil
+		}
+
+		// parent is the ID of the organization, and is the same for every call, so we cache it
+		org, err := crmService.Organizations.Get(parent).Context(ctx).Do()
+		if err != nil {
+			log.Errorf("error fetching GCP Org: %s, error: %s", parent, err)
+			return ""
+		}
+		orgName = org.DisplayName
+		return orgName
+	}
+	return wrapper, nil
 }
 
 func (c *ResourceManagerWrapper) GetCloudMetadata(ctx context.Context, asset *assetpb.Asset) *fetching.CloudAccountMetadata {
@@ -94,9 +106,14 @@ func (c *ResourceManagerWrapper) GetCloudMetadata(ctx context.Context, asset *as
 		}
 		c.log.Errorf("error casting cloud account metadata for key: %s", key)
 	}
-	cloudAccountMetadata := c.getMetadata(ctx, orgId, projectId)
-	c.accountMetadataCache.Store(key, cloudAccountMetadata)
-	return cloudAccountMetadata
+
+	result, _, _ := c.metadataGroup.Do(key, func() (any, error) {
+		metadata := c.getMetadata(ctx, orgId, projectId)
+		c.accountMetadataCache.Store(key, metadata)
+		return metadata, nil
+	})
+
+	return result.(*fetching.CloudAccountMetadata)
 }
 
 func (c *ResourceManagerWrapper) getMetadata(ctx context.Context, orgId string, projectId string) *fetching.CloudAccountMetadata {
@@ -127,9 +144,7 @@ func (c *ResourceManagerWrapper) getMetadata(ctx context.Context, orgId string,
 	}
 }
 
-func (c *ResourceManagerWrapper) Clear() {
-	c.accountMetadataCache.Clear()
-}
+func (c *ResourceManagerWrapper) Clear() {}
 
 func getOrganizationId(ancestors []string) string {
 	last := ancestors[len(ancestors)-1]
diff --git a/internal/resources/providers/gcplib/inventory/resource_manager_test.go b/internal/resources/providers/gcplib/inventory/resource_manager_test.go
@@ -20,10 +20,13 @@ package inventory
 import (
 	"context"
 	"sync"
+	"sync/atomic"
 	"testing"
+	"time"
 
 	"cloud.google.com/go/asset/apiv1/assetpb"
 	"github.com/stretchr/testify/suite"
+	"golang.org/x/time/rate"
 	"google.golang.org/api/option"
 
 	"github.com/elastic/cloudbeat/internal/resources/providers/gcplib/auth"
@@ -165,3 +168,63 @@ func (s *ResourceManagerTestSuite) TestGetOrganizationId() {
 func (s *ResourceManagerTestSuite) TestGetProjectId() {
 	s.Equal("5", getProjectId(ancestors))
 }
+
+func (s *ResourceManagerTestSuite) TestProjectsRateLimiterWait() {
+	t := s.T()
+	ctx := t.Context()
+	duration := time.Millisecond
+	crm := s.NewMockResourceManagerWrapper()
+	crm.projectsRateLimiter = rate.NewLimiter(rate.Every(duration), 1)
+	crm.getProjectDisplayName = func(ctx context.Context, _ string) string {
+		if err := crm.projectsRateLimiter.Wait(ctx); err != nil {
+			return ""
+		}
+		return "projectName"
+	}
+
+	totalRequests := 5
+	startTime := time.Now()
+	for range totalRequests {
+		crm.getProjectDisplayName(ctx, "projects/test")
+	}
+	endTime := time.Now()
+
+	actualDuration := endTime.Sub(startTime)
+	minDuration := duration * time.Duration(totalRequests-1) // 1st request is instant, rest wait
+	s.GreaterOrEqual(actualDuration, minDuration)
+}
+
+func (s *ResourceManagerTestSuite) TestSingleflightDeduplicatesConcurrentRequests() {
+	t := s.T()
+	ctx := t.Context()
+	crm := s.NewMockResourceManagerWrapper()
+
+	var fetchCount atomic.Int32
+	crm.getProjectDisplayName = func(_ context.Context, _ string) string {
+		fetchCount.Add(1)
+		time.Sleep(10 * time.Millisecond)
+		return "projectName"
+	}
+
+	asset := &assetpb.Asset{
+		Name: "projects/1",
+		Ancestors: []string{
+			"projects/1",
+			"organizations/1",
+		},
+	}
+
+	const numRequests = 10
+	var wg sync.WaitGroup
+	wg.Add(numRequests)
+	for range numRequests {
+		go func() {
+			defer wg.Done()
+			crm.GetCloudMetadata(ctx, asset)
+		}()
+	}
+	wg.Wait()
+
+	// only 1 fetch, rest waited on singleflight
+	s.Equal(int32(1), fetchCount.Load())
+}

Original file line number	Diff line number	Diff line change
`@@ -69,7 +69,6 @@ func (s *GcpAssetsFetcherTestSuite) TestFetcher_Fetch() {`
`69`	`69`	`AssetType: "compute.googleapis.com/Instance",`
`70`	`70`	`},`
`71`	`71`	`}`
`72`		`- mockInventoryService.EXPECT().Clear()`
`73`	`72`	`mockInventoryService.On("ListAssetTypes", mock.Anything, mock.Anything, mock.Anything).`
`74`	`73`	`Run(func(args mock.Arguments) {`
`75`	`74`	`ch := args.Get(2).(chan<- *inventory.ExtendedGcpAsset)`
Original file line number	Diff line number	Diff line change
`@@ -70,7 +70,6 @@ func (s *GcpLogSinkFetcherTestSuite) TestLogSinkFetcher_Fetch_Success() {`
`70`	`70`	`AccountId: "1",`
`71`	`71`	`},`
`72`	`72`	`}`
`73`		`- mockInventoryService.EXPECT().Clear()`
`74`	`73`	`mockInventoryService.On("ListProjectAssets", mock.Anything, []string{inventory.LogSinkAssetType}, mock.Anything).`
`75`	`74`	`Run(func(args mock.Arguments) {`
`76`	`75`	`ch := args.Get(2).(chan<- *inventory.ProjectAssets)`
Original file line number	Diff line number	Diff line change
`@@ -75,7 +75,6 @@ func (s *GcpMonitoringFetcherTestSuite) TestMonitoringFetcher_Fetch_Success() {`
`75`	`75`	`},`
`76`	`76`	`}`
`77`	`77`
`78`		`- mockInventoryService.EXPECT().Clear()`
`79`	`78`	`mockInventoryService.On("ListMonitoringAssets", mock.Anything, mock.Anything).`
`80`	`79`	`Run(func(args mock.Arguments) {`
`81`	`80`	`ch := args.Get(1).(chan<- *inventory.MonitoringAsset)`
Original file line number	Diff line number	Diff line change
`@@ -67,7 +67,6 @@ func (s *GcpNetworksFetcherTestSuite) TestNetworksFetcher_Fetch_Success() {`
`67`	`67`	`AccountId: "1",`
`68`	`68`	`},`
`69`	`69`	`}`
`70`		`- mockInventoryService.EXPECT().Clear()`
`71`	`70`	`mockInventoryService.On("ListNetworkAssets", mock.Anything, mock.Anything).`
`72`	`71`	`Run(func(args mock.Arguments) {`
`73`	`72`	`ch := args.Get(1).(chan<- *inventory.ExtendedGcpAsset)`