feat(service-graphs): add database_name_attributes config to service graph processor

CHANGELOG.md

@@ -20,6 +20,7 @@
 * [FEATURE] New block encoding vParquet5-preview5 with virtual span row numbers. This is a preview, breaking changes are expected. [#5943](https://github.com/grafana/tempo/pull/5943) (@stoewer)
 * [FEATURE] Add SSE-C encryption support to S3 backend [#5789](https://github.com/grafana/tempo/pull/5789) (@steffsas)
 * [FEATURE] Add support for `Accept: application/vnd.grafana.llm` to Tempo endpoints. Currently supported directly by trace by id and tag values [#5961](https://github.com/grafana/tempo/pull/5961) (@joe-elliott)
+* [FEATURE] add database_name_attributes config to service graph processor [#5398](https://github.com/grafana/tempo/pull/5398) (@KyriosGN0)
   This response is subject to change and should not be relied on. It is intended for LLM consumption only. Even a fundamental change to its representation (yaml? markdown?) would not be considered breaking.
 * [ENHANCEMENT] docs: Add explicit notes about authentication [#5735](https://github.com/grafana/tempo/pull/5735) (@electron0zero)
 * [ENHANCEMENT] On startup, first record for live store to consume is not older than two complete block timeouts [#5693](https://github.com/grafana/tempo/pull/5693) (@ruslan-mikhailov)

docs/sources/tempo/configuration/_index.md

@@ -470,6 +470,9 @@ metrics_generator:
             # Enables additional labels for services and virtual nodes.
             [enable_virtual_node_label: <bool> | default = false]
 
+            # List of attribute names used to identify the database name from span attributes. If it isn't set, the order is peer.service -> server.address -> network.peer.address -> db.name
+            [database_name_attributes: <list of string> | default = ["db.namespace","db.name","db.system"]]
+
         span_metrics:
 
             # Buckets for the latency histogram in seconds.

docs/sources/tempo/configuration/manifest.md

@@ -641,6 +641,10 @@ metrics_generator:
                 - db.system
             span_multiplier_key: ""
             enable_virtual_node_label: false
+            database_name_attributes:
+                - db.namespace
+                - db.name
+                - db.system
         span_metrics:
             histogram_buckets:
                 - 0.002

modules/generator/processor/servicegraphs/config.go

@@ -6,6 +6,8 @@ import (
 
 	"github.com/grafana/tempo/modules/generator/registry"
 	"github.com/prometheus/client_golang/prometheus"
+	semconv "go.opentelemetry.io/otel/semconv/v1.25.0"
+	semconvnew "go.opentelemetry.io/otel/semconv/v1.34.0"
 )
 
 type Config struct {
@@ -45,6 +47,9 @@ type Config struct {
 
 	// EnableVirtualNodeLabel enables additional labels for uninstrumented services
 	EnableVirtualNodeLabel bool `yaml:"enable_virtual_node_label"`
+
+	// DatabaseNameAttributes the list of attribute names used to identify the database name from span attributes.
+	DatabaseNameAttributes []string `yaml:"database_name_attributes"`
 }
 
 func (cfg *Config) RegisterFlagsAndApplyDefaults(string, *flag.FlagSet) {
@@ -62,4 +67,10 @@ func (cfg *Config) RegisterFlagsAndApplyDefaults(string, *flag.FlagSet) {
 	cfg.PeerAttributes = peerAttr
 
 	cfg.EnableMessagingSystemLatencyHistogram = false
+
+	cfg.DatabaseNameAttributes = []string{
+		string(semconvnew.DBNamespaceKey),
+		string(semconv.DBNameKey),
+		string(semconv.DBSystemKey),
+	}
 }

modules/generator/processor/servicegraphs/servicegraphs.go

@@ -14,7 +14,6 @@ import (
 	"github.com/prometheus/client_golang/prometheus/promauto"
 	"go.opentelemetry.io/otel/attribute"
 	semconv "go.opentelemetry.io/otel/semconv/v1.25.0"
-	semconvnew "go.opentelemetry.io/otel/semconv/v1.34.0"
 
 	gen "github.com/grafana/tempo/modules/generator/processor"
 	"github.com/grafana/tempo/modules/generator/processor/servicegraphs/store"
@@ -276,19 +275,11 @@ func (p *Processor) upsertDatabaseRequest(e *store.Edge, resourceAttr []*v1_comm
 	)
 
 	// Check for db.name or db.namespace first.  The dbName is set initially to maintain backwards compatbility.
-	if name, ok := processor_util.FindAttributeValue(string(semconv.DBNameKey), resourceAttr, span.Attributes); ok {
-		dbName = name
-		isDatabase = true
-	}
-	if name, ok := processor_util.FindAttributeValue(string(semconvnew.DBNamespaceKey), span.Attributes); ok {
-		dbName = name
-		isDatabase = true
-	}
-
-	// Check for db.system only if we don't have db.name above
-	if !isDatabase {
-		if _, ok := processor_util.FindAttributeValue(string(semconv.DBSystemKey), resourceAttr, span.Attributes); ok {
+	for _, attrName := range p.Cfg.DatabaseNameAttributes {
+		if name, ok := processor_util.FindAttributeValue(attrName, resourceAttr, span.Attributes); ok {
+			dbName = name
 			isDatabase = true
+			break
 		}
 	}
 
@@ -299,8 +290,6 @@ func (p *Processor) upsertDatabaseRequest(e *store.Edge, resourceAttr []*v1_comm
 	e.ConnectionType = store.Database
 	e.ServerLatencySec = spanDurationSec(span)
 
-	// Set the service name by order of precedence
-
 	// Check for peer.service
 	if name, ok := processor_util.FindAttributeValue(string(semconv.PeerServiceKey), resourceAttr, span.Attributes); ok {
 		e.ServerService = name

modules/generator/processor/servicegraphs/servicegraphs_test.go

@@ -11,16 +11,15 @@ import (
 
 	"github.com/go-kit/log"
 	"github.com/gogo/protobuf/jsonpb"
+	"github.com/grafana/tempo/modules/generator/registry"
+	"github.com/grafana/tempo/pkg/tempopb"
+	"github.com/grafana/tempo/pkg/util/test"
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/prometheus/model/labels"
 	"github.com/stretchr/testify/assert"
 	"github.com/stretchr/testify/require"
 	semconv "go.opentelemetry.io/otel/semconv/v1.25.0"
 	semconvnew "go.opentelemetry.io/otel/semconv/v1.34.0"
-
-	"github.com/grafana/tempo/modules/generator/registry"
-	"github.com/grafana/tempo/pkg/tempopb"
-	"github.com/grafana/tempo/pkg/util/test"
 )
 
 // NOTE: This is a way to know if the contents of the semconv package have changed.
@@ -543,6 +542,36 @@ func TestServiceGraphs_prefixDimensionsAndEnableExtraLabels(t *testing.T) {
 	assert.Equal(t, 0.0, testRegistry.Query(`traces_service_graph_request_failed_total`, dbSystemSystemLabels))
 }
 
+func TestServiceGraphs_DatabaseNameAttributes(t *testing.T) {
+	testRegistry := registry.NewTestRegistry()
+
+	cfg := Config{}
+	cfg.RegisterFlagsAndApplyDefaults("", nil)
+
+	cfg.HistogramBuckets = []float64{0.04}
+	cfg.Dimensions = []string{"beast", "god"}
+	cfg.DatabaseNameAttributes = []string{"db.system"}
+
+	p := New(cfg, "test", testRegistry, log.NewNopLogger(), prometheus.NewCounter(prometheus.CounterOpts{}))
+	defer p.Shutdown(context.Background())
+
+	request, err := loadTestData("testdata/trace-with-queue-database.json")
+	require.NoError(t, err)
+
+	p.PushSpans(context.Background(), request)
+
+	// The server label should be set to the value of db.system
+	labels := labels.FromMap(map[string]string{
+		"client":          "mythical-server",
+		"server":          "postgresql",
+		"connection_type": "database",
+		"beast":           "",
+		"god":             "",
+	})
+	print(testRegistry.Query(`traces_service_graph_request_total`, labels))
+	assert.Equal(t, 1.0, testRegistry.Query(`traces_service_graph_request_total`, labels))
+}
+
 func BenchmarkServiceGraphs(b *testing.B) {
 	testRegistry := registry.NewTestRegistry()