Add cloud_provider block to database_observability.postgres

Copy over functionality from `database_observability.mysql` to populate
cloud provider info in `connection_info` collector for Postgres.

Author: cristiangreco

docs/sources/reference/components/database_observability/database_observability.postgres.md

@@ -48,16 +48,39 @@ You can use the following blocks with `database_observability.postgres`:
 
 | Block                              | Description                                       | Required |
 |------------------------------------|---------------------------------------------------|----------|
+| [`cloud_provider`][cloud_provider]   | Provide Cloud Provider information.               | no       |
+| `cloud_provider` > [`aws`][aws]      | Provide AWS database host information.            | no       |
 | [`query_details`][query_details]   | Configure the queries collector.                  | no       |
 | [`query_samples`][query_samples]   | Configure the query samples collector.            | no       |
 | [`schema_details`][schema_details] | Configure the schema and table details collector. | no       |
 | [`explain_plans`][explain_plans]   | Configure the explain plans collector.            | no       |
 
+The > symbol indicates deeper levels of nesting.
+For example, `cloud_provider` > `aws` refers to a `aws` block defined inside an `cloud_provider` block.
+
+[cloud_provider]: #cloud_provider
+[aws]: #aws
 [query_details]: #query_details
 [query_samples]: #query_samples
 [schema_details]: #schema_details
 [explain_plans]: #explain_plans
 
+### `cloud_provider`
+
+The `cloud_provider` block has no attributes.
+It contains zero or more [`aws`][aws] blocks.
+You use the `cloud_provider` block to provide information related to the cloud provider that hosts the database under observation.
+This information is appended as labels to the collected metrics.
+The labels make it easier for you to filter and group your metrics.
+
+### `aws`
+
+The `aws` block supplies the [ARN](https://docs.aws.amazon.com/IAM/latest/UserGuide/reference-arns.html) identifier for the database being monitored.
+
+| Name  | Type     | Description                                             | Default | Required |
+|-------|----------|---------------------------------------------------------|---------|----------|
+| `arn` | `string` | The ARN associated with the database under observation. |         | yes      |
+
 ### `query_details`
 
 | Name               | Type       | Description                                          | Default | Required |
@@ -96,6 +119,12 @@ database_observability.postgres "orders_db" {
   data_source_name = "postgres://user:pass@localhost:5432/mydb"
   forward_to = [loki.write.logs_service.receiver]
   enable_collectors = ["query_details", "query_samples", "schema_details"]
+
+  cloud_provider {
+    aws {
+      arn = "your-rds-db-arn"
+    }
+  }
 }
 
 prometheus.scrape "orders_db" {

internal/component/database_observability/mysql/collector/connection_info.go

@@ -7,10 +7,9 @@ import (
 	"strings"
 
 	"github.com/go-sql-driver/mysql"
+	"github.com/grafana/alloy/internal/component/database_observability"
 	"github.com/prometheus/client_golang/prometheus"
 	"go.uber.org/atomic"
-
-	"github.com/grafana/alloy/internal/component/database_observability"
 )
 
 const ConnectionInfoName = "connection_info"

internal/component/database_observability/postgres/collector/connection_info.go

@@ -5,6 +5,7 @@ import (
 	"regexp"
 	"strings"
 
+	"github.com/grafana/alloy/internal/component/database_observability"
 	"github.com/prometheus/client_golang/prometheus"
 	"go.uber.org/atomic"
 )
@@ -22,13 +23,15 @@ type ConnectionInfoArguments struct {
 	DSN           string
 	Registry      *prometheus.Registry
 	EngineVersion string
+	CloudProvider *database_observability.CloudProvider
 }
 
 type ConnectionInfo struct {
 	DSN           string
 	Registry      *prometheus.Registry
 	EngineVersion string
 	InfoMetric    *prometheus.GaugeVec
+	CloudProvider *database_observability.CloudProvider
 
 	running *atomic.Bool
 }
@@ -38,7 +41,7 @@ func NewConnectionInfo(args ConnectionInfoArguments) (*ConnectionInfo, error) {
 		Namespace: "database_observability",
 		Name:      "connection_info",
 		Help:      "Information about the connection",
-	}, []string{"provider_name", "provider_region", "db_instance_identifier", "engine", "engine_version"})
+	}, []string{"provider_name", "provider_region", "provider_account", "db_instance_identifier", "engine", "engine_version"})
 
 	args.Registry.MustRegister(infoMetric)
 
@@ -47,6 +50,7 @@ func NewConnectionInfo(args ConnectionInfoArguments) (*ConnectionInfo, error) {
 		Registry:      args.Registry,
 		EngineVersion: args.EngineVersion,
 		InfoMetric:    infoMetric,
+		CloudProvider: args.CloudProvider,
 		running:       &atomic.Bool{},
 	}, nil
 }
@@ -56,34 +60,45 @@ func (c *ConnectionInfo) Name() string {
 }
 
 func (c *ConnectionInfo) Start(ctx context.Context) error {
-	c.running.Store(true)
-
 	var (
 		providerName         = "unknown"
 		providerRegion       = "unknown"
+		providerAccount      = "unknown"
 		dbInstanceIdentifier = "unknown"
 		engine               = "postgres"
 		engineVersion        = "unknown"
 	)
 
-	parts, err := ParseURL(c.DSN)
-	if err != nil {
-		return err
-	}
-
-	if host, ok := parts["host"]; ok {
-		if strings.HasSuffix(host, "rds.amazonaws.com") {
+	if c.CloudProvider != nil {
+		if c.CloudProvider.AWS != nil {
 			providerName = "aws"
-			matches := rdsRegex.FindStringSubmatch(host)
-			if len(matches) > 3 {
-				dbInstanceIdentifier = matches[1]
-				providerRegion = matches[3]
+			providerAccount = c.CloudProvider.AWS.ARN.AccountID
+			providerRegion = c.CloudProvider.AWS.ARN.Region
+
+			// We only support RDS database for now. Resource types and ARN formats are documented at: https://docs.aws.amazon.com/service-authorization/latest/reference/list_amazonrds.html#amazonrds-resources-for-iam-policies
+			if resource := c.CloudProvider.AWS.ARN.Resource; strings.HasPrefix(resource, "db:") {
+				dbInstanceIdentifier = strings.TrimPrefix(resource, "db:")
 			}
-		} else if strings.HasSuffix(host, "postgres.database.azure.com") {
-			providerName = "azure"
-			matches := azureRegex.FindStringSubmatch(host)
-			if len(matches) > 1 {
-				dbInstanceIdentifier = matches[1]
+		}
+	} else {
+		parts, err := ParseURL(c.DSN)
+		if err != nil {
+			return err
+		}
+		if host, ok := parts["host"]; ok {
+			if strings.HasSuffix(host, "rds.amazonaws.com") {
+				providerName = "aws"
+				matches := rdsRegex.FindStringSubmatch(host)
+				if len(matches) > 3 {
+					dbInstanceIdentifier = matches[1]
+					providerRegion = matches[3]
+				}
+			} else if strings.HasSuffix(host, "postgres.database.azure.com") {
+				providerName = "azure"
+				matches := azureRegex.FindStringSubmatch(host)
+				if len(matches) > 1 {
+					dbInstanceIdentifier = matches[1]
+				}
 			}
 		}
 	}
@@ -93,7 +108,9 @@ func (c *ConnectionInfo) Start(ctx context.Context) error {
 		engineVersion = matches[1]
 	}
 
-	c.InfoMetric.WithLabelValues(providerName, providerRegion, dbInstanceIdentifier, engine, engineVersion).Set(1)
+	c.running.Store(true)
+
+	c.InfoMetric.WithLabelValues(providerName, providerRegion, providerAccount, dbInstanceIdentifier, engine, engineVersion).Set(1)
 	return nil
 }
 

internal/component/database_observability/postgres/collector/connection_info_test.go

@@ -5,10 +5,13 @@ import (
 	"strings"
 	"testing"
 
+	"github.com/aws/aws-sdk-go-v2/aws/arn"
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/client_golang/prometheus/testutil"
 	"github.com/stretchr/testify/require"
 	"go.uber.org/goleak"
+
+	"github.com/grafana/alloy/internal/component/database_observability"
 )
 
 func TestConnectionInfo(t *testing.T) {
@@ -17,32 +20,48 @@ func TestConnectionInfo(t *testing.T) {
 	const baseExpectedMetrics = `
 	# HELP database_observability_connection_info Information about the connection
 	# TYPE database_observability_connection_info gauge
-	database_observability_connection_info{db_instance_identifier="%s",engine="%s",engine_version="%s",provider_name="%s",provider_region="%s"} 1
+	database_observability_connection_info{db_instance_identifier="%s",engine="%s",engine_version="%s",provider_account="%s",provider_name="%s",provider_region="%s"} 1
 `
 
 	testCases := []struct {
 		name            string
 		dsn             string
 		engineVersion   string
+		cloudProvider   *database_observability.CloudProvider
 		expectedMetrics string
 	}{
 		{
 			name:            "generic dsn",
 			dsn:             "postgres://user:pass@localhost:5432/mydb",
 			engineVersion:   "15.4",
-			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "unknown", "postgres", "15.4", "unknown", "unknown"),
+			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "unknown", "postgres", "15.4", "unknown", "unknown", "unknown"),
 		},
 		{
 			name:            "AWS/RDS dsn",
 			dsn:             "postgres://user:[email protected]:5432/mydb",
 			engineVersion:   "15.4",
-			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "products-db", "postgres", "15.4", "aws", "us-east-1"),
+			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "products-db", "postgres", "15.4", "unknown", "aws", "us-east-1"),
+		},
+		{
+			name:          "AWS/RDS dsn with cloud provider info supplied",
+			dsn:           "postgres://user:[email protected]:5432/mydb",
+			engineVersion: "15.4",
+			cloudProvider: &database_observability.CloudProvider{
+				AWS: &database_observability.AWSCloudProviderInfo{
+					ARN: arn.ARN{
+						Region:    "us-east-1",
+						AccountID: "some-account-123",
+						Resource:  "db:products-db",
+					},
+				},
+			},
+			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "products-db", "postgres", "15.4", "some-account-123", "aws", "us-east-1"),
 		},
 		{
 			name:            "Azure flexibleservers dsn",
 			dsn:             "postgres://user:[email protected]:5432/mydb",
 			engineVersion:   "15.4",
-			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "products-db", "postgres", "15.4", "azure", "unknown"),
+			expectedMetrics: fmt.Sprintf(baseExpectedMetrics, "products-db", "postgres", "15.4", "unknown", "azure", "unknown"),
 		},
 	}
 
@@ -53,6 +72,7 @@ func TestConnectionInfo(t *testing.T) {
 			DSN:           tc.dsn,
 			Registry:      reg,
 			EngineVersion: tc.engineVersion,
+			CloudProvider: tc.cloudProvider,
 		})
 		require.NoError(t, err)
 		require.NotNil(t, collector)

internal/component/database_observability/postgres/component.go

@@ -11,6 +11,7 @@ import (
 	"sync"
 	"time"
 
+	"github.com/aws/aws-sdk-go-v2/aws/arn"
 	"github.com/blang/semver/v4"
 	"github.com/lib/pq"
 	"github.com/prometheus/client_golang/prometheus"
@@ -67,11 +68,19 @@ type Arguments struct {
 	EnableCollectors  []string            `alloy:"enable_collectors,attr,optional"`
 	DisableCollectors []string            `alloy:"disable_collectors,attr,optional"`
 
+	CloudProvider          *CloudProvider         `alloy:"cloud_provider,block,optional"`
 	QuerySampleArguments   QuerySampleArguments   `alloy:"query_samples,block,optional"`
 	QueryTablesArguments   QueryTablesArguments   `alloy:"query_details,block,optional"`
 	SchemaDetailsArguments SchemaDetailsArguments `alloy:"schema_details,block,optional"`
+	ExplainPlanArguments   ExplainPlanArguments   `alloy:"explain_plans,block,optional"`
+}
+
+type CloudProvider struct {
+	AWS *AWSCloudProviderInfo `alloy:"aws,block,optional"`
+}
 
-	ExplainPlanArguments ExplainPlanArguments `alloy:"explain_plans,block,optional"`
+type AWSCloudProviderInfo struct {
+	ARN string `alloy:"arn,attr"`
 }
 
 type QuerySampleArguments struct {
@@ -342,6 +351,19 @@ func (c *Component) startCollectors(systemID string, engineVersion string) error
 		startErrors = append(startErrors, errorString)
 	}
 
+	var cloudProviderInfo *database_observability.CloudProvider
+	if c.args.CloudProvider != nil && c.args.CloudProvider.AWS != nil {
+		arn, err := arn.Parse(c.args.CloudProvider.AWS.ARN)
+		if err != nil {
+			level.Error(c.opts.Logger).Log("msg", "failed to parse AWS cloud provider ARN", "err", err)
+		}
+		cloudProviderInfo = &database_observability.CloudProvider{
+			AWS: &database_observability.AWSCloudProviderInfo{
+				ARN: arn,
+			},
+		}
+	}
+
 	entryHandler := addLokiLabels(loki.NewEntryHandler(c.handler.Chan(), func() {}), c.instanceKey, systemID)
 
 	collectors := enableOrDisableCollectors(c.args)
@@ -384,6 +406,7 @@ func (c *Component) startCollectors(systemID string, engineVersion string) error
 		DSN:           string(c.args.DataSourceName),
 		Registry:      c.registry,
 		EngineVersion: engineVersion,
+		CloudProvider: cloudProviderInfo,
 	})
 	if err != nil {
 		logStartError(collector.ConnectionInfoName, "create", err)