[extension/opamp] Implement ReportsHealth capability

.chloggen/opamp-extension-reportshealth.yaml

@@ -0,0 +1,27 @@
+# Use this changelog template to create an entry for release notes.
+
+# One of 'breaking', 'deprecation', 'new_component', 'enhancement', 'bug_fix'
+change_type: enhancement
+
+# The name of the component, or a single word describing the area of concern, (e.g. filelogreceiver)
+component: opampextension
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Implement `ReportsHealth` capability in OpAMP extension
+
+# Mandatory: One or more tracking issues related to the change. You can use the PR number here if no issue exists.
+issues: [35433]
+
+# (Optional) One or more lines of additional information to render under the primary note.
+# These lines will be padded with 2 spaces and then inserted directly into the document.
+# Use pipe (|) for multiline entries.
+subtext:
+
+# If your change doesn't affect end users or the exported elements of any package,
+# you should instead start your pull request title with [chore] or use the "Skip Changelog" label.
+# Optional: The change log or logs in which this entry should be included.
+# e.g. '[user]' or '[user, api]'
+# Include 'user' if the change is relevant to end users.
+# Include 'api' if there is a change to a library API.
+# Default: '[user]'
+change_logs: []

extension/opampextension/config.go

@@ -53,6 +53,8 @@ type AgentDescription struct {
 type Capabilities struct {
 	// ReportsEffectiveConfig enables the OpAMP ReportsEffectiveConfig Capability. (default: true)
 	ReportsEffectiveConfig bool `mapstructure:"reports_effective_config"`
+	// ReportsHealth enables the OpAMP ReportsHealth Capability. (default: true)
+	ReportsHealth bool `mapstructure:"reports_health"`
 }
 
 func (caps Capabilities) toAgentCapabilities() protobufs.AgentCapabilities {
@@ -62,6 +64,9 @@ func (caps Capabilities) toAgentCapabilities() protobufs.AgentCapabilities {
 	if caps.ReportsEffectiveConfig {
 		agentCapabilities |= protobufs.AgentCapabilities_AgentCapabilities_ReportsEffectiveConfig
 	}
+	if caps.ReportsHealth {
+		agentCapabilities |= protobufs.AgentCapabilities_AgentCapabilities_ReportsHealth
+	}
 
 	return agentCapabilities
 }

extension/opampextension/config_test.go

@@ -4,6 +4,7 @@
 package opampextension
 
 import (
+	"github.com/open-telemetry/opamp-go/protobufs"
 	"path/filepath"
 	"testing"
 	"time"
@@ -39,6 +40,7 @@ func TestUnmarshalConfig(t *testing.T) {
 			InstanceUID: "01BX5ZZKBKACTAV9WEVGEMMVRZ",
 			Capabilities: Capabilities{
 				ReportsEffectiveConfig: true,
+				ReportsHealth:          true,
 			},
 			PPIDPollInterval: 5 * time.Second,
 		}, cfg)
@@ -63,6 +65,7 @@ func TestUnmarshalHttpConfig(t *testing.T) {
 			InstanceUID: "01BX5ZZKBKACTAV9WEVGEMMVRZ",
 			Capabilities: Capabilities{
 				ReportsEffectiveConfig: true,
+				ReportsHealth:          true,
 			},
 			PPIDPollInterval: 5 * time.Second,
 		}, cfg)
@@ -286,3 +289,41 @@ func TestConfig_Validate(t *testing.T) {
 		})
 	}
 }
+
+func TestCapabilities_toAgentCapabilities(t *testing.T) {
+	type fields struct {
+		ReportsEffectiveConfig bool
+		ReportsHealth          bool
+	}
+	tests := []struct {
+		name   string
+		fields fields
+		want   protobufs.AgentCapabilities
+	}{
+		{
+			name: "default capabilities",
+			fields: fields{
+				ReportsEffectiveConfig: false,
+				ReportsHealth:          false,
+			},
+			want: protobufs.AgentCapabilities_AgentCapabilities_ReportsStatus,
+		},
+		{
+			name: "all supported capabilities enabled",
+			fields: fields{
+				ReportsEffectiveConfig: true,
+				ReportsHealth:          true,
+			},
+			want: protobufs.AgentCapabilities_AgentCapabilities_ReportsStatus | protobufs.AgentCapabilities_AgentCapabilities_ReportsEffectiveConfig | protobufs.AgentCapabilities_AgentCapabilities_ReportsHealth,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			caps := Capabilities{
+				ReportsEffectiveConfig: tt.fields.ReportsEffectiveConfig,
+				ReportsHealth:          tt.fields.ReportsHealth,
+			}
+			assert.Equalf(t, tt.want, caps.toAgentCapabilities(), "toAgentCapabilities()")
+		})
+	}
+}

extension/opampextension/factory.go

@@ -27,6 +27,7 @@ func createDefaultConfig() component.Config {
 		Server: &OpAMPServer{},
 		Capabilities: Capabilities{
 			ReportsEffectiveConfig: true,
+			ReportsHealth:          true,
 		},
 		PPIDPollInterval: 5 * time.Second,
 	}

extension/opampextension/opamp_agent.go

@@ -7,13 +7,6 @@ import (
 	"context"
 	"errors"
 	"fmt"
-	"net/http"
-	"os"
-	"runtime"
-	"sort"
-	"strings"
-	"sync"
-
 	"github.com/google/uuid"
 	"github.com/oklog/ulid/v2"
 	"github.com/open-telemetry/opamp-go/client"
@@ -27,6 +20,12 @@ import (
 	"go.uber.org/zap"
 	"golang.org/x/exp/maps"
 	"gopkg.in/yaml.v3"
+	"net/http"
+	"os"
+	"runtime"
+	"sort"
+	"strings"
+	"sync"
 
 	"github.com/open-telemetry/opentelemetry-collector-contrib/extension/opampcustommessages"
 )
@@ -112,6 +111,10 @@ func (o *opampAgent) Start(ctx context.Context, host component.Host) error {
 		return err
 	}
 
+	if err := o.opampClient.SetHealth(&protobufs.ComponentHealth{Healthy: false}); err != nil {
+		o.logger.Error("Could not report health to OpAMP server", zap.Error(err))
+	}
+
 	o.logger.Debug("Starting OpAMP client...")
 
 	if err := o.opampClient.Start(context.Background(), settings); err != nil {
@@ -120,6 +123,10 @@ func (o *opampAgent) Start(ctx context.Context, host component.Host) error {
 
 	o.logger.Debug("OpAMP client started")
 
+	if err := o.opampClient.SetHealth(&protobufs.ComponentHealth{Healthy: true}); err != nil {
+		o.logger.Error("Could not report health to OpAMP server", zap.Error(err))
+	}
+
 	return nil
 }
 
@@ -132,8 +139,18 @@ func (o *opampAgent) Shutdown(ctx context.Context) error {
 	if o.opampClient == nil {
 		return nil
 	}
+
+	err := o.opampClient.SetHealth(
+		&protobufs.ComponentHealth{
+			Healthy: false, LastError: "Agent is shut down",
+		},
+	)
+	if err != nil {
+		o.logger.Error("Could not report health to OpAMP server", zap.Error(err))
+	}
+
 	o.logger.Debug("Stopping OpAMP client...")
-	err := o.opampClient.Stop(ctx)
+	err = o.opampClient.Stop(ctx)
 	// Opamp-go considers this an error, but the collector does not.
 	// https://github.com/open-telemetry/opamp-go/issues/255
 	if err != nil && strings.EqualFold(err.Error(), "cannot stop because not started") {

extension/opampextension/opamp_agent_test.go

@@ -31,6 +31,7 @@ func TestNewOpampAgent(t *testing.T) {
 	assert.Equal(t, "test version", o.agentVersion)
 	assert.NotEmpty(t, o.instanceID.String())
 	assert.True(t, o.capabilities.ReportsEffectiveConfig)
+	assert.True(t, o.capabilities.ReportsHealth)
 	assert.Empty(t, o.effectiveConfig)
 	assert.Nil(t, o.agentDescription)
 }