[cmd/opampsupervisor] Support AcceptsRestartCommand Capability

.chloggen/supervisor_restart_command_capability.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: cmd/opampsupervisor
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Support AcceptsRestartCommand Capability.
+
+# Mandatory: One or more tracking issues related to the change. You can use the PR number here if no issue exists.
+issues: [21077]
+
+# (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: []

cmd/opampsupervisor/e2e_test.go

@@ -473,3 +473,40 @@ func waitForSupervisorConnection(connection chan bool, connected bool) {
 		}
 	}
 }
+
+func TestSupervisorRestartCommand(t *testing.T) {
+
+	var connectedCount atomic.Int32
+	server := newOpAMPServer(
+		t,
+		defaultConnectingHandler,
+		server.ConnectionCallbacksStruct{
+			OnConnectedFunc: func(_ types.Connection) {
+				connectedCount.Add(1)
+			},
+			OnMessageFunc: func(_ types.Connection, message *protobufs.AgentToServer) *protobufs.ServerToAgent {
+				return &protobufs.ServerToAgent{}
+			},
+		})
+
+	s := newSupervisor(t, "basic", map[string]string{"url": server.addr})
+	defer s.Shutdown()
+
+	waitForSupervisorConnection(server.supervisorConnected, true)
+
+	require.Eventually(t, func() bool {
+
+		return connectedCount.Load() == 1
+
+	}, 5*time.Second, 500*time.Millisecond, "Collector never connected")
+
+	server.sendToSupervisor(&protobufs.ServerToAgent{
+		Command: &protobufs.ServerToAgentCommand{
+			Type: protobufs.CommandType_CommandType_Restart,
+		},
+	})
+
+	require.Eventually(t, func() bool {
+		return s.LastRestartError() == nil
+	}, 10*time.Second, 250*time.Millisecond, "Collector didn't connect after restart")
+}

cmd/opampsupervisor/supervisor/commander/commander.go

@@ -82,6 +82,7 @@ func (c *Commander) Start(ctx context.Context) error {
 }
 
 func (c *Commander) Restart(ctx context.Context) error {
+	c.logger.Debug("Restarting agent", zap.String("agent", c.cfg.Executable))
 	if err := c.Stop(ctx); err != nil {
 		return err
 	}

cmd/opampsupervisor/supervisor/config/config.go

@@ -21,6 +21,7 @@ type Capabilities struct {
 	ReportsOwnMetrics      *bool `mapstructure:"reports_own_metrics"`
 	ReportsHealth          *bool `mapstructure:"reports_health"`
 	ReportsRemoteConfig    *bool `mapstructure:"reports_remote_config"`
+	AcceptsRestartCommand  *bool `mapstructure:"accepts_restart_command"`
 }
 
 type OpAMPServer struct {

cmd/opampsupervisor/supervisor/supervisor.go

@@ -61,6 +61,7 @@ type Supervisor struct {
 	healthCheckTicker  *backoff.Ticker
 	healthChecker      *healthchecker.HTTPHealthChecker
 	lastHealthCheckErr error
+	lastRestartErr     error
 
 	// Supervisor's own config.
 	config config.Supervisor
@@ -334,6 +335,10 @@ func (s *Supervisor) Capabilities() protobufs.AgentCapabilities {
 		if c.ReportsRemoteConfig != nil && *c.ReportsRemoteConfig {
 			supportedCapabilities |= protobufs.AgentCapabilities_AgentCapabilities_ReportsRemoteConfig
 		}
+
+		if c.AcceptsRestartCommand != nil && *c.AcceptsRestartCommand {
+			supportedCapabilities |= protobufs.AgentCapabilities_AgentCapabilities_AcceptsRestartCommand
+		}
 	}
 	return supportedCapabilities
 }
@@ -373,8 +378,7 @@ func (s *Supervisor) startOpAMP() error {
 			OnCommandFunc: func(command *protobufs.ServerToAgentCommand) error {
 				cmdType := command.GetType()
 				if *cmdType.Enum() == protobufs.CommandType_CommandType_Restart {
-					// TODO: https://github.com/open-telemetry/opentelemetry-collector-contrib/issues/21077
-					s.logger.Debug("Received restart command")
+					return s.handleRestartCommand()
 				}
 				return nil
 			},
@@ -604,6 +608,20 @@ func (s *Supervisor) recalcEffectiveConfig() (configChanged bool, err error) {
 	return configChanged, nil
 }
 
+func (s *Supervisor) handleRestartCommand() error {
+	err := s.commander.Restart(context.Background())
+	s.lastRestartErr = err
+	if err != nil {
+		s.logger.Error("Could not restart agent process", zap.Error(err))
+	}
+	return err
+}
+
+// Used in tests.
+func (s *Supervisor) LastRestartError() error {
+	return s.lastRestartErr
+}
+
 func (s *Supervisor) startAgent() {
 	err := s.commander.Start(context.Background())
 	if err != nil {

cmd/opampsupervisor/testdata/supervisor/supervisor_basic.yaml

@@ -9,6 +9,7 @@ capabilities:
   reports_health: true
   accepts_remote_config: true
   reports_remote_config: true
+  accepts_restart_command: true
 
 agent:
   executable: ../../bin/otelcontribcol_{{.goos}}_{{.goarch}}{{.extension}}