backport of commit 5e111f7b31

internal/cmd/config/config.go

@@ -210,6 +210,12 @@ type Controller struct {
 	LivenessTimeToStale         any           `hcl:"liveness_time_to_stale"`
 	LivenessTimeToStaleDuration time.Duration `hcl:"-"`
 
+	// TODO: This isn't documented (on purpose) because the right place for this
+	// is central configuration so you can't drift across controllers and workers
+	// but we don't have that yet.
+	GetDownstreamWorkersTimeout         any           `hcl:"get_downstream_workers_timeout"`
+	GetDownstreamWorkersTimeoutDuration time.Duration `hcl:"-"`
+
 	// SchedulerRunJobInterval is the time interval between waking up the
 	// scheduler to run pending jobs.
 	//
@@ -279,6 +285,13 @@ type Worker struct {
 	StatusCallTimeout         any           `hcl:"status_call_timeout"`
 	StatusCallTimeoutDuration time.Duration `hcl:"-"`
 
+	// GetDownstreamWorkersTimeout represents the period of time (as a duration) timeout
+	// for GetDownstreamWorkers call in DownstreamWorkerTicker
+	//
+	// TODO: This is currently not documented and considered internal.
+	GetDownstreamWorkersTimeout         any           `hcl:"get_downstream_workers_timeout"`
+	GetDownstreamWorkersTimeoutDuration time.Duration `hcl:"-"`
+
 	// SuccessfulStatusGracePeriod represents the period of time (as a duration)
 	// that the worker will wait before disconnecting connections if it cannot
 	// successfully complete a status report to a controller. This cannot be
@@ -648,6 +661,21 @@ func Parse(d string) (*Config, error) {
 			return nil, errors.New("Controller liveness time to stale value is negative")
 		}
 
+		getDownstreamWorkersTimeout := result.Controller.GetDownstreamWorkersTimeout
+		if util.IsNil(getDownstreamWorkersTimeout) {
+			getDownstreamWorkersTimeout = os.Getenv("BOUNDARY_CONTROLLER_GET_DOWNSTREAM_WORKERS_TIMEOUT")
+		}
+		if getDownstreamWorkersTimeout != nil {
+			t, err := parseutil.ParseDurationSecond(getDownstreamWorkersTimeout)
+			if err != nil {
+				return result, err
+			}
+			result.Controller.GetDownstreamWorkersTimeoutDuration = t
+		}
+		if result.Controller.GetDownstreamWorkersTimeoutDuration < 0 {
+			return nil, errors.New("GetDownstreamWorkersTimeoutDuration value is negative")
+		}
+
 		if result.Controller.MaxPageSizeRaw != nil {
 			switch t := result.Controller.MaxPageSizeRaw.(type) {
 			case string:
@@ -785,6 +813,21 @@ func Parse(d string) (*Config, error) {
 			return nil, errors.New("Status call timeout value is negative")
 		}
 
+		getDownstreamWorkersTimeoutDuration := result.Worker.GetDownstreamWorkersTimeout
+		if util.IsNil(getDownstreamWorkersTimeoutDuration) {
+			getDownstreamWorkersTimeoutDuration = os.Getenv("BOUNDARY_WORKER_GET_DOWNSTREAM_WORKERS_TIMEOUT")
+		}
+		if getDownstreamWorkersTimeoutDuration != nil {
+			t, err := parseutil.ParseDurationSecond(getDownstreamWorkersTimeoutDuration)
+			if err != nil {
+				return result, fmt.Errorf("error parsing get_downstream_worker_call_timeout: %w", err)
+			}
+			result.Worker.GetDownstreamWorkersTimeoutDuration = t
+		}
+		if result.Worker.GetDownstreamWorkersTimeoutDuration < 0 {
+			return nil, errors.New("GetDownstreamWorkersTimeoutDuration call timeout value is negative")
+		}
+
 		successfulStatusGracePeriod := result.Worker.SuccessfulStatusGracePeriod
 		if util.IsNil(successfulStatusGracePeriod) {
 			successfulStatusGracePeriod = os.Getenv("BOUNDARY_WORKER_SUCCESSFUL_STATUS_GRACE_PERIOD")

internal/daemon/controller/controller.go

@@ -10,6 +10,7 @@ import (
 	"strings"
 	"sync"
 	"sync/atomic"
+	"time"
 
 	"github.com/hashicorp/boundary/internal/alias"
 	talias "github.com/hashicorp/boundary/internal/alias/target"
@@ -91,7 +92,7 @@ var (
 	downstreamReceiverFactory func() downstreamReceiver
 
 	downstreamersFactory           func(context.Context, string, string) (common.Downstreamers, error)
-	downstreamWorkersTickerFactory func(context.Context, string, string, common.Downstreamers, downstreamReceiver) (downstreamWorkersTicker, error)
+	downstreamWorkersTickerFactory func(context.Context, string, string, common.Downstreamers, downstreamReceiver, time.Duration) (downstreamWorkersTicker, error)
 	commandClientFactory           func(context.Context, *Controller) error
 	extControllerFactory           func(ctx context.Context, c *Controller, r db.Reader, w db.Writer, kms *kms.Kms) (intglobals.ControllerExtension, error)
 )
@@ -121,8 +122,9 @@ type Controller struct {
 
 	// Timing variables. These are atomics for SIGHUP support, and are int64
 	// because they are casted to time.Duration.
-	workerStatusGracePeriod *atomic.Int64
-	livenessTimeToStale     *atomic.Int64
+	workerStatusGracePeriod     *atomic.Int64
+	livenessTimeToStale         *atomic.Int64
+	getDownstreamWorkersTimeout *atomic.Pointer[time.Duration]
 
 	apiGrpcServer         *grpc.Server
 	apiGrpcServerListener grpcServerListener
@@ -176,18 +178,19 @@ func New(ctx context.Context, conf *Config) (*Controller, error) {
 	metric.InitializeApiCollectors(conf.PrometheusRegisterer)
 	ratelimit.InitializeMetrics(conf.PrometheusRegisterer)
 	c := &Controller{
-		conf:                    conf,
-		logger:                  conf.Logger.Named("controller"),
-		started:                 ua.NewBool(false),
-		tickerWg:                new(sync.WaitGroup),
-		schedulerWg:             new(sync.WaitGroup),
-		workerAuthCache:         new(sync.Map),
-		workerStatusUpdateTimes: new(sync.Map),
-		enabledPlugins:          conf.Server.EnabledPlugins,
-		apiListeners:            make([]*base.ServerListener, 0),
-		downstreamConnManager:   cluster.NewDownstreamManager(),
-		workerStatusGracePeriod: new(atomic.Int64),
-		livenessTimeToStale:     new(atomic.Int64),
+		conf:                        conf,
+		logger:                      conf.Logger.Named("controller"),
+		started:                     ua.NewBool(false),
+		tickerWg:                    new(sync.WaitGroup),
+		schedulerWg:                 new(sync.WaitGroup),
+		workerAuthCache:             new(sync.Map),
+		workerStatusUpdateTimes:     new(sync.Map),
+		enabledPlugins:              conf.Server.EnabledPlugins,
+		apiListeners:                make([]*base.ServerListener, 0),
+		downstreamConnManager:       cluster.NewDownstreamManager(),
+		workerStatusGracePeriod:     new(atomic.Int64),
+		livenessTimeToStale:         new(atomic.Int64),
+		getDownstreamWorkersTimeout: new(atomic.Pointer[time.Duration]),
 	}
 
 	if downstreamReceiverFactory != nil {
@@ -238,6 +241,15 @@ func New(ctx context.Context, conf *Config) (*Controller, error) {
 		c.livenessTimeToStale.Store(int64(conf.RawConfig.Controller.LivenessTimeToStaleDuration))
 	}
 
+	switch conf.RawConfig.Controller.GetDownstreamWorkersTimeoutDuration {
+	case 0:
+		to := server.DefaultLiveness
+		c.getDownstreamWorkersTimeout.Store(&to)
+	default:
+		to := conf.RawConfig.Controller.GetDownstreamWorkersTimeoutDuration
+		c.getDownstreamWorkersTimeout.Store(&to)
+	}
+
 	clusterListeners := make([]*base.ServerListener, 0)
 	for i := range conf.Listeners {
 		l := conf.Listeners[i]
@@ -579,7 +591,7 @@ func (c *Controller) Start() error {
 		// we'll use "root" to designate that this is the root of the graph (aka
 		// a controller)
 		boundVer := version.Get().VersionNumber()
-		dswTicker, err := downstreamWorkersTickerFactory(c.baseContext, "root", boundVer, c.downstreamWorkers, c.downstreamConns)
+		dswTicker, err := downstreamWorkersTickerFactory(c.baseContext, "root", boundVer, c.downstreamWorkers, c.downstreamConns, *c.getDownstreamWorkersTimeout.Load())
 		if err != nil {
 			return fmt.Errorf("error creating downstream workers ticker: %w", err)
 		}

internal/daemon/worker/worker.go

@@ -187,8 +187,9 @@ type Worker struct {
 
 	// Timing variables. These are atomics for SIGHUP support, and are int64
 	// because they are casted to time.Duration.
-	successfulStatusGracePeriod *atomic.Int64
-	statusCallTimeoutDuration   *atomic.Int64
+	successfulStatusGracePeriod         *atomic.Int64
+	statusCallTimeoutDuration           *atomic.Int64
+	getDownstreamWorkersTimeoutDuration *atomic.Pointer[time.Duration]
 
 	// AuthRotationNextRotation is useful in tests to understand how long to
 	// sleep
@@ -222,17 +223,18 @@ func New(ctx context.Context, conf *Config) (*Worker, error) {
 		lastStatusSuccess:      new(atomic.Value),
 		controllerMultihopConn: new(atomic.Value),
 		// controllerUpstreamMsgConn:   new(atomic.Value),
-		tags:                        new(atomic.Value),
-		updateTags:                  ua.NewBool(false),
-		nonceFn:                     base62.Random,
-		WorkerAuthCurrentKeyId:      new(ua.String),
-		operationalState:            new(atomic.Value),
-		downstreamConnManager:       cluster.NewDownstreamManager(),
-		localStorageState:           new(atomic.Value),
-		successfulStatusGracePeriod: new(atomic.Int64),
-		statusCallTimeoutDuration:   new(atomic.Int64),
-		upstreamConnectionState:     new(atomic.Value),
-		downstreamWorkers:           new(atomic.Pointer[downstreamersContainer]),
+		tags:                                new(atomic.Value),
+		updateTags:                          ua.NewBool(false),
+		nonceFn:                             base62.Random,
+		WorkerAuthCurrentKeyId:              new(ua.String),
+		operationalState:                    new(atomic.Value),
+		downstreamConnManager:               cluster.NewDownstreamManager(),
+		localStorageState:                   new(atomic.Value),
+		successfulStatusGracePeriod:         new(atomic.Int64),
+		statusCallTimeoutDuration:           new(atomic.Int64),
+		getDownstreamWorkersTimeoutDuration: new(atomic.Pointer[time.Duration]),
+		upstreamConnectionState:             new(atomic.Value),
+		downstreamWorkers:                   new(atomic.Pointer[downstreamersContainer]),
 	}
 
 	w.operationalState.Store(server.UnknownOperationalState)
@@ -334,6 +336,14 @@ func New(ctx context.Context, conf *Config) (*Worker, error) {
 	default:
 		w.statusCallTimeoutDuration.Store(int64(conf.RawConfig.Worker.StatusCallTimeoutDuration))
 	}
+	switch conf.RawConfig.Worker.GetDownstreamWorkersTimeoutDuration {
+	case 0:
+		to := server.DefaultLiveness
+		w.getDownstreamWorkersTimeoutDuration.Store(&to)
+	default:
+		to := conf.RawConfig.Worker.GetDownstreamWorkersTimeoutDuration
+		w.getDownstreamWorkersTimeoutDuration.Store(&to)
+	}
 	// FIXME: This is really ugly, but works.
 	session.CloseCallTimeout.Store(w.successfulStatusGracePeriod.Load())