Remove anti-pattern of ignoring irrecoverable error channel

cmd/scaffold_test.go

@@ -64,7 +64,7 @@ func TestLoadSecretsEncryptionKey(t *testing.T) {
 // Test the components are started in the correct order, and are run serially
 func TestComponentsRunSerially(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 
 	nb := FlowNode("scaffold test")
 	nb.componentBuilder = component.NewComponentManagerBuilder()
@@ -165,7 +165,7 @@ func TestPostShutdown(t *testing.T) {
 
 func TestOverrideComponent(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 
 	nb := FlowNode("scaffold test")
 	nb.componentBuilder = component.NewComponentManagerBuilder()
@@ -225,7 +225,7 @@ func TestOverrideComponent(t *testing.T) {
 
 func TestOverrideModules(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 
 	nb := FlowNode("scaffold test")
 	nb.componentBuilder = component.NewComponentManagerBuilder()
@@ -602,7 +602,7 @@ func TestDependableComponentWaitForDependencies(t *testing.T) {
 
 func testDependableComponentWaitForDependencies(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 
 	nb := FlowNode("scaffold test")
 	nb.componentBuilder = component.NewComponentManagerBuilder()

consensus/hotstuff/eventloop/event_loop_test.go

@@ -52,7 +52,7 @@ func (s *EventLoopTestSuite) SetupTest() {
 
 	ctx, cancel := context.WithCancel(context.Background())
 	s.cancel = cancel
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(s.T(), ctx)
 
 	s.eventLoop.Start(signalerCtx)
 	unittest.RequireCloseBefore(s.T(), s.eventLoop.Ready(), 100*time.Millisecond, "event loop not started")
@@ -207,7 +207,7 @@ func TestEventLoop_Timeout(t *testing.T) {
 	eh.On("TimeoutChannel").Return(time.After(100 * time.Millisecond))
 
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 	eventLoop.Start(signalerCtx)
 
 	unittest.RequireCloseBefore(t, eventLoop.Ready(), 100*time.Millisecond, "event loop not stopped")
@@ -264,7 +264,7 @@ func TestReadyDoneWithStartTime(t *testing.T) {
 	}).Return(nil).Once()
 
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 	eventLoop.Start(signalerCtx)
 
 	unittest.RequireCloseBefore(t, eventLoop.Ready(), 100*time.Millisecond, "event loop not started")

consensus/hotstuff/integration/instance_test.go

@@ -548,13 +548,13 @@ func NewInstance(t *testing.T, options ...Option) *Instance {
 	return &in
 }
 
-func (in *Instance) Run() error {
+func (in *Instance) Run(t *testing.T) error {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer func() {
 		cancel()
 		<-util.AllDone(in.voteAggregator, in.timeoutAggregator)
 	}()
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 	in.voteAggregator.Start(signalerCtx)
 	in.timeoutAggregator.Start(signalerCtx)
 	<-util.AllReady(in.voteAggregator, in.timeoutAggregator)

consensus/hotstuff/integration/integration_test.go

@@ -32,7 +32,7 @@ func TestSingleInstance(t *testing.T) {
 	)
 
 	// run the event handler until we reach a stop condition
-	err := in.Run()
+	err := in.Run(t)
 	require.ErrorIs(t, err, errStopCondition, "should run until stop condition")
 
 	// check if forks and pacemaker are in expected view state
@@ -78,7 +78,7 @@ func TestThreeInstances(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.True(t, errors.Is(err, errStopCondition), "should run until stop condition")
 			wg.Done()
 		}(in)
@@ -151,7 +151,7 @@ func TestSevenInstances(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.True(t, errors.Is(err, errStopCondition), "should run until stop condition")
 			wg.Done()
 		}(in)

consensus/hotstuff/integration/liveness_test.go

@@ -73,7 +73,7 @@ func Test2TimeoutOutof7Instances(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.ErrorIs(t, err, errStopCondition)
 			wg.Done()
 		}(in)
@@ -141,7 +141,7 @@ func Test2TimeoutOutof4Instances(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.True(t, errors.Is(err, errStopCondition), "should run until stop condition")
 			wg.Done()
 		}(in)
@@ -209,7 +209,7 @@ func Test1TimeoutOutof5Instances(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.ErrorIs(t, err, errStopCondition)
 			wg.Done()
 		}(in)
@@ -306,7 +306,7 @@ func TestBlockDelayIsHigherThanTimeout(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.ErrorIs(t, err, errStopCondition)
 			wg.Done()
 		}(in)
@@ -388,7 +388,7 @@ func TestAsyncClusterStartup(t *testing.T) {
 	for _, in := range instances {
 		wg.Add(1)
 		go func(in *Instance) {
-			err := in.Run()
+			err := in.Run(t)
 			require.ErrorIs(t, err, errStopCondition)
 			wg.Done()
 		}(in)

consensus/hotstuff/timeoutaggregator/timeout_aggregator_test.go

@@ -55,7 +55,7 @@ func (s *TimeoutAggregatorTestSuite) SetupTest() {
 	require.NoError(s.T(), err)
 
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(s.T(), ctx)
 	s.stopAggregator = cancel
 	s.aggregator.Start(signalerCtx)
 	unittest.RequireCloseBefore(s.T(), s.aggregator.Ready(), 100*time.Millisecond, "should close before timeout")

consensus/hotstuff/voteaggregator/vote_aggregator.go

@@ -18,6 +18,7 @@ import (
 	"github.com/onflow/flow-go/module/irrecoverable"
 	"github.com/onflow/flow-go/module/mempool"
 	"github.com/onflow/flow-go/module/metrics"
+	"github.com/onflow/flow-go/module/util"
 )
 
 // defaultVoteAggregatorWorkers number of workers to dispatch events for vote aggregators
@@ -99,24 +100,36 @@ func NewVoteAggregator(
 			aggregator.queuedMessagesProcessingLoop(ctx)
 		})
 	}
-	componentBuilder.AddWorker(func(_ irrecoverable.SignalerContext, ready component.ReadyFunc) {
+	componentBuilder.AddWorker(func(parentCtx irrecoverable.SignalerContext, ready component.ReadyFunc) {
 		// create new context which is not connected to parent
 		// we need to ensure that our internal workers stop before asking
 		// vote collectors to stop. We want to avoid delivering events to already stopped vote collectors
 		ctx, cancel := context.WithCancel(context.Background())
-		signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+		signalerCtx, errCh := irrecoverable.WithSignaler(ctx)
+
 		// start vote collectors
 		collectors.Start(signalerCtx)
-		<-collectors.Ready()
 
-		ready()
+		// Handle the component lifecycle in a separate goroutine so we can capture any errors
+		// thrown during initialization in the main goroutine.
+		go func() {
+			if err := util.WaitClosed(parentCtx, collectors.Ready()); err == nil {
+				// only signal ready when collectors are ready, but always handle shutdown
+				ready()
+			}
 
-		// wait for internal workers to stop
-		wg.Wait()
-		// signal vote collectors to stop
-		cancel()
-		// wait for it to stop
-		<-collectors.Done()
+			// wait for internal workers to stop, then signal vote collectors to stop
+			wg.Wait()
+			cancel()
+		}()
+
+		// since we are breaking the connection between parentCtx and signalerCtx, we need to
+		// explicitly rethrow any errors from signalerCtx to parentCtx, otherwise they are dropped.
+		// Handle errors in the main worker goroutine to guarantee that they are rethrown to the parent
+		// before the component is marked done.
+		if err := util.WaitError(errCh, collectors.Done()); err != nil {
+			parentCtx.Throw(err)
+		}
 	})
 	componentBuilder.AddWorker(func(ctx irrecoverable.SignalerContext, ready component.ReadyFunc) {
 		ready()

consensus/integration/nodes_test.go

@@ -226,7 +226,7 @@ func createNodes(t *testing.T, participants *ConsensusParticipants, rootSnapshot
 
 	// create a context which will be used for all nodes
 	ctx, cancel := context.WithCancel(context.Background())
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 
 	// create a function to return which the test case can use to run the nodes for some maximum duration
 	// and gracefully stop after.

engine/access/access_test.go

@@ -262,7 +262,7 @@ func (suite *Suite) TestSendExpiredTransaction() {
 			Return(referenceBlock, nil).
 			Twice()
 
-		//Advancing final state to expire ref block
+		// Advancing final state to expire ref block
 		suite.finalizedBlock = latestBlock
 
 		req := &accessproto.SendTransactionRequest{
@@ -728,7 +728,7 @@ func (suite *Suite) TestGetSealedTransaction() {
 		background, cancel := context.WithCancel(context.Background())
 		defer cancel()
 
-		ctx, _ := irrecoverable.WithSignaler(background)
+		ctx := irrecoverable.NewMockSignalerContext(suite.T(), background)
 		ingestEng.Start(ctx)
 		<-ingestEng.Ready()
 
@@ -1169,7 +1169,7 @@ func (suite *Suite) TestExecuteScript() {
 		require.NoError(suite.T(), err)
 		err = db.Update(operation.IndexBlockHeight(lastBlock.Header.Height, lastBlock.ID()))
 		require.NoError(suite.T(), err)
-		//update latest sealed block
+		// update latest sealed block
 		suite.sealedBlock = lastBlock.Header
 		// create execution receipts for each of the execution node and the last block
 		executionReceipts := unittest.ReceiptsForBlockFixture(lastBlock, identities.NodeIDs())

engine/collection/ingest/engine_test.go

@@ -297,7 +297,7 @@ func (suite *Suite) TestComponentShutdown() {
 
 	// start then shut down the engine
 	parentCtx, cancel := context.WithCancel(context.Background())
-	ctx, _ := irrecoverable.WithSignaler(parentCtx)
+	ctx := irrecoverable.NewMockSignalerContext(suite.T(), parentCtx)
 	suite.engine.Start(ctx)
 	unittest.AssertClosesBefore(suite.T(), suite.engine.Ready(), 10*time.Millisecond)
 	cancel()

engine/common/worker/worker_builder_test.go

@@ -38,7 +38,7 @@ func TestWorkerPool_SingleEvent_SingleWorker(t *testing.T) {
 
 	cancelCtx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	ctx, _ := irrecoverable.WithSignaler(cancelCtx)
+	ctx := irrecoverable.NewMockSignalerContext(t, cancelCtx)
 	cm := component.NewComponentManagerBuilder().
 		AddWorker(pool.WorkerLogic()).
 		Build()
@@ -75,7 +75,7 @@ func TestWorkerBuilder_UnhappyPaths(t *testing.T) {
 
 	cancelCtx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	ctx, _ := irrecoverable.WithSignaler(cancelCtx)
+	ctx := irrecoverable.NewMockSignalerContext(t, cancelCtx)
 	cm := component.NewComponentManagerBuilder().
 		AddWorker(pool.WorkerLogic()).
 		Build()
@@ -138,7 +138,7 @@ func TestWorkerPool_TwoWorkers_ConcurrentEvents(t *testing.T) {
 
 	cancelCtx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	ctx, _ := irrecoverable.WithSignaler(cancelCtx)
+	ctx := irrecoverable.NewMockSignalerContext(t, cancelCtx)
 	cm := component.NewComponentManagerBuilder().
 		AddWorker(pool.WorkerLogic()).
 		AddWorker(pool.WorkerLogic()).

engine/consensus/ingestion/engine_test.go

@@ -55,7 +55,7 @@ func (s *IngestionSuite) SetupTest() {
 
 	ctx, cancel := context.WithCancel(context.Background())
 	s.cancel = cancel
-	signalerCtx, _ := irrecoverable.WithSignaler(ctx)
+	signalerCtx := irrecoverable.NewMockSignalerContext(s.T(), ctx)
 
 	metrics := metrics.NewNoopCollector()
 	ingest, err := New(unittest.Logger(), metrics, s.net, me, s.core)

engine/execution/execution_test.go

@@ -67,7 +67,7 @@ func TestExecutionFlow(t *testing.T) {
 
 	ctx, cancel := context.WithCancel(context.Background())
 	unittest.RequireReturnsBefore(t, func() {
-		exeNode.Ready(ctx)
+		exeNode.Ready(t, ctx)
 	}, 1*time.Second, "could not start execution node on time")
 	defer exeNode.Done(cancel)
 
@@ -400,7 +400,7 @@ func TestFailedTxWillNotChangeStateCommitment(t *testing.T) {
 
 	ctx, cancel := context.WithCancel(context.Background())
 	unittest.RequireReturnsBefore(t, func() {
-		exe1Node.Ready(ctx)
+		exe1Node.Ready(t, ctx)
 	}, 1*time.Second, "could not start execution node on time")
 	defer exe1Node.Done(cancel)
 
@@ -495,7 +495,7 @@ func TestFailedTxWillNotChangeStateCommitment(t *testing.T) {
 	scExe1Block2, err := exe1Node.ExecutionState.StateCommitmentByBlockID(block2.ID())
 	assert.NoError(t, err)
 	// TODO this is no longer valid because the system chunk can change the state
-	//assert.Equal(t, scExe1Block1, scExe1Block2)
+	// assert.Equal(t, scExe1Block1, scExe1Block2)
 	_ = scExe1Block2
 
 	collectionEngine.AssertExpectations(t)
@@ -567,7 +567,7 @@ func TestBroadcastToMultipleVerificationNodes(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 
 	unittest.RequireReturnsBefore(t, func() {
-		exeNode.Ready(ctx)
+		exeNode.Ready(t, ctx)
 	}, 1*time.Second, "could not start execution node on time")
 	defer exeNode.Done(cancel)
 

engine/execution/ingestion/core_test.go

@@ -37,7 +37,7 @@ func TestInogestionCoreExecuteBlock(t *testing.T) {
 
 	// start the core
 	ctx, cancel := context.WithCancel(context.Background())
-	irrecoverableCtx, _ := irrecoverable.WithSignaler(ctx)
+	irrecoverableCtx := irrecoverable.NewMockSignalerContext(t, ctx)
 	core.Start(irrecoverableCtx)
 	<-core.Ready()
 	defer func() {

engine/execution/provider/engine_test.go

@@ -47,7 +47,7 @@ func TestProviderEngine_onChunkDataRequest(t *testing.T) {
 
 		cancelCtx, cancel := context.WithCancel(context.Background())
 		defer cancel()
-		ctx, _ := irrecoverable.WithSignaler(cancelCtx)
+		ctx := irrecoverable.NewMockSignalerContext(t, cancelCtx)
 		e.Start(ctx)
 		// submit using non-existing origin ID
 		unittest.RequireCloseBefore(t, e.Ready(), 100*time.Millisecond, "could not start engine")
@@ -96,7 +96,7 @@ func TestProviderEngine_onChunkDataRequest(t *testing.T) {
 
 		cancelCtx, cancel := context.WithCancel(context.Background())
 		defer cancel()
-		ctx, _ := irrecoverable.WithSignaler(cancelCtx)
+		ctx := irrecoverable.NewMockSignalerContext(t, cancelCtx)
 		e.Start(ctx)
 		// submit using non-existing origin ID
 		unittest.RequireCloseBefore(t, e.Ready(), 100*time.Millisecond, "could not start engine")

engine/testutil/mock/nodes.go

@@ -228,11 +228,11 @@ type ExecutionNode struct {
 	StorehouseEnabled   bool
 }
 
-func (en ExecutionNode) Ready(ctx context.Context) {
+func (en ExecutionNode) Ready(t *testing.T, ctx context.Context) {
 	// TODO: receipt engine has been migrated to the new component interface, hence
 	// is using Start. Other engines' startup should be refactored once migrated to
 	// new interface.
-	irctx, _ := irrecoverable.WithSignaler(ctx)
+	irctx := irrecoverable.NewMockSignalerContext(t, ctx)
 	en.ReceiptsEngine.Start(irctx)
 	en.IngestionEngine.Start(irctx)
 	en.FollowerCore.Start(irctx)

module/component/component_manager_test.go

@@ -639,7 +639,7 @@ func TestComponentManagerShutdown(t *testing.T) {
 		}).Build()
 
 	parent, cancel := context.WithCancel(context.Background())
-	ctx, _ := irrecoverable.WithSignaler(parent)
+	ctx := irrecoverable.NewMockSignalerContext(t, parent)
 
 	mgr.Start(ctx)
 	unittest.AssertClosesBefore(t, mgr.Ready(), 10*time.Millisecond)