Change Queue interface to return a callback instead of an index (#12230)

This change will allow to simplify significantly the implementation for
when queue is disabled and caller must block until the batch is
constructed and processed downstream

Updates
#8122

It is ok to have this as a breaking change since Queue is still
experimental.

Signed-off-by: Bogdan Drutu <[email protected]>

Author: bogdandrutu

.chloggen/done-callback.yaml

@@ -0,0 +1,25 @@
+# Use this changelog template to create an entry for release notes.
+
+# One of 'breaking', 'deprecation', 'new_component', 'enhancement', 'bug_fix'
+change_type: breaking
+
+# The name of the component, or a single word describing the area of concern, (e.g. otlpreceiver)
+component: exporterqueue
+
+# A brief description of the change.  Surround your text with quotes ("") if it needs to start with a backtick (`).
+note: Change Queue interface to return a callback instead of an index
+
+# One or more tracking issues or pull requests related to the change
+issues: [8122]
+
+# (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:
+
+# 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: [api]

exporter/exporterqueue/bounded_memory_queue.go

@@ -11,6 +11,8 @@ import (
 	"go.opentelemetry.io/collector/component"
 )
 
+var noopDone DoneCallback = func(error) {}
+
 // boundedMemoryQueue implements a producer-consumer exchange similar to a ring buffer queue,
 // where the queue is bounded and if it fills up due to slow consumers, the new items written by
 // the producer are dropped.
@@ -34,11 +36,7 @@ func newBoundedMemoryQueue[T any](set memoryQueueSettings[T]) Queue[T] {
 	}
 }
 
-func (q *boundedMemoryQueue[T]) Read(context.Context) (uint64, context.Context, T, bool) {
+func (q *boundedMemoryQueue[T]) Read(context.Context) (context.Context, T, DoneCallback, bool) {
 	ctx, req, ok := q.sizedQueue.pop()
-	return 0, ctx, req, ok
+	return ctx, req, noopDone, ok
 }
-
-// OnProcessingFinished should be called to remove the item of the given index from the queue once processing is finished.
-// For in memory queue, this function is noop.
-func (q *boundedMemoryQueue[T]) OnProcessingFinished(uint64, error) {}

exporter/exporterqueue/bounded_memory_queue_test.go

@@ -183,11 +183,11 @@ func TestZeroSizeNoConsumers(t *testing.T) {
 }
 
 func consume[T any](q Queue[T], consumeFunc func(context.Context, T) error) bool {
-	index, ctx, req, ok := q.Read(context.Background())
+	ctx, req, done, ok := q.Read(context.Background())
 	if !ok {
 		return false
 	}
-	q.OnProcessingFinished(index, consumeFunc(ctx, req))
+	done(consumeFunc(ctx, req))
 	return true
 }
 
@@ -203,11 +203,11 @@ func newAsyncConsumer[T any](q Queue[T], numConsumers int, consumeFunc func(cont
 		go func() {
 			defer ac.stopWG.Done()
 			for {
-				index, ctx, req, ok := q.Read(context.Background())
+				ctx, req, done, ok := q.Read(context.Background())
 				if !ok {
 					return
 				}
-				q.OnProcessingFinished(index, consumeFunc(ctx, req))
+				done(consumeFunc(ctx, req))
 			}
 		}()
 	}

exporter/exporterqueue/persistent_queue.go

@@ -273,14 +273,14 @@ func (pq *persistentQueue[T]) putInternal(ctx context.Context, req T) error {
 	return nil
 }
 
-func (pq *persistentQueue[T]) Read(ctx context.Context) (uint64, context.Context, T, bool) {
+func (pq *persistentQueue[T]) Read(ctx context.Context) (context.Context, T, DoneCallback, bool) {
 	pq.mu.Lock()
 	defer pq.mu.Unlock()
 
 	for {
 		if pq.stopped {
 			var req T
-			return 0, context.Background(), req, false
+			return context.Background(), req, nil, false
 		}
 
 		// Read until either a successful retrieved element or no more elements in the storage.
@@ -301,7 +301,7 @@ func (pq *persistentQueue[T]) Read(ctx context.Context) (uint64, context.Context
 				}
 				pq.hasMoreSpace.Signal()
 
-				return index, context.Background(), req, true
+				return context.Background(), req, func(processErr error) { pq.onDone(index, processErr) }, true
 			}
 		}
 
@@ -312,7 +312,7 @@ func (pq *persistentQueue[T]) Read(ctx context.Context) (uint64, context.Context
 }
 
 // getNextItem pulls the next available item from the persistent storage along with its index. Once processing is
-// finished, the index should be called with OnProcessingFinished to clean up the storage. If no new item is available,
+// finished, the index should be called with onDone to clean up the storage. If no new item is available,
 // returns false.
 func (pq *persistentQueue[T]) getNextItem(ctx context.Context) (uint64, T, bool) {
 	index := pq.readIndex
@@ -347,8 +347,8 @@ func (pq *persistentQueue[T]) getNextItem(ctx context.Context) (uint64, T, bool)
 	return index, request, true
 }
 
-// OnProcessingFinished should be called to remove the item of the given index from the queue once processing is finished.
-func (pq *persistentQueue[T]) OnProcessingFinished(index uint64, consumeErr error) {
+// onDone should be called to remove the item of the given index from the queue once processing is finished.
+func (pq *persistentQueue[T]) onDone(index uint64, consumeErr error) {
 	// Delete the item from the persistent storage after it was processed.
 	pq.mu.Lock()
 	// Always unref client even if the consumer is shutdown because we always ref it for every valid request.

exporter/exporterqueue/persistent_queue_test.go

@@ -659,19 +659,19 @@ func TestPersistentQueue_CurrentlyProcessedItems(t *testing.T) {
 	requireCurrentlyDispatchedItemsEqual(t, ps, []uint64{})
 
 	// Takes index 0 in process.
-	_, _, readReq, found := ps.Read(context.Background())
+	_, readReq, _, found := ps.Read(context.Background())
 	require.True(t, found)
 	assert.Equal(t, req, readReq)
 	requireCurrentlyDispatchedItemsEqual(t, ps, []uint64{0})
 
 	// This takes item 1 to process.
-	secondIndex, _, secondReadReq, found := ps.Read(context.Background())
+	_, secondReadReq, secondDone, found := ps.Read(context.Background())
 	require.True(t, found)
 	assert.Equal(t, req, secondReadReq)
 	requireCurrentlyDispatchedItemsEqual(t, ps, []uint64{0, 1})
 
 	// Lets mark item 1 as finished, it will remove it from the currently dispatched items list.
-	ps.OnProcessingFinished(secondIndex, nil)
+	secondDone(nil)
 	requireCurrentlyDispatchedItemsEqual(t, ps, []uint64{0})
 
 	// Reload the storage. Since items 0 was not finished, this should be re-enqueued at the end.
@@ -910,12 +910,12 @@ func TestPersistentQueue_ShutdownWhileConsuming(t *testing.T) {
 
 	require.NoError(t, ps.Offer(context.Background(), uint64(50)))
 
-	index, _, _, ok := ps.Read(context.Background())
+	_, _, done, ok := ps.Read(context.Background())
 	require.True(t, ok)
 	assert.False(t, ps.client.(*storagetest.MockStorageClient).IsClosed())
 	require.NoError(t, ps.Shutdown(context.Background()))
 	assert.False(t, ps.client.(*storagetest.MockStorageClient).IsClosed())
-	ps.OnProcessingFinished(index, nil)
+	done(nil)
 	assert.True(t, ps.client.(*storagetest.MockStorageClient).IsClosed())
 }
 

exporter/exporterqueue/queue.go

@@ -12,11 +12,18 @@ import (
 	"go.opentelemetry.io/collector/pipeline"
 )
 
-// ErrQueueIsFull is the error returned when an item is offered to the Queue and the queue is full.
+// ErrQueueIsFull is the error returned when an item is offered to the Queue and the queue is full and setup to
+// not block.
 // Experimental: This API is at the early stage of development and may change without backward compatibility
 // until https://github.com/open-telemetry/opentelemetry-collector/issues/8122 is resolved.
 var ErrQueueIsFull = errors.New("sending queue is full")
 
+// DoneCallback represents the callback that will be called when the read request is completely processed by the
+// downstream components.
+// Experimental: This API is at the early stage of development and may change without backward compatibility
+// until https://github.com/open-telemetry/opentelemetry-collector/issues/8122 is resolved.
+type DoneCallback func(processErr error)
+
 // Queue defines a producer-consumer exchange which can be backed by e.g. the memory-based ring buffer queue
 // (boundedMemoryQueue) or via a disk-based queue (persistentQueue)
 // Experimental: This API is at the early stage of development and may change without backward compatibility
@@ -31,13 +38,11 @@ type Queue[T any] interface {
 	Size() int64
 	// Capacity returns the capacity of the queue.
 	Capacity() int64
-	// Read pulls the next available item from the queue along with its index. Once processing is
-	// finished, the index should be called with OnProcessingFinished to clean up the storage.
+	// Read pulls the next available item from the queue along with its done callback. Once processing is
+	// finished, the done callback must be called to clean up the storage.
 	// The function blocks until an item is available or if the queue is stopped.
-	// Returns false if reading failed or if the queue is stopped.
-	Read(context.Context) (uint64, context.Context, T, bool)
-	// OnProcessingFinished should be called to remove the item of the given index from the queue once processing is finished.
-	OnProcessingFinished(index uint64, consumeErr error)
+	// If the queue is stopped returns false, otherwise true.
+	Read(context.Context) (context.Context, T, DoneCallback, bool)
 }
 
 // Settings defines settings for creating a queue.

exporter/internal/queue/batcher.go

@@ -14,9 +14,9 @@ import (
 )
 
 type batch struct {
-	ctx     context.Context
-	req     internal.Request
-	idxList []uint64
+	ctx   context.Context
+	req   internal.Request
+	dones []exporterqueue.DoneCallback
 }
 
 // Batcher is in charge of reading items from the queue and send them out asynchronously.
@@ -68,16 +68,16 @@ func newBaseBatcher(batchCfg exporterbatcher.Config,
 }
 
 // flush starts a goroutine that calls exportFunc. It blocks until a worker is available if necessary.
-func (qb *BaseBatcher) flush(batchToFlush batch) {
+func (qb *BaseBatcher) flush(ctx context.Context, req internal.Request, dones []exporterqueue.DoneCallback) {
 	qb.stopWG.Add(1)
 	if qb.workerPool != nil {
 		<-qb.workerPool
 	}
 	go func() {
 		defer qb.stopWG.Done()
-		err := qb.exportFunc(batchToFlush.ctx, batchToFlush.req)
-		for _, idx := range batchToFlush.idxList {
-			qb.queue.OnProcessingFinished(idx, err)
+		err := qb.exportFunc(ctx, req)
+		for _, done := range dones {
+			done(err)
 		}
 		if qb.workerPool != nil {
 			qb.workerPool <- true

exporter/internal/queue/consumers.go

@@ -36,12 +36,11 @@ func (qc *Consumers[T]) Start(_ context.Context, _ component.Host) error {
 			startWG.Done()
 			defer qc.stopWG.Done()
 			for {
-				index, ctx, req, ok := qc.queue.Read(context.Background())
+				ctx, req, done, ok := qc.queue.Read(context.Background())
 				if !ok {
 					return
 				}
-				consumeErr := qc.consumeFunc(ctx, req)
-				qc.queue.OnProcessingFinished(index, consumeErr)
+				done(qc.consumeFunc(ctx, req))
 			}
 		}()
 	}

exporter/internal/queue/default_batcher.go

@@ -10,6 +10,7 @@ import (
 	"time"
 
 	"go.opentelemetry.io/collector/component"
+	"go.opentelemetry.io/collector/exporter/exporterqueue"
 	"go.opentelemetry.io/collector/exporter/internal"
 )
 
@@ -35,7 +36,7 @@ func (qb *DefaultBatcher) startReadingFlushingGoroutine() {
 		defer qb.stopWG.Done()
 		for {
 			// Read() blocks until the queue is non-empty or until the queue is stopped.
-			idx, ctx, req, ok := qb.queue.Read(context.Background())
+			ctx, req, done, ok := qb.queue.Read(context.Background())
 			if !ok {
 				qb.shutdownCh <- true
 				return
@@ -54,7 +55,7 @@ func (qb *DefaultBatcher) startReadingFlushingGoroutine() {
 				}
 
 				if mergeSplitErr != nil || reqList == nil {
-					qb.queue.OnProcessingFinished(idx, mergeSplitErr)
+					done(mergeSplitErr)
 					qb.currentBatchMu.Unlock()
 					continue
 				}
@@ -64,42 +65,38 @@ func (qb *DefaultBatcher) startReadingFlushingGoroutine() {
 					qb.currentBatch = nil
 					qb.currentBatchMu.Unlock()
 					for i := 0; i < len(reqList); i++ {
-						qb.flush(batch{
-							req:     reqList[i],
-							ctx:     ctx,
-							idxList: []uint64{idx},
-						})
+						qb.flush(ctx, reqList[i], []exporterqueue.DoneCallback{done})
 						// TODO: handle partial failure
 					}
 					qb.resetTimer()
 				} else {
 					qb.currentBatch = &batch{
-						req:     reqList[0],
-						ctx:     ctx,
-						idxList: []uint64{idx},
+						req:   reqList[0],
+						ctx:   ctx,
+						dones: []exporterqueue.DoneCallback{done},
 					}
 					qb.currentBatchMu.Unlock()
 				}
 			} else {
 				if qb.currentBatch == nil || qb.currentBatch.req == nil {
 					qb.resetTimer()
 					qb.currentBatch = &batch{
-						req:     req,
-						ctx:     ctx,
-						idxList: []uint64{idx},
+						req:   req,
+						ctx:   ctx,
+						dones: []exporterqueue.DoneCallback{done},
 					}
 				} else {
 					// TODO: consolidate implementation for the cases where MaxSizeConfig is specified and the case where it is not specified
 					mergedReq, mergeErr := qb.currentBatch.req.MergeSplit(qb.currentBatch.ctx, qb.batchCfg.MaxSizeConfig, req)
 					if mergeErr != nil {
-						qb.queue.OnProcessingFinished(idx, mergeErr)
+						done(mergeErr)
 						qb.currentBatchMu.Unlock()
 						continue
 					}
 					qb.currentBatch = &batch{
-						req:     mergedReq[0],
-						ctx:     qb.currentBatch.ctx,
-						idxList: append(qb.currentBatch.idxList, idx),
+						req:   mergedReq[0],
+						ctx:   qb.currentBatch.ctx,
+						dones: append(qb.currentBatch.dones, done),
 					}
 				}
 
@@ -109,7 +106,7 @@ func (qb *DefaultBatcher) startReadingFlushingGoroutine() {
 					qb.currentBatchMu.Unlock()
 
 					// flush() blocks until successfully started a goroutine for flushing.
-					qb.flush(batchToFlush)
+					qb.flush(batchToFlush.ctx, batchToFlush.req, batchToFlush.dones)
 					qb.resetTimer()
 				} else {
 					qb.currentBatchMu.Unlock()
@@ -168,7 +165,7 @@ func (qb *DefaultBatcher) flushCurrentBatchIfNecessary() {
 	qb.currentBatchMu.Unlock()
 
 	// flush() blocks until successfully started a goroutine for flushing.
-	qb.flush(batchToFlush)
+	qb.flush(batchToFlush.ctx, batchToFlush.req, batchToFlush.dones)
 	qb.resetTimer()
 }
 

exporter/internal/queue/disabled_batcher.go

@@ -7,6 +7,7 @@ import (
 	"context"
 
 	"go.opentelemetry.io/collector/component"
+	"go.opentelemetry.io/collector/exporter/exporterqueue"
 )
 
 // DisabledBatcher is a special-case of Batcher that has no size limit for sending. Any items read from the queue will
@@ -29,15 +30,11 @@ func (qb *DisabledBatcher) Start(_ context.Context, _ component.Host) error {
 	go func() {
 		defer qb.stopWG.Done()
 		for {
-			idx, _, req, ok := qb.queue.Read(context.Background())
+			_, req, done, ok := qb.queue.Read(context.Background())
 			if !ok {
 				return
 			}
-			qb.flush(batch{
-				req:     req,
-				ctx:     context.Background(),
-				idxList: []uint64{idx},
-			})
+			qb.flush(context.Background(), req, []exporterqueue.DoneCallback{done})
 		}
 	}()
 	return nil