Merge pull request '[CI] workerqueue attempt to fix flacky test' (#2721) from oliverpool/forgejo:queue_flaky_fix into forgejo

Reviewed-on: https://codeberg.org/forgejo/forgejo/pulls/2721
Reviewed-by: Earl Warren <earl-warren@noreply.codeberg.org>

modules/queue/workergroup.go

@@ -146,8 +146,6 @@ func (q *WorkerPoolQueue[T]) doStartNewWorker(wp *workerGroup[T]) {
 		log.Debug("Queue %q starts new worker", q.GetName())
 		defer log.Debug("Queue %q stops idle worker", q.GetName())
 
-		atomic.AddInt32(&q.workerStartedCounter, 1) // Only increase counter, used for debugging
-
 		t := time.NewTicker(workerIdleDuration)
 		defer t.Stop()
 

modules/queue/workerqueue.go

@@ -40,8 +40,6 @@ type WorkerPoolQueue[T any] struct {
 	workerMaxNum    int
 	workerActiveNum int
 	workerNumMu     sync.Mutex
-
-	workerStartedCounter int32
 }
 
 type flushType chan struct{}

modules/queue/workerqueue_test.go

@@ -4,7 +4,9 @@
 package queue
 
 import (
+	"bytes"
 	"context"
+	"runtime"
 	"strconv"
 	"sync"
 	"testing"
@@ -14,6 +16,7 @@ import (
 	"code.gitea.io/gitea/modules/test"
 
 	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
 )
 
 func runWorkerPoolQueue[T any](q *WorkerPoolQueue[T]) func() {
@@ -249,23 +252,40 @@ func TestWorkerPoolQueueShutdown(t *testing.T) {
 }
 
 func TestWorkerPoolQueueWorkerIdleReset(t *testing.T) {
-	defer test.MockVariableValue(&workerIdleDuration, 10*time.Millisecond)()
+	defer test.MockVariableValue(&workerIdleDuration, 1*time.Millisecond)()
 
+	chGoroutineIDs := make(chan string)
 	handler := func(items ...int) (unhandled []int) {
-		time.Sleep(50 * time.Millisecond)
+		time.Sleep(10 * workerIdleDuration)
+		chGoroutineIDs <- goroutineID() // hacky way to identify a worker
 		return nil
 	}
 
 	q, _ := newWorkerPoolQueueForTest("test-workpoolqueue", setting.QueueSettings{Type: "channel", BatchLength: 1, MaxWorkers: 2, Length: 100}, handler, false)
 	stop := runWorkerPoolQueue(q)
-	for i := 0; i < 20; i++ {
+
+	const workloadSize = 12
+	for i := 0; i < workloadSize; i++ {
 		assert.NoError(t, q.Push(i))
 	}
 
-	time.Sleep(500 * time.Millisecond)
-	assert.EqualValues(t, 2, q.GetWorkerNumber())
-	assert.EqualValues(t, 2, q.GetWorkerActiveNumber())
-	// when the queue never becomes empty, the existing workers should keep working
-	assert.LessOrEqual(t, q.workerStartedCounter, int32(4)) // counter should be 2, but sometimes it gets bigger
+	workerIDs := make(map[string]struct{})
+	for i := 0; i < workloadSize; i++ {
+		c := <-chGoroutineIDs
+		workerIDs[c] = struct{}{}
+		t.Logf("%d workers: overall=%d current=%d", i, len(workerIDs), q.GetWorkerNumber())
+
+		// ensure that no more than qs.MaxWorkers workers are created over the whole lifetime of the queue
+		// (otherwise it would mean that some workers got shut down while the queue was full)
+		require.LessOrEqual(t, len(workerIDs), q.GetWorkerMaxNumber())
+	}
+	close(chGoroutineIDs)
+
 	stop()
 }
+
+func goroutineID() string {
+	var buffer [31]byte
+	_ = runtime.Stack(buffer[:], false)
+	return string(bytes.Fields(buffer[10:])[0])
+}