[8.17](backport #42966) fix(libbeat): mitigate race condition in ratelimit processor

CHANGELOG.next.asciidoc

@@ -31,6 +31,23 @@ https://github.com/elastic/beats/compare/v8.8.1\...main[Check the HEAD diff]
 - Redis: Added replication role as a field to submitted slowlogs
 - Change log.file.path field in awscloudwatch input to nested object. {pull}41099[41099]
 
+- Remove deprecated awscloudwatch field from Filebeat. {pull}41089[41089]
+- The performance of ingesting SQS data with the S3 input has improved by up to 60x for queues with many small events. `max_number_of_messages` config for SQS mode is now ignored, as the new design no longer needs a manual cap on messages. Instead, use `number_of_workers` to scale ingestion rate in both S3 and SQS modes. The increased efficiency may increase network bandwidth consumption, which can be throttled by lowering `number_of_workers`. It may also increase number of events stored in memory, which can be throttled by lowering the configured size of the internal queue. {pull}40699[40699]
+- Fixes filestream logging the error "filestream input with ID 'ID' already exists, this will lead to data duplication[...]" on Kubernetes when using autodiscover. {pull}41585[41585]
+- Add kafka compression support for ZSTD.
+- Filebeat fails to start if there is any input with a duplicated ID. It logs the duplicated IDs and the offending inputs configurations. {pull}41731[41731]
+- Filestream inputs with duplicated IDs will fail to start. An error is logged showing the ID and the full input configuration. {issue}41938[41938] {pull}41954[41954]
+- Filestream inputs can define `allow_deprecated_id_duplication: true` to run keep the previous behaviour of running inputs with duplicated IDs. {issue}41938[41938] {pull}41954[41954]
+- The Filestream input only starts to ingest a file when it is >= 1024 bytes in size. This happens because the fingerprint` is the default file identity now. To restore the previous behaviour, set `file_identity.native: ~` and `prospector.scanner.fingerprint.enabled: false` {issue}40197[40197] {pull}41762[41762]
+- Filebeat fails to start when its configuration contains usage of the deprecated `log` or `container` inputs. However, they can still be using while setting `allow_deprecated_use: true` in their configuration {pull}42295[42295]
+- The fields produced by the Journald input are updated to better match ECS. Renamed fields:
+Dropped fields: `syslog.priority` and `syslog.facility` while keeping their duplicated equivalent:
+`log.syslog.priority`,`log.syslog.facility.code`. Renamed fields: `syslog.identifier` -> `log.syslog.appname`,
+`syslog.pid` -> `log.syslog.procid`. `container.id_truncated` is dropped because the full container ID is
+already present as `container.id` and `container.log.tag` is dropped because it is already present as
+`log.syslog.appname`. The field `container.partial` is replaced by the tag `partial_message` if it was `true`,
+otherwise no tag is added. {issue}42208[42208] {pull}42403[42403]
+- Fixed race conditions in the global ratelimit processor that could drop events or apply rate limiting incorrectly.
 
 *Heartbeat*
 

libbeat/processors/ratelimit/rate_limit_test.go

@@ -176,3 +176,31 @@ func TestRateLimit(t *testing.T) {
 		})
 	}
 }
+
+func TestAllocs(t *testing.T) {
+	p, err := new(conf.MustNewConfigFrom(mapstr.M{
+		"limit": "100/s",
+	}))
+	require.NoError(t, err)
+	event := beat.Event{Fields: mapstr.M{"field": 1}}
+
+	allocs := testing.AllocsPerRun(1000, func() {
+		p.Run(&event) //nolint:errcheck // ignore
+	})
+	if allocs > 0 {
+		t.Errorf("allocs = %v; want 0", allocs)
+	}
+}
+
+func BenchmarkRateLimit(b *testing.B) {
+	p, err := new(conf.MustNewConfigFrom(mapstr.M{
+		"limit": "100/s",
+	}))
+	require.NoError(b, err)
+	event := beat.Event{Fields: mapstr.M{"field": 1}}
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		p.Run(&event) //nolint:errcheck // ignore
+	}
+}

libbeat/processors/ratelimit/token_bucket.go

@@ -35,10 +35,36 @@
 }
 
 type bucket struct {
+	mu sync.Mutex
+
 	tokens        float64
 	lastReplenish time.Time
 }
 
+func (b *bucket) withdraw() bool {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	if b.tokens < 1 {
+		return false
+	}
+
+	b.tokens--
+	return true
+}
+
+func (b *bucket) replenish(rate rate, clock clockwork.Clock) float64 {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	secsSinceLastReplenish := clock.Now().Sub(b.lastReplenish).Seconds()
+	tokensToReplenish := secsSinceLastReplenish * rate.valuePerSecond()
+
+	b.tokens += tokensToReplenish
+	b.lastReplenish = clock.Now()
+	return b.tokens
+}
+
 type tokenBucket struct {
 	mu unison.Mutex
 
@@ -122,34 +148,29 @@
 }
 
 func (t *tokenBucket) getBucket(key uint64) *bucket {
+<<<<<<< HEAD
 	v, exists := t.buckets.LoadOrStore(key, &bucket{
 		tokens:        t.depth,
 		lastReplenish: t.clock.Now(),
 	})
 	b := v.(*bucket)
 
+=======
+	v, exists := t.buckets.Load(key)
+>>>>>>> acce29cf2 (fix(libbeat): mitigate race condition in ratelimit processor (#42966))
 	if exists {
+		//nolint:errcheck // ignore
+		b := v.(*bucket)
 		b.replenish(t.limit, t.clock)
 		return b
 	}
 
-	return b
-}
-
-func (b *bucket) withdraw() bool {
-	if b.tokens < 1 {
-		return false
-	}
-	b.tokens--
-	return true
-}
-
-func (b *bucket) replenish(rate rate, clock clockwork.Clock) {
-	secsSinceLastReplenish := clock.Now().Sub(b.lastReplenish).Seconds()
-	tokensToReplenish := secsSinceLastReplenish * rate.valuePerSecond()
-
-	b.tokens += tokensToReplenish
-	b.lastReplenish = clock.Now()
+	v, _ = t.buckets.LoadOrStore(key, &bucket{
+		tokens:        t.depth,
+		lastReplenish: t.clock.Now(),
+	})
+	//nolint:errcheck // ignore
+	return v.(*bucket)
 }
 
 func (t *tokenBucket) runGC() {
@@ -174,9 +195,8 @@
 			key := k.(uint64)
 			b := v.(*bucket)
 
-			b.replenish(t.limit, t.clock)
-
-			if b.tokens >= t.depth {
+			tokens := b.replenish(t.limit, t.clock)
+			if tokens >= t.depth {
 				toDelete = append(toDelete, key)
 			}
 
@@ -190,7 +210,11 @@
 		}
 
 		// Reset GC metrics
+<<<<<<< HEAD
 		t.gc.metrics.numCalls = atomic.MakeUint(0)
+=======
+		t.gc.metrics.numCalls.Store(0)
+>>>>>>> acce29cf2 (fix(libbeat): mitigate race condition in ratelimit processor (#42966))
 
 		gcDuration := time.Now().Sub(gcStartTime)
 		numBucketsDeleted := len(toDelete)
@@ -198,4 +222,4 @@
 		t.logger.Debugf("gc duration: %v, buckets: (before: %v, deleted: %v, after: %v)",
 			gcDuration, numBucketsBefore, numBucketsDeleted, numBucketsAfter)
 	}()
 }