Add batch_size config

sinks/httpSink.go

@@ -38,7 +38,13 @@ type HttpSinkConfig struct {
 	IdleConnTimeout string `json:"idle_connection_timeout,omitempty"`
 	idleConnTimeout time.Duration
 
+	// Maximum number of points sent to server in single request.
+	// Default: 1000
+	BatchSize int `json:"batch_size,omitempty"`
+
 	// Batch all writes arriving in during this duration
+	// If batch size is reached before the end of this interval,
+	// the metrics are sent without further delay.
 	// (default '5s', batching can be disabled by setting it to 0)
 	FlushDelay string `json:"flush_delay,omitempty"`
 	flushDelay time.Duration
@@ -62,6 +68,7 @@ type HttpSink struct {
 	// Flush() runs in another goroutine and accesses the influx line protocol encoder,
 	// so this encoderLock has to protect the encoder
 	encoderLock        sync.Mutex
+	encoderRecordCount int
 
 	// timer to run Flush()
 	flushTimer *time.Timer
@@ -133,6 +140,8 @@ func (s *HttpSink) Write(m lp.CCMetric) error {
 	// Encode time stamp
 	s.encoder.EndLine(m.Time())
 
+	s.encoderRecordCount++
+
 	// Check for encoder errors
 	err := s.encoder.Err()
 
@@ -149,6 +158,27 @@ func (s *HttpSink) Write(m lp.CCMetric) error {
 
 		// Directly flush if no flush delay is configured
 		return s.Flush()
+	} else if s.encoderRecordCount > s.config.BatchSize {
+
+		// Flush if batch size limit is reached
+
+		// Stop flush timer if possible
+		if s.flushTimer != nil {
+			if ok := s.flushTimer.Stop(); ok {
+				cclog.ComponentDebug(s.name, "Stopped flush timer. Batch size limit reached before flush delay")
+				s.timerLock.Unlock()
+			}
+		}
+
+		// Start asynchronous flush operation
+		go func() {
+			cclog.ComponentDebug(s.name, "Starting flush triggered by batch size limit")
+			if err := s.Flush(); err != nil {
+				cclog.ComponentError(s.name, "Flush triggered by batch size limit: flush failed:", err)
+			}
+		}()
+		return nil
+
 	} else if s.timerLock.TryLock() {
 
 		// Setup flush timer when flush delay is configured
@@ -166,9 +196,9 @@ func (s *HttpSink) Write(m lp.CCMetric) error {
 				s.config.flushDelay,
 				func() {
 					defer s.timerLock.Unlock()
-					cclog.ComponentDebug(s.name, "Starting flush in flush timer")
+					cclog.ComponentDebug(s.name, "Starting flush triggered by flush timer")
 					if err := s.Flush(); err != nil {
-						cclog.ComponentError(s.name, "Flush timer: flush failed:", err)
+						cclog.ComponentError(s.name, "Flush triggered by flush timer: flush failed:", err)
 					}
 				})
 		}
@@ -184,6 +214,8 @@ func (s *HttpSink) Flush() error {
 
 	buf := slices.Clone(s.encoder.Bytes())
 	s.encoder.Reset()
+	recordCount := s.encoderRecordCount
+	s.encoderRecordCount = 0
 
 	// Unlock encoder usage
 	s.encoderLock.Unlock()
@@ -192,6 +224,8 @@ func (s *HttpSink) Flush() error {
 		return nil
 	}
 
+	cclog.ComponentDebug(s.name, "Flushing", recordCount, "records")
+
 	var res *http.Response
 	for i := 0; i < s.config.MaxRetries; i++ {
 		// Create new request to send buffer
@@ -253,6 +287,7 @@ func NewHttpSink(name string, config json.RawMessage) (Sink, error) {
 	// should be larger than the measurement interval to keep the connection open
 	s.config.IdleConnTimeout = "120s"
 	s.config.Timeout = "5s"
+	s.config.BatchSize = 1000
 	s.config.FlushDelay = "5s"
 	s.config.MaxRetries = 3
 	cclog.ComponentDebug(s.name, "init")

sinks/httpSink.md

@@ -18,6 +18,7 @@ The `http` sink uses POST requests to a HTTP server to submit the metrics in the
     "timeout": "5s",
     "idle_connection_timeout" : "5s",
     "flush_delay": "2s",
+    "batch_size": 1000
   }
 }
 ```
@@ -32,3 +33,4 @@ The `http` sink uses POST requests to a HTTP server to submit the metrics in the
 - `max_retries`: Maximum number of retries to connect to the http server
 - `idle_connection_timeout`: Timeout for idle connections (default '120s'). Should be larger than the measurement interval to keep the connection open
 - `flush_delay`: Batch all writes arriving in during this duration (default '1s', batching can be disabled by setting it to 0)
+- `batch_size`: Maximal batch size. If `batch_size` is reached before the end of `flush_delay`, the metrics are sent without further delay