Implement telegraf's own full metric type

main reasons behind this:
- make adding/removing tags cheap
- make adding/removing fields cheap
- make parsing cheaper
- make parse -> decorate -> write out bytes metric flow much faster

Refactor serializer to use byte buffer

internal/models/makemetric.go

@@ -6,6 +6,7 @@ import (
 	"time"
 
 	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/metric"
 )
 
 // makemetric is used by both RunningAggregator & RunningInput
@@ -135,16 +136,7 @@ func makemetric(
 		}
 	}
 
-	var m telegraf.Metric
-	var err error
-	switch mType {
-	case telegraf.Counter:
-		m, err = telegraf.NewCounterMetric(measurement, tags, fields, t)
-	case telegraf.Gauge:
-		m, err = telegraf.NewGaugeMetric(measurement, tags, fields, t)
-	default:
-		m, err = telegraf.NewMetric(measurement, tags, fields, t)
-	}
+	m, err := metric.New(measurement, tags, fields, t, mType)
 	if err != nil {
 		log.Printf("Error adding point [%s]: %s\n", measurement, err.Error())
 		return nil

internal/models/running_aggregator.go

@@ -4,6 +4,7 @@ import (
 	"time"
 
 	"github.com/influxdata/telegraf"
+	"github.com/influxdata/telegraf/metric"
 )
 
 type RunningAggregator struct {
@@ -90,7 +91,7 @@ func (r *RunningAggregator) Add(in telegraf.Metric) bool {
 			return false
 		}
 
-		in, _ = telegraf.NewMetric(name, tags, fields, t)
+		in, _ = metric.New(name, tags, fields, t)
 	}
 
 	r.metrics <- in

internal/models/running_output.go

@@ -6,6 +6,7 @@ import (
 
 	"github.com/influxdata/telegraf"
 	"github.com/influxdata/telegraf/internal/buffer"
+	"github.com/influxdata/telegraf/metric"
 )
 
 const (
@@ -56,23 +57,23 @@ func NewRunningOutput(
 
 // AddMetric adds a metric to the output. This function can also write cached
 // points if FlushBufferWhenFull is true.
-func (ro *RunningOutput) AddMetric(metric telegraf.Metric) {
+func (ro *RunningOutput) AddMetric(m telegraf.Metric) {
 	// Filter any tagexclude/taginclude parameters before adding metric
 	if ro.Config.Filter.IsActive() {
 		// In order to filter out tags, we need to create a new metric, since
 		// metrics are immutable once created.
-		name := metric.Name()
-		tags := metric.Tags()
-		fields := metric.Fields()
-		t := metric.Time()
+		name := m.Name()
+		tags := m.Tags()
+		fields := m.Fields()
+		t := m.Time()
 		if ok := ro.Config.Filter.Apply(name, fields, tags); !ok {
 			return
 		}
 		// error is not possible if creating from another metric, so ignore.
-		metric, _ = telegraf.NewMetric(name, tags, fields, t)
+		m, _ = metric.New(name, tags, fields, t)
 	}
 
-	ro.metrics.Add(metric)
+	ro.metrics.Add(m)
 	if ro.metrics.Len() == ro.MetricBatchSize {
 		batch := ro.metrics.Batch(ro.MetricBatchSize)
 		err := ro.write(batch)