Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add support for exporting internal metrics via OpenTelemetry library. #3816

Merged
merged 1 commit into from
Sep 23, 2021
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions config/configtelemetry/configtelemetry.go
Original file line number Diff line number Diff line change
Expand Up @@ -38,6 +38,8 @@ const (
metricsLevelCfg = "metrics-level"
)

const UseOpenTelemetryForInternalMetrics = false

var metricsLevelPtr = new(Level)

// Flags is a helper function to add telemetry config flags to the service that exposes
Expand Down
3 changes: 3 additions & 0 deletions go.mod
Original file line number Diff line number Diff line change
Expand Up @@ -23,8 +23,11 @@ require (
go.opentelemetry.io/contrib/instrumentation/net/http/otelhttp v0.23.0
go.opentelemetry.io/contrib/zpages v0.23.0
go.opentelemetry.io/otel v1.0.0
go.opentelemetry.io/otel/exporters/prometheus v0.23.0
go.opentelemetry.io/otel/metric v0.23.0
go.opentelemetry.io/otel/sdk v1.0.0
go.opentelemetry.io/otel/sdk/export/metric v0.23.0
go.opentelemetry.io/otel/sdk/metric v0.23.0
go.opentelemetry.io/otel/trace v1.0.0
go.uber.org/atomic v1.9.0
go.uber.org/zap v1.19.1
Expand Down
6 changes: 6 additions & 0 deletions go.sum
Original file line number Diff line number Diff line change
Expand Up @@ -408,13 +408,19 @@ go.opentelemetry.io/contrib/zpages v0.23.0/go.mod h1:i5BVZTRftVMBmYLP/T++in2G5MA
go.opentelemetry.io/otel v1.0.0-RC3/go.mod h1:Ka5j3ua8tZs4Rkq4Ex3hwgBgOchyPVq5S6P2lz//nKQ=
go.opentelemetry.io/otel v1.0.0 h1:qTTn6x71GVBvoafHK/yaRUmFzI4LcONZD0/kXxl5PHI=
go.opentelemetry.io/otel v1.0.0/go.mod h1:AjRVh9A5/5DE7S+mZtTR6t8vpKKryam+0lREnfmS4cg=
go.opentelemetry.io/otel/exporters/prometheus v0.23.0 h1:ZFx1kUjUSBF7H1mTPHHOqglEDQsxYBrDnYZ8i41v3iE=
go.opentelemetry.io/otel/exporters/prometheus v0.23.0/go.mod h1:kjCXbxQnnEm5l3HrUw4IPyuALu7Uqb/bEK7vWQnbd8s=
go.opentelemetry.io/otel/internal/metric v0.23.0 h1:mPfzm9Iqhw7G2nDBmUAjFTfPqLZPbOW2k7QI57ITbaI=
go.opentelemetry.io/otel/internal/metric v0.23.0/go.mod h1:z+RPiDJe30YnCrOhFGivwBS+DU1JU/PiLKkk4re2DNY=
go.opentelemetry.io/otel/metric v0.23.0 h1:mYCcDxi60P4T27/0jchIDFa1WHEfQeU3zH9UEMpnj2c=
go.opentelemetry.io/otel/metric v0.23.0/go.mod h1:G/Nn9InyNnIv7J6YVkQfpc0JCfKBNJaERBGw08nqmVQ=
go.opentelemetry.io/otel/sdk v1.0.0-RC3/go.mod h1:78H6hyg2fka0NYT9fqGuFLvly2yCxiBXDJAgLKo/2Us=
go.opentelemetry.io/otel/sdk v1.0.0 h1:BNPMYUONPNbLneMttKSjQhOTlFLOD9U22HNG1KrIN2Y=
go.opentelemetry.io/otel/sdk v1.0.0/go.mod h1:PCrDHlSy5x1kjezSdL37PhbFUMjrsLRshJ2zCzeXwbM=
go.opentelemetry.io/otel/sdk/export/metric v0.23.0 h1:7NeoKPPx6NdZBVHLEp/LY5Lq85Ff1WNZnuJkuRy+azw=
go.opentelemetry.io/otel/sdk/export/metric v0.23.0/go.mod h1:SuMiREmKVRIwFKq73zvGTvwFpxb/ZAYkMfyqMoOtDqs=
go.opentelemetry.io/otel/sdk/metric v0.23.0 h1:xlZhPbiue1+jjSFEth94q9QCmX8Q24mOtue9IAmlVyI=
go.opentelemetry.io/otel/sdk/metric v0.23.0/go.mod h1:wa0sKK13eeIFW+0OFjcC3S1i7FTRRiLAXe1kjBVbhwg=
go.opentelemetry.io/otel/trace v1.0.0-RC3/go.mod h1:VUt2TUYd8S2/ZRX09ZDFZQwn2RqfMB5MzO17jBojGxo=
go.opentelemetry.io/otel/trace v1.0.0 h1:TSBr8GTEtKevYMG/2d21M989r5WJYVimhTHBKVEZuh4=
go.opentelemetry.io/otel/trace v1.0.0/go.mod h1:PXTWqayeFUlJV1YDNhsJYB184+IvAH814St6o6ajzIs=
Expand Down
99 changes: 74 additions & 25 deletions service/telemetry.go
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,13 @@ import (
"contrib.go.opencensus.io/exporter/prometheus"
"github.com/google/uuid"
"go.opencensus.io/stats/view"
otelprometheus "go.opentelemetry.io/otel/exporters/prometheus"
"go.opentelemetry.io/otel/metric/global"
export "go.opentelemetry.io/otel/sdk/export/metric"
"go.opentelemetry.io/otel/sdk/metric/aggregator/histogram"
controller "go.opentelemetry.io/otel/sdk/metric/controller/basic"
processor "go.opentelemetry.io/otel/sdk/metric/processor/basic"
selector "go.opentelemetry.io/otel/sdk/metric/selector/simple"
"go.uber.org/zap"

"go.opentelemetry.io/collector/config/configtelemetry"
Expand Down Expand Up @@ -71,9 +78,56 @@ func (tel *colTelemetry) initOnce(asyncErrorChannel chan<- error, ballastSizeByt
return nil
}

var instanceID string
if telemetry.GetAddInstanceID() {
instanceUUID, _ := uuid.NewRandom()
instanceID = instanceUUID.String()
}

var pe http.Handler
if configtelemetry.UseOpenTelemetryForInternalMetrics {
otelHandler, err := tel.initOpenTelemetry()
tigrannajaryan marked this conversation as resolved.
Show resolved Hide resolved
if err != nil {
return err
}
pe = otelHandler
} else {
ocHandler, err := tel.initOpenCensus(level, instanceID, ballastSizeBytes)
if err != nil {
return err
}
pe = ocHandler
}

logger.Info(
"Serving Prometheus metrics",
zap.String("address", metricsAddr),
zap.Int8("level", int8(level)), // TODO: make it human friendly
zap.String(semconv.AttributeServiceInstanceID, instanceID),
)

mux := http.NewServeMux()
mux.Handle("/metrics", pe)

tel.server = &http.Server{
Addr: metricsAddr,
Handler: mux,
}

go func() {
serveErr := tel.server.ListenAndServe()
if serveErr != nil && serveErr != http.ErrServerClosed {
asyncErrorChannel <- serveErr
}
}()

return nil
}

func (tel *colTelemetry) initOpenCensus(level configtelemetry.Level, instanceID string, ballastSizeBytes uint64) (http.Handler, error) {
processMetricsViews, err := telemetry2.NewProcessMetricsViews(ballastSizeBytes)
if err != nil {
return err
return nil, err
}

var views []*view.View
Expand All @@ -84,7 +138,7 @@ func (tel *colTelemetry) initOnce(asyncErrorChannel chan<- error, ballastSizeByt

tel.views = views
if err = view.Register(views...); err != nil {
return err
return nil, err
}

processMetricsViews.StartCollection()
Expand All @@ -94,45 +148,40 @@ func (tel *colTelemetry) initOnce(asyncErrorChannel chan<- error, ballastSizeByt
Namespace: telemetry.GetMetricsPrefix(),
}

var instanceID string
if telemetry.GetAddInstanceID() {
instanceUUID, _ := uuid.NewRandom()
instanceID = instanceUUID.String()
opts.ConstLabels = map[string]string{
sanitizePrometheusKey(semconv.AttributeServiceInstanceID): instanceID,
}
}

pe, err := prometheus.NewExporter(opts)
if err != nil {
return err
return nil, err
}

view.RegisterExporter(pe)
return pe, nil
}

logger.Info(
"Serving Prometheus metrics",
zap.String("address", metricsAddr),
zap.Int8("level", int8(level)), // TODO: make it human friendly
zap.String(semconv.AttributeServiceInstanceID, instanceID),
func (tel *colTelemetry) initOpenTelemetry() (http.Handler, error) {
config := otelprometheus.Config{}
c := controller.New(
processor.New(
selector.NewWithHistogramDistribution(
histogram.WithExplicitBoundaries(config.DefaultHistogramBoundaries),
),
export.CumulativeExportKindSelector(),
processor.WithMemory(true),
),
)

mux := http.NewServeMux()
mux.Handle("/metrics", pe)

tel.server = &http.Server{
Addr: metricsAddr,
Handler: mux,
pe, err := otelprometheus.New(config, c)
if err != nil {
return nil, err
}

go func() {
serveErr := tel.server.ListenAndServe()
if serveErr != nil && serveErr != http.ErrServerClosed {
asyncErrorChannel <- serveErr
}
}()

return nil
global.SetMeterProvider(pe.MeterProvider())
return pe, err
}

func (tel *colTelemetry) shutdown() error {
Expand Down