Support remote write v2

Signed-off-by: SungJin1212 <[email protected]>

Author: SungJin1212

Diff for: .github/workflows/test-build-deploy.yml

@@ -144,6 +144,7 @@ jobs:
           - integration_querier
           - integration_ruler
           - integration_query_fuzz
+          - integration_remote_write_v2
     steps:
       - name: Upgrade golang
         uses: actions/setup-go@bfdd3570ce990073878bf10f6b2d79082de49492 # v2.2.0

Diff for: .golangci.yml

@@ -49,3 +49,4 @@ run:
     - integration_querier
     - integration_ruler
     - integration_query_fuzz
+    - integration_remote_write_v2

Diff for: CHANGELOG.md

@@ -9,6 +9,7 @@
 * [FEATURE] Ruler: Minimize chances of missed rule group evaluations that can occur due to OOM kills, bad underlying nodes, or due to an unhealthy ruler that appears in the ring as healthy. This feature is enabled via `-ruler.enable-ha-evaluation` flag. #6129
 * [FEATURE] Store Gateway: Add an in-memory chunk cache. #6245
 * [FEATURE] Chunk Cache: Support multi level cache and add metrics. #6249
+* [FEATURE] Distributor/Ingester: Support remote write 2.0. It includes proto, samples, and (native) histograms ingestion. #6292
 * [ENHANCEMENT] S3 Bucket Client: Add a list objects version configs to configure list api object version. #6280
 * [ENHANCEMENT] OpenStack Swift: Add application credential configs for Openstack swift object storage backend. #6255
 * [ENHANCEMENT] Query Frontend: Add new query stats metrics `cortex_query_samples_scanned_total` and `cortex_query_peak_samples` to track scannedSamples and peakSample per user. #6228

Diff for: Makefile

@@ -85,6 +85,7 @@ $(foreach exe, $(EXES), $(eval $(call dep_exe, $(exe))))
 
 # Manually declared dependencies And what goes into each exe
 pkg/cortexpb/cortex.pb.go: pkg/cortexpb/cortex.proto
+pkg/cortexpbv2/cortexv2.pb.go: pkg/cortexpbv2/cortexv2.proto
 pkg/ingester/client/ingester.pb.go: pkg/ingester/client/ingester.proto
 pkg/distributor/distributorpb/distributor.pb.go: pkg/distributor/distributorpb/distributor.proto
 pkg/ingester/wal.pb.go: pkg/ingester/wal.proto

Diff for: integration/e2e/util.go

@@ -24,6 +24,7 @@ import (
 	"github.com/thanos-io/thanos/pkg/block/metadata"
 	"github.com/thanos-io/thanos/pkg/runutil"
 
+	"github.com/cortexproject/cortex/pkg/cortexpbv2"
 	cortex_tsdb "github.com/cortexproject/cortex/pkg/storage/tsdb"
 )
 
@@ -149,6 +150,46 @@ func GenerateSeries(name string, ts time.Time, additionalLabels ...prompb.Label)
 	return
 }
 
+func GenerateHistogramSeriesV2(name string, ts time.Time, i uint32, floatHistogram bool, additionalLabels ...prompb.Label) (symbols []string, series []cortexpbv2.TimeSeries) {
+	tsMillis := TimeToMilliseconds(ts)
+
+	labelRefs := []uint32{1, 2}
+	symbols = []string{"", "__name__", name}
+
+	for i, label := range additionalLabels {
+		symbols = append(symbols, label.Name, label.Value)
+		labelRefs = append(labelRefs, uint32(i*2+3), uint32(i*2+4))
+	}
+
+	// Generate the expected vector when querying it
+	metric := model.Metric{}
+	metric[labels.MetricName] = model.LabelValue(name)
+	for _, lbl := range additionalLabels {
+		metric[model.LabelName(lbl.Name)] = model.LabelValue(lbl.Value)
+	}
+
+	var (
+		h  *histogram.Histogram
+		fh *histogram.FloatHistogram
+		ph cortexpbv2.Histogram
+	)
+	if floatHistogram {
+		fh = tsdbutil.GenerateTestFloatHistogram(int(i))
+		ph = cortexpbv2.FloatHistogramToHistogramProto(tsMillis, fh)
+	} else {
+		h = tsdbutil.GenerateTestHistogram(int(i))
+		ph = cortexpbv2.HistogramToHistogramProto(tsMillis, h)
+	}
+
+	// Generate the series
+	series = append(series, cortexpbv2.TimeSeries{
+		LabelsRefs: labelRefs,
+		Histograms: []cortexpbv2.Histogram{ph},
+	})
+
+	return
+}
+
 func GenerateHistogramSeries(name string, ts time.Time, i uint32, floatHistogram bool, additionalLabels ...prompb.Label) (series []prompb.TimeSeries) {
 	tsMillis := TimeToMilliseconds(ts)
 
@@ -188,6 +229,76 @@ func GenerateHistogramSeries(name string, ts time.Time, i uint32, floatHistogram
 	return
 }
 
+func GenerateSeriesV2(name string, ts time.Time, additionalLabels ...prompb.Label) (symbols []string, series []cortexpbv2.TimeSeries, vector model.Vector) {
+	tsMillis := TimeToMilliseconds(ts)
+	value := rand.Float64()
+
+	labelRefs := []uint32{1, 2}
+	symbols = []string{"", "__name__", name}
+
+	for i, label := range additionalLabels {
+		symbols = append(symbols, label.Name, label.Value)
+		labelRefs = append(labelRefs, uint32(i*2+3), uint32(i*2+4))
+	}
+
+	// Generate the series
+	series = append(series, cortexpbv2.TimeSeries{
+		LabelsRefs: labelRefs,
+		Samples: []cortexpbv2.Sample{
+			{Value: value, Timestamp: tsMillis},
+		},
+	})
+
+	// Generate the expected vector when querying it
+	metric := model.Metric{}
+	metric[labels.MetricName] = model.LabelValue(name)
+	for _, lbl := range additionalLabels {
+		metric[model.LabelName(lbl.Name)] = model.LabelValue(lbl.Value)
+	}
+
+	vector = append(vector, &model.Sample{
+		Metric:    metric,
+		Value:     model.SampleValue(value),
+		Timestamp: model.Time(tsMillis),
+	})
+
+	return
+}
+
+func GenerateSeriesV2WithSamples(
+	name string,
+	startTime time.Time,
+	scrapeInterval time.Duration,
+	startValue int, numSamples int,
+	additionalLabels ...prompb.Label,
+) (symbols []string, series cortexpbv2.TimeSeries) {
+	tsMillis := TimeToMilliseconds(startTime)
+	durMillis := scrapeInterval.Milliseconds()
+
+	symbols = []string{"", "__name__", name}
+	labelRefs := []uint32{1, 2}
+
+	for i, label := range additionalLabels {
+		symbols = append(symbols, label.Name, label.Value)
+		labelRefs = append(labelRefs, uint32(i*2+3), uint32(i*2+4))
+	}
+
+	startTMillis := tsMillis
+	samples := make([]cortexpbv2.Sample, numSamples)
+	for i := 0; i < numSamples; i++ {
+		samples[i] = cortexpbv2.Sample{
+			Timestamp: startTMillis,
+			Value:     float64(i + startValue),
+		}
+		startTMillis += durMillis
+	}
+
+	return symbols, cortexpbv2.TimeSeries{
+		LabelsRefs: labelRefs,
+		Samples:    samples,
+	}
+}
+
 func GenerateSeriesWithSamples(
 	name string,
 	startTime time.Time,

Diff for: integration/e2ecortex/client.go

@@ -25,12 +25,12 @@ import (
 	"github.com/prometheus/prometheus/prompb"
 	"github.com/prometheus/prometheus/storage"
 	"github.com/prometheus/prometheus/storage/remote"
-	yaml "gopkg.in/yaml.v3"
-
 	"go.opentelemetry.io/collector/pdata/pcommon"
 	"go.opentelemetry.io/collector/pdata/pmetric"
 	"go.opentelemetry.io/collector/pdata/pmetric/pmetricotlp"
+	yaml "gopkg.in/yaml.v3"
 
+	"github.com/cortexproject/cortex/pkg/cortexpbv2"
 	"github.com/cortexproject/cortex/pkg/ruler"
 	"github.com/cortexproject/cortex/pkg/util/backoff"
 )
@@ -113,6 +113,39 @@ func NewPromQueryClient(address string) (*Client, error) {
 	return c, nil
 }
 
+// PushV2 the input timeseries to the remote endpoint
+func (c *Client) PushV2(symbols []string, timeseries []cortexpbv2.TimeSeries) (*http.Response, error) {
+	// Create write request
+	data, err := proto.Marshal(&cortexpbv2.WriteRequest{Symbols: symbols, Timeseries: timeseries})
+	if err != nil {
+		return nil, err
+	}
+
+	// Create HTTP request
+	compressed := snappy.Encode(nil, data)
+	req, err := http.NewRequest("POST", fmt.Sprintf("http://%s/api/prom/push", c.distributorAddress), bytes.NewReader(compressed))
+	if err != nil {
+		return nil, err
+	}
+
+	req.Header.Add("Content-Encoding", "snappy")
+	req.Header.Set("Content-Type", "application/x-protobuf;proto=io.prometheus.write.v2.Request")
+	req.Header.Set("X-Prometheus-Remote-Write-Version", "2.0.0")
+	req.Header.Set("X-Scope-OrgID", c.orgID)
+
+	ctx, cancel := context.WithTimeout(context.Background(), c.timeout)
+	defer cancel()
+
+	// Execute HTTP request
+	res, err := c.httpClient.Do(req.WithContext(ctx))
+	if err != nil {
+		return nil, err
+	}
+
+	defer res.Body.Close()
+	return res, nil
+}
+
 // Push the input timeseries to the remote endpoint
 func (c *Client) Push(timeseries []prompb.TimeSeries) (*http.Response, error) {
 	// Create write request