Merge branch 'main' into njpm/rw2-sync-main-conflicts

This commit is meant to be used to get a diff with the resolved
conflicts in another branch.

.github/CODEOWNERS

@@ -1,6 +1,6 @@
 /web/ui @juliusv
 /web/ui/module @juliusv @nexucis
-/storage/remote @csmarchbanks @cstyan @bwplotka @tomwilkie
+/storage/remote @cstyan @bwplotka @tomwilkie
 /storage/remote/otlptranslator @gouthamve @jesusvazquez
 /discovery/kubernetes @brancz
 /tsdb @jesusvazquez

.github/workflows/buf.yml

@@ -23,7 +23,7 @@ jobs:
         with:
           input: 'prompb'
           against: 'https://github.com/prometheus/prometheus.git#branch=main,ref=HEAD~1,subdir=prompb'
-      - uses: bufbuild/buf-push-action@342fc4cdcf29115a01cf12a2c6dd6aac68dc51e1 # v1.1.1
+      - uses: bufbuild/buf-push-action@a654ff18effe4641ebea4a4ce242c49800728459 # v1.1.1
         with:
           input: 'prompb'
           buf_token: ${{ secrets.BUF_TOKEN }}

.github/workflows/ci.yml

@@ -197,7 +197,7 @@ jobs:
         uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1
       - uses: prometheus/promci@3cb0c3871f223bd5ce1226995bd52ffb314798b6 # v0.1.0
       - name: Install nodejs
-        uses: actions/setup-node@5e21ff4d9bc1a8cf6de233a3057d20ec6b3fb69d # v3.8.1
+        uses: actions/setup-node@b39b52d1213e96004bfcb1c61a8a6fa8ab84f3e8 # v4.0.1
         with:
           node-version-file: "web/ui/.nvmrc"
           registry-url: "https://registry.npmjs.org"

.github/workflows/codeql-analysis.yml

@@ -30,12 +30,12 @@ jobs:
           go-version: 1.21.x
 
       - name: Initialize CodeQL
-        uses: github/codeql-action/init@407ffafae6a767df3e0230c3df91b6443ae8df75 # v2.22.8
+        uses: github/codeql-action/init@012739e5082ff0c22ca6d6ab32e07c36df03c4a4 # v3.22.12
         with:
           languages: ${{ matrix.language }}
 
       - name: Autobuild
-        uses: github/codeql-action/autobuild@407ffafae6a767df3e0230c3df91b6443ae8df75 # v2.22.8
+        uses: github/codeql-action/autobuild@012739e5082ff0c22ca6d6ab32e07c36df03c4a4 # v3.22.12
 
       - name: Perform CodeQL Analysis
-        uses: github/codeql-action/analyze@407ffafae6a767df3e0230c3df91b6443ae8df75 # v2.22.8
+        uses: github/codeql-action/analyze@012739e5082ff0c22ca6d6ab32e07c36df03c4a4 # v3.22.12

.github/workflows/fuzzing.yml

@@ -21,7 +21,7 @@ jobs:
           fuzz-seconds: 600
           dry-run: false
       - name: Upload Crash
-        uses: actions/upload-artifact@a8a3f3ad30e3422c9c7b888a15615d19a852ae32 # v3.1.3
+        uses: actions/upload-artifact@c7d193f32edcb7bfad88892161225aeda64e9392 # v4.0.0
         if: failure() && steps.build.outcome == 'success'
         with:
           name: artifacts

.github/workflows/scorecards.yml

@@ -37,7 +37,7 @@ jobs:
       # Upload the results as artifacts (optional). Commenting out will disable uploads of run results in SARIF
       # format to the repository Actions tab.
       - name: "Upload artifact"
-        uses: actions/upload-artifact@a8a3f3ad30e3422c9c7b888a15615d19a852ae32 # tag=v3.1.3
+        uses: actions/upload-artifact@c7d193f32edcb7bfad88892161225aeda64e9392 # tag=v4.0.0
         with:
           name: SARIF file
           path: results.sarif
@@ -45,6 +45,6 @@ jobs:
 
       # Upload the results to GitHub's code scanning dashboard.
       - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@407ffafae6a767df3e0230c3df91b6443ae8df75 # tag=v2.22.8
+        uses: github/codeql-action/upload-sarif@012739e5082ff0c22ca6d6ab32e07c36df03c4a4 # tag=v3.22.12
         with:
           sarif_file: results.sarif

CHANGELOG.md

@@ -1,9 +1,47 @@
 # Changelog
 
-## unreleased
+## 2.49.1 / 2024-01-15
 
-* [ENHANCEMENT] TSDB: Make the wlog watcher read segments synchronously when not tailing. #13224
+* [BUGFIX] TSDB: Fixed a wrong `q=` value in scrape accept header #13313
-* [BUGFIX] Agent: Participate in notify calls. #13223
+
+## 2.49.0 / 2024-01-15
+
+* [FEATURE] Promtool: Add `--run` flag promtool test rules command. #12206
+* [FEATURE] SD: Add support for `NS` records to DNS SD. #13219
+* [FEATURE] UI: Add heatmap visualization setting in the Graph tab, useful histograms. #13096 #13371
+* [FEATURE] Scraping: Add `scrape_config.enable_compression` (default true) to disable gzip compression when scraping the target. #13166
+* [FEATURE] PromQL: Add a `promql-experimental-functions` feature flag containing some new experimental PromQL functions. #13103 NOTE: More experimental functions might be added behind the same feature flag in the future. Added functions:
+  * Experimental `mad_over_time` (median absolute deviation around the median) function. #13059
+  * Experimental `sort_by_label` and `sort_by_label_desc` functions allowing sorting returned series by labels. #11299
+* [FEATURE] SD: Add `__meta_linode_gpus` label to Linode SD. #13097
+* [FEATURE] API: Add `exclude_alerts` query parameter to `/api/v1/rules` to only return recording rules. #12999
+* [FEATURE] TSDB: --storage.tsdb.retention.time flag value is now exposed as a `prometheus_tsdb_retention_limit_seconds` metric. #12986
+* [FEATURE] Scraping: Add ability to specify priority of scrape protocols to accept during scrape (e.g. to scrape Prometheus proto format for certain jobs). This can be changed by setting `global.scrape_protocols` and `scrape_config.scrape_protocols`. #12738
+* [ENHANCEMENT] Scraping: Automated handling of scraping histograms that violate `scrape_config.native_histogram_bucket_limit` setting. #13129
+* [ENHANCEMENT] Scraping: Optimized memory allocations when scraping. #12992
+* [ENHANCEMENT] SD: Added cache for Azure SD to avoid rate-limits. #12622
+* [ENHANCEMENT] TSDB: Various improvements to OOO exemplar scraping. E.g. allowing ingestion of exemplars with the same timestamp, but with different labels. #13021
+* [ENHANCEMENT] API: Optimize `/api/v1/labels` and `/api/v1/label/<label_name>/values` when 1 set of matchers are used. #12888
+* [ENHANCEMENT] TSDB: Various optimizations for TSDB block index, head mmap chunks and WAL, reducing latency and memory allocations (improving API calls, compaction queries etc). #12997 #13058 #13056 #13040
+* [ENHANCEMENT] PromQL: Optimize memory allocations and latency when querying float histograms. #12954
+* [ENHANCEMENT] Rules: Instrument TraceID in log lines for rule evaluations. #13034
+* [ENHANCEMENT] PromQL: Optimize memory allocations in query_range calls. #13043
+* [ENHANCEMENT] Promtool: unittest interval now defaults to evaluation_intervals when not set. #12729
+* [BUGFIX] SD: Fixed Azure SD public IP reporting #13241
+* [BUGFIX] API: Fix inaccuracies in posting cardinality statistics. #12653
+* [BUGFIX] PromQL: Fix inaccuracies of `histogram_quantile` with classic histograms. #13153
+* [BUGFIX] TSDB: Fix rare fails or inaccurate queries with OOO samples. #13115
+* [BUGFIX] TSDB: Fix rare panics on append commit when exemplars are used. #13092
+* [BUGFIX] TSDB: Fix exemplar WAL storage, so remote write can send/receive samples before exemplars. #13113
+* [BUGFIX] Mixins: Fix `url` filter on remote write dashboards. #10721
+* [BUGFIX] PromQL/TSDB: Various fixes to float histogram operations. #12891 #12977 #12609 #13190 #13189 #13191 #13201 #13212 #13208
+* [BUGFIX] Promtool: Fix int32 overflow issues for 32-bit architectures. #12978
+* [BUGFIX] SD: Fix Azure VM Scale Set NIC issue. #13283
+
+## 2.48.1 / 2023-12-07
+
+* [BUGFIX] TSDB: Make the wlog watcher read segments synchronously when not tailing. #13224
+* [BUGFIX] Agent: Participate in notify calls (fixes slow down in remote write handling introduced in 2.45). #13223
 
 ## 2.48.0 / 2023-11-16
 

MAINTAINERS.md

@@ -8,8 +8,10 @@ Julien Pivotto (<roidelapluie@prometheus.io> / @roidelapluie) and Levi Harrison
   * `k8s`: Frederic Branczyk (<fbranczyk@gmail.com> / @brancz)
 * `documentation`
   * `prometheus-mixin`: Matthias Loibl (<mail@matthiasloibl.com> / @metalmatze)
+* `model/histogram` and other code related to native histograms: Björn Rabenstein (<beorn@grafana.com> / @beorn7), 
+George Krajcsovits (<gyorgy.krajcsovits@grafana.com> / @krajorama)
 * `storage`
-  * `remote`: Chris Marchbanks (<csmarchbanks@gmail.com> / @csmarchbanks), Callum Styan (<callumstyan@gmail.com> / @cstyan), Bartłomiej Płotka (<bwplotka@gmail.com> / @bwplotka), Tom Wilkie (<tom.wilkie@gmail.com> / @tomwilkie)
+  * `remote`: Callum Styan (<callumstyan@gmail.com> / @cstyan), Bartłomiej Płotka (<bwplotka@gmail.com> / @bwplotka), Tom Wilkie (<tom.wilkie@gmail.com> / @tomwilkie)
 * `tsdb`: Ganesh Vernekar (<ganesh@grafana.com> / @codesome), Bartłomiej Płotka (<bwplotka@gmail.com> / @bwplotka), Jesús Vázquez (<jesus.vazquez@grafana.com> / @jesusvazquez)
   * `agent`: Robert Fratto (<robert.fratto@grafana.com> / @rfratto)
 * `web`
@@ -17,6 +19,7 @@ Julien Pivotto (<roidelapluie@prometheus.io> / @roidelapluie) and Levi Harrison
     * `module`: Augustin Husson (<husson.augustin@gmail.com> @nexucis)
 * `Makefile` and related build configuration: Simon Pasquier (<pasquier.simon@gmail.com> / @simonpasquier), Ben Kochie (<superq@gmail.com> / @SuperQ)
 
+
 For the sake of brevity, not all subtrees are explicitly listed. Due to the
 size of this repository, the natural changes in focus of maintainers over time,
 and nuances of where particular features live, this list will always be

README.md

@@ -14,7 +14,7 @@ examples and guides.</p>
 [![CII Best Practices](https://bestpractices.coreinfrastructure.org/projects/486/badge)](https://bestpractices.coreinfrastructure.org/projects/486)
 [![Gitpod ready-to-code](https://img.shields.io/badge/Gitpod-ready--to--code-blue?logo=gitpod)](https://gitpod.io/#https://github.com/prometheus/prometheus)
 [![Fuzzing Status](https://oss-fuzz-build-logs.storage.googleapis.com/badges/prometheus.svg)](https://bugs.chromium.org/p/oss-fuzz/issues/list?sort=-opened&can=1&q=proj:prometheus)
-[![OpenSSF Scorecard](https://api.securityscorecards.dev/projects/github.com/prometheus/prometheus/badge)](https://api.securityscorecards.dev/projects/github.com/prometheus/prometheus)
+[![OpenSSF Scorecard](https://api.securityscorecards.dev/projects/github.com/prometheus/prometheus/badge)](https://securityscorecards.dev/viewer/?uri=github.com/prometheus/prometheus)
 
 </div>
 

RELEASE.md

@@ -54,7 +54,8 @@ Release cadence of first pre-releases being cut is 6 weeks.
 | v2.47          | 2023-08-23                                 | Bryan Boreham (GitHub: @bboreham)           |
 | v2.48          | 2023-10-04                                 | Levi Harrison (GitHub: @LeviHarrison)       |
 | v2.49          | 2023-12-05                                 | Bartek Plotka (GitHub: @bwplotka)           |
-| v2.50          | 2024-01-16                                 | **searching for volunteer**                 |
+| v2.50          | 2024-01-16                                 | Augustin Husson (GitHub: @nexucis)          |
+| v2.51          | 2024-02-13                                 | **searching for volunteer**                 |
 
 If you are interested in volunteering please create a pull request against the [prometheus/prometheus](https://github.com/prometheus/prometheus) repository and propose yourself for the release series of your choice.
 

VERSION

@@ -1 +1 @@
-2.48.0
+2.49.1

cmd/prometheus/main.go

@@ -33,6 +33,7 @@ import (
 	"syscall"
 	"time"
 
+	"github.com/KimMachineGun/automemlimit/memlimit"
 	"github.com/alecthomas/kingpin/v2"
 	"github.com/alecthomas/units"
 	"github.com/go-kit/log"
@@ -136,6 +137,7 @@ type flagConfig struct {
 	forGracePeriod      model.Duration
 	outageTolerance     model.Duration
 	resendDelay         model.Duration
+	maxConcurrentEvals  int64
 	web                 web.Options
 	scrape              scrape.Options
 	tsdb                tsdbOptions
@@ -148,14 +150,20 @@ type flagConfig struct {
 	RemoteFlushDeadline model.Duration
 
 	featureList   []string
+	memlimitRatio float64
 	// These options are extracted from featureList
 	// for ease of use.
 	enableExpandExternalLabels bool
 	enableNewSDManager         bool
 	enablePerStepStats         bool
 	enableAutoGOMAXPROCS       bool
+<<<<<<< HEAD
 	// todo: how to use the enable feature flag properly + use the remote format enum type
 	rwFormat int
+=======
+	enableAutoGOMEMLIMIT       bool
+	enableConcurrentRuleEval   bool
+>>>>>>> main
 
 	prometheusURL   string
 	corsRegexString string
@@ -202,6 +210,12 @@ func (c *flagConfig) setFeatureListOptions(logger log.Logger) error {
 			case "auto-gomaxprocs":
 				c.enableAutoGOMAXPROCS = true
 				level.Info(logger).Log("msg", "Automatically set GOMAXPROCS to match Linux container CPU quota")
+			case "auto-gomemlimit":
+				c.enableAutoGOMEMLIMIT = true
+				level.Info(logger).Log("msg", "Automatically set GOMEMLIMIT to match Linux container or system memory limit")
+			case "concurrent-rule-eval":
+				c.enableConcurrentRuleEval = true
+				level.Info(logger).Log("msg", "Experimental concurrent rule evaluation enabled.")
 			case "no-default-scrape-port":
 				c.scrape.NoDefaultPort = true
 				level.Info(logger).Log("msg", "No default port will be appended to scrape targets' addresses.")
@@ -267,6 +281,9 @@ func main() {
 	a.Flag("web.listen-address", "Address to listen on for UI, API, and telemetry.").
 		Default("0.0.0.0:9090").StringVar(&cfg.web.ListenAddress)
 
+	a.Flag("auto-gomemlimit.ratio", "The ratio of reserved GOMEMLIMIT memory to the detected maximum container or system memory").
+		Default("0.9").FloatVar(&cfg.memlimitRatio)
+
 	webConfig := a.Flag(
 		"web.config.file",
 		"[EXPERIMENTAL] Path to configuration file that can enable TLS or authentication.",
@@ -407,6 +424,9 @@ func main() {
 	serverOnlyFlag(a, "rules.alert.resend-delay", "Minimum amount of time to wait before resending an alert to Alertmanager.").
 		Default("1m").SetValue(&cfg.resendDelay)
 
+	serverOnlyFlag(a, "rules.max-concurrent-evals", "Global concurrency limit for independent rules that can run concurrently.").
+		Default("4").Int64Var(&cfg.maxConcurrentEvals)
+
 	a.Flag("scrape.adjust-timestamps", "Adjust scrape timestamps by up to `scrape.timestamp-tolerance` to align them to the intended schedule. See https://github.com/prometheus/prometheus/issues/7846 for more context. Experimental. This flag will be removed in a future release.").
 		Hidden().Default("true").BoolVar(&scrape.AlignScrapeTimestamps)
 
@@ -434,7 +454,11 @@ func main() {
 	a.Flag("scrape.discovery-reload-interval", "Interval used by scrape manager to throttle target groups updates.").
 		Hidden().Default("5s").SetValue(&cfg.scrape.DiscoveryReloadInterval)
 
+<<<<<<< HEAD
 	a.Flag("enable-feature", "Comma separated feature names to enable. Valid options: agent, exemplar-storage, expand-external-labels, memory-snapshot-on-shutdown, promql-at-modifier, promql-negative-offset, promql-per-step-stats, promql-experimental-functions, remote-write-receiver (DEPRECATED), extra-scrape-metrics, new-service-discovery-manager, auto-gomaxprocs, no-default-scrape-port, native-histograms, otlp-write-receiver, metadata-wal-records. See https://prometheus.io/docs/prometheus/latest/feature_flags/ for more details.").
+=======
+	a.Flag("enable-feature", "Comma separated feature names to enable. Valid options: agent, auto-gomemlimit, exemplar-storage, expand-external-labels, memory-snapshot-on-shutdown, promql-at-modifier, promql-negative-offset, promql-per-step-stats, promql-experimental-functions, remote-write-receiver (DEPRECATED), extra-scrape-metrics, new-service-discovery-manager, auto-gomaxprocs, no-default-scrape-port, native-histograms, otlp-write-receiver. See https://prometheus.io/docs/prometheus/latest/feature_flags/ for more details.").
+>>>>>>> main
 		Default("").StringsVar(&cfg.featureList)
 
 	a.Flag("remote-write-format", "remote write proto format to use, valid options: 0 (1.0), 1 (reduced format), 3 (min64 format)").
@@ -475,6 +499,11 @@ func main() {
 		os.Exit(3)
 	}
 
+	if cfg.memlimitRatio <= 0.0 || cfg.memlimitRatio > 1.0 {
+		fmt.Fprintf(os.Stderr, "--auto-gomemlimit.ratio must be greater than 0 and less than or equal to 1.")
+		os.Exit(1)
+	}
+
 	localStoragePath := cfg.serverStoragePath
 	if agentMode {
 		localStoragePath = cfg.agentStoragePath
@@ -638,9 +667,16 @@ func main() {
 		level.Error(logger).Log("msg", "failed to register Kubernetes client metrics", "err", err)
 		os.Exit(1)
 	}
+
+	sdMetrics, err := discovery.CreateAndRegisterSDMetrics(prometheus.DefaultRegisterer)
+	if err != nil {
+		level.Error(logger).Log("msg", "failed to register service discovery metrics", "err", err)
+		os.Exit(1)
+	}
+
 	if cfg.enableNewSDManager {
 		{
-			discMgr := discovery.NewManager(ctxScrape, log.With(logger, "component", "discovery manager scrape"), prometheus.DefaultRegisterer, discovery.Name("scrape"))
+			discMgr := discovery.NewManager(ctxScrape, log.With(logger, "component", "discovery manager scrape"), prometheus.DefaultRegisterer, sdMetrics, discovery.Name("scrape"))
 			if discMgr == nil {
 				level.Error(logger).Log("msg", "failed to create a discovery manager scrape")
 				os.Exit(1)
@@ -649,7 +685,7 @@ func main() {
 		}
 
 		{
-			discMgr := discovery.NewManager(ctxNotify, log.With(logger, "component", "discovery manager notify"), prometheus.DefaultRegisterer, discovery.Name("notify"))
+			discMgr := discovery.NewManager(ctxNotify, log.With(logger, "component", "discovery manager notify"), prometheus.DefaultRegisterer, sdMetrics, discovery.Name("notify"))
 			if discMgr == nil {
 				level.Error(logger).Log("msg", "failed to create a discovery manager notify")
 				os.Exit(1)
@@ -658,7 +694,7 @@ func main() {
 		}
 	} else {
 		{
-			discMgr := legacymanager.NewManager(ctxScrape, log.With(logger, "component", "discovery manager scrape"), prometheus.DefaultRegisterer, legacymanager.Name("scrape"))
+			discMgr := legacymanager.NewManager(ctxScrape, log.With(logger, "component", "discovery manager scrape"), prometheus.DefaultRegisterer, sdMetrics, legacymanager.Name("scrape"))
 			if discMgr == nil {
 				level.Error(logger).Log("msg", "failed to create a discovery manager scrape")
 				os.Exit(1)
@@ -667,7 +703,7 @@ func main() {
 		}
 
 		{
-			discMgr := legacymanager.NewManager(ctxNotify, log.With(logger, "component", "discovery manager notify"), prometheus.DefaultRegisterer, legacymanager.Name("notify"))
+			discMgr := legacymanager.NewManager(ctxNotify, log.With(logger, "component", "discovery manager notify"), prometheus.DefaultRegisterer, sdMetrics, legacymanager.Name("notify"))
 			if discMgr == nil {
 				level.Error(logger).Log("msg", "failed to create a discovery manager notify")
 				os.Exit(1)
@@ -703,6 +739,20 @@ func main() {
 		}
 	}
 
+	if cfg.enableAutoGOMEMLIMIT {
+		if _, err := memlimit.SetGoMemLimitWithOpts(
+			memlimit.WithRatio(cfg.memlimitRatio),
+			memlimit.WithProvider(
+				memlimit.ApplyFallback(
+					memlimit.FromCgroup,
+					memlimit.FromSystem,
+				),
+			),
+		); err != nil {
+			level.Warn(logger).Log("component", "automemlimit", "msg", "Failed to set GOMEMLIMIT automatically", "err", err)
+		}
+	}
+
 	if !agentMode {
 		opts := promql.EngineOpts{
 			Logger:                   log.With(logger, "component", "query engine"),
@@ -733,6 +783,8 @@ func main() {
 			OutageTolerance:        time.Duration(cfg.outageTolerance),
 			ForGracePeriod:         time.Duration(cfg.forGracePeriod),
 			ResendDelay:            time.Duration(cfg.resendDelay),
+			MaxConcurrentEvals:     cfg.maxConcurrentEvals,
+			ConcurrentEvalsEnabled: cfg.enableConcurrentRuleEval,
 		})
 	}
 
@@ -1655,6 +1707,7 @@ func (opts tsdbOptions) ToTSDBOptions() tsdb.Options {
 		EnableMemorySnapshotOnShutdown: opts.EnableMemorySnapshotOnShutdown,
 		EnableNativeHistograms:         opts.EnableNativeHistograms,
 		OutOfOrderTimeWindow:           opts.OutOfOrderTimeWindow,
+		EnableOverlappingCompaction:    true,
 	}
 }
 

cmd/prometheus/main_unix_test.go

@@ -12,7 +12,6 @@
 // limitations under the License.
 //
 //go:build !windows
-// +build !windows
 
 package main
 

cmd/promtool/analyze.go

@@ -0,0 +1,370 @@
+// Copyright 2023 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package main
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"io"
+	"math"
+	"net/http"
+	"net/url"
+	"os"
+	"sort"
+	"strconv"
+	"strings"
+	"time"
+
+	v1 "github.com/prometheus/client_golang/api/prometheus/v1"
+	"github.com/prometheus/common/model"
+
+	"github.com/prometheus/prometheus/model/labels"
+)
+
+var (
+	errNotNativeHistogram = fmt.Errorf("not a native histogram")
+	errNotEnoughData      = fmt.Errorf("not enough data")
+
+	outputHeader = `Bucket stats for each histogram series over time
+------------------------------------------------
+First the min, avg, and max number of populated buckets, followed by the total
+number of buckets (only if different from the max number of populated buckets
+which is typical for classic but not native histograms).`
+	outputFooter = `Aggregated bucket stats
+-----------------------
+Each line shows min/avg/max over the series above.`
+)
+
+type QueryAnalyzeConfig struct {
+	metricType string
+	duration   time.Duration
+	time       string
+	matchers   []string
+}
+
+// run retrieves metrics that look like conventional histograms (i.e. have _bucket
+// suffixes) or native histograms, depending on metricType flag.
+func (c *QueryAnalyzeConfig) run(url *url.URL, roundtripper http.RoundTripper) error {
+	if c.metricType != "histogram" {
+		return fmt.Errorf("analyze type is %s, must be 'histogram'", c.metricType)
+	}
+
+	ctx := context.Background()
+
+	api, err := newAPI(url, roundtripper, nil)
+	if err != nil {
+		return err
+	}
+
+	var endTime time.Time
+	if c.time != "" {
+		endTime, err = parseTime(c.time)
+		if err != nil {
+			return fmt.Errorf("error parsing time '%s': %w", c.time, err)
+		}
+	} else {
+		endTime = time.Now()
+	}
+
+	return c.getStatsFromMetrics(ctx, api, endTime, os.Stdout, c.matchers)
+}
+
+func (c *QueryAnalyzeConfig) getStatsFromMetrics(ctx context.Context, api v1.API, endTime time.Time, out io.Writer, matchers []string) error {
+	fmt.Fprintf(out, "%s\n\n", outputHeader)
+	metastatsNative := newMetaStatistics()
+	metastatsClassic := newMetaStatistics()
+	for _, matcher := range matchers {
+		seriesSel := seriesSelector(matcher, c.duration)
+		matrix, err := querySamples(ctx, api, seriesSel, endTime)
+		if err != nil {
+			return err
+		}
+
+		matrices := make(map[string]model.Matrix)
+		for _, series := range matrix {
+			// We do not handle mixed types. If there are float values, we assume it is a
+			// classic histogram, otherwise we assume it is a native histogram, and we
+			// ignore series with errors if they do not match the expected type.
+			if len(series.Values) == 0 {
+				stats, err := calcNativeBucketStatistics(series)
+				if err != nil {
+					if errors.Is(err, errNotNativeHistogram) || errors.Is(err, errNotEnoughData) {
+						continue
+					}
+					return err
+				}
+				fmt.Fprintf(out, "- %s (native): %v\n", series.Metric, *stats)
+				metastatsNative.update(stats)
+			} else {
+				lbs := model.LabelSet(series.Metric).Clone()
+				if _, ok := lbs["le"]; !ok {
+					continue
+				}
+				metricName := string(lbs[labels.MetricName])
+				if !strings.HasSuffix(metricName, "_bucket") {
+					continue
+				}
+				delete(lbs, labels.MetricName)
+				delete(lbs, "le")
+				key := formatSeriesName(metricName, lbs)
+				matrices[key] = append(matrices[key], series)
+			}
+		}
+
+		for key, matrix := range matrices {
+			stats, err := calcClassicBucketStatistics(matrix)
+			if err != nil {
+				if errors.Is(err, errNotEnoughData) {
+					continue
+				}
+				return err
+			}
+			fmt.Fprintf(out, "- %s (classic): %v\n", key, *stats)
+			metastatsClassic.update(stats)
+		}
+	}
+	fmt.Fprintf(out, "\n%s\n", outputFooter)
+	if metastatsNative.Count() > 0 {
+		fmt.Fprintf(out, "\nNative %s\n", metastatsNative)
+	}
+	if metastatsClassic.Count() > 0 {
+		fmt.Fprintf(out, "\nClassic %s\n", metastatsClassic)
+	}
+	return nil
+}
+
+func seriesSelector(metricName string, duration time.Duration) string {
+	builder := strings.Builder{}
+	builder.WriteString(metricName)
+	builder.WriteRune('[')
+	builder.WriteString(duration.String())
+	builder.WriteRune(']')
+	return builder.String()
+}
+
+func formatSeriesName(metricName string, lbs model.LabelSet) string {
+	builder := strings.Builder{}
+	builder.WriteString(metricName)
+	builder.WriteString(lbs.String())
+	return builder.String()
+}
+
+func querySamples(ctx context.Context, api v1.API, query string, end time.Time) (model.Matrix, error) {
+	values, _, err := api.Query(ctx, query, end)
+	if err != nil {
+		return nil, err
+	}
+
+	matrix, ok := values.(model.Matrix)
+	if !ok {
+		return nil, fmt.Errorf("query of buckets resulted in non-Matrix")
+	}
+
+	return matrix, nil
+}
+
+// minPop/avgPop/maxPop is for the number of populated (non-zero) buckets.
+// total is the total number of buckets across all samples in the series,
+// populated or not.
+type statistics struct {
+	minPop, maxPop, total int
+	avgPop                float64
+}
+
+func (s statistics) String() string {
+	if s.maxPop == s.total {
+		return fmt.Sprintf("%d/%.3f/%d", s.minPop, s.avgPop, s.maxPop)
+	}
+	return fmt.Sprintf("%d/%.3f/%d/%d", s.minPop, s.avgPop, s.maxPop, s.total)
+}
+
+func calcClassicBucketStatistics(matrix model.Matrix) (*statistics, error) {
+	numBuckets := len(matrix)
+
+	stats := &statistics{
+		minPop: math.MaxInt,
+		total:  numBuckets,
+	}
+
+	if numBuckets == 0 || len(matrix[0].Values) < 2 {
+		return stats, errNotEnoughData
+	}
+
+	numSamples := len(matrix[0].Values)
+
+	sortMatrix(matrix)
+
+	totalPop := 0
+	for timeIdx := 0; timeIdx < numSamples; timeIdx++ {
+		curr, err := getBucketCountsAtTime(matrix, numBuckets, timeIdx)
+		if err != nil {
+			return stats, err
+		}
+		countPop := 0
+		for _, b := range curr {
+			if b != 0 {
+				countPop++
+			}
+		}
+
+		totalPop += countPop
+		if stats.minPop > countPop {
+			stats.minPop = countPop
+		}
+		if stats.maxPop < countPop {
+			stats.maxPop = countPop
+		}
+	}
+	stats.avgPop = float64(totalPop) / float64(numSamples)
+	return stats, nil
+}
+
+func sortMatrix(matrix model.Matrix) {
+	sort.SliceStable(matrix, func(i, j int) bool {
+		return getLe(matrix[i]) < getLe(matrix[j])
+	})
+}
+
+func getLe(series *model.SampleStream) float64 {
+	lbs := model.LabelSet(series.Metric)
+	le, _ := strconv.ParseFloat(string(lbs["le"]), 64)
+	return le
+}
+
+func getBucketCountsAtTime(matrix model.Matrix, numBuckets, timeIdx int) ([]int, error) {
+	counts := make([]int, numBuckets)
+	if timeIdx >= len(matrix[0].Values) {
+		// Just return zeroes instead of erroring out so we can get partial results.
+		return counts, nil
+	}
+	counts[0] = int(matrix[0].Values[timeIdx].Value)
+	for i, bucket := range matrix[1:] {
+		if timeIdx >= len(bucket.Values) {
+			// Just return zeroes instead of erroring out so we can get partial results.
+			return counts, nil
+		}
+		curr := bucket.Values[timeIdx]
+		prev := matrix[i].Values[timeIdx]
+		// Assume the results are nicely aligned.
+		if curr.Timestamp != prev.Timestamp {
+			return counts, fmt.Errorf("matrix result is not time aligned")
+		}
+		counts[i+1] = int(curr.Value - prev.Value)
+	}
+	return counts, nil
+}
+
+type bucketBounds struct {
+	boundaries   int32
+	upper, lower float64
+}
+
+func makeBucketBounds(b *model.HistogramBucket) bucketBounds {
+	return bucketBounds{
+		boundaries: b.Boundaries,
+		upper:      float64(b.Upper),
+		lower:      float64(b.Lower),
+	}
+}
+
+func calcNativeBucketStatistics(series *model.SampleStream) (*statistics, error) {
+	stats := &statistics{
+		minPop: math.MaxInt,
+	}
+
+	overall := make(map[bucketBounds]struct{})
+	totalPop := 0
+	if len(series.Histograms) == 0 {
+		return nil, errNotNativeHistogram
+	}
+	if len(series.Histograms) == 1 {
+		return nil, errNotEnoughData
+	}
+	for _, histogram := range series.Histograms {
+		for _, bucket := range histogram.Histogram.Buckets {
+			bb := makeBucketBounds(bucket)
+			overall[bb] = struct{}{}
+		}
+		countPop := len(histogram.Histogram.Buckets)
+
+		totalPop += countPop
+		if stats.minPop > countPop {
+			stats.minPop = countPop
+		}
+		if stats.maxPop < countPop {
+			stats.maxPop = countPop
+		}
+	}
+	stats.avgPop = float64(totalPop) / float64(len(series.Histograms))
+	stats.total = len(overall)
+	return stats, nil
+}
+
+type distribution struct {
+	min, max, count int
+	avg             float64
+}
+
+func newDistribution() distribution {
+	return distribution{
+		min: math.MaxInt,
+	}
+}
+
+func (d *distribution) update(num int) {
+	if d.min > num {
+		d.min = num
+	}
+	if d.max < num {
+		d.max = num
+	}
+	d.count++
+	d.avg += float64(num)/float64(d.count) - d.avg/float64(d.count)
+}
+
+func (d distribution) String() string {
+	return fmt.Sprintf("%d/%.3f/%d", d.min, d.avg, d.max)
+}
+
+type metaStatistics struct {
+	minPop, avgPop, maxPop, total distribution
+}
+
+func newMetaStatistics() *metaStatistics {
+	return &metaStatistics{
+		minPop: newDistribution(),
+		avgPop: newDistribution(),
+		maxPop: newDistribution(),
+		total:  newDistribution(),
+	}
+}
+
+func (ms metaStatistics) Count() int {
+	return ms.minPop.count
+}
+
+func (ms metaStatistics) String() string {
+	if ms.maxPop == ms.total {
+		return fmt.Sprintf("histogram series (%d in total):\n- min populated: %v\n- avg populated: %v\n- max populated: %v", ms.Count(), ms.minPop, ms.avgPop, ms.maxPop)
+	}
+	return fmt.Sprintf("histogram series (%d in total):\n- min populated: %v\n- avg populated: %v\n- max populated: %v\n- total: %v", ms.Count(), ms.minPop, ms.avgPop, ms.maxPop, ms.total)
+}
+
+func (ms *metaStatistics) update(s *statistics) {
+	ms.minPop.update(s.minPop)
+	ms.avgPop.update(int(s.avgPop))
+	ms.maxPop.update(s.maxPop)
+	ms.total.update(s.total)
+}

cmd/promtool/analyze_test.go

@@ -0,0 +1,170 @@
+// Copyright 2023 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package main
+
+import (
+	"fmt"
+	"testing"
+
+	"github.com/stretchr/testify/require"
+
+	"github.com/prometheus/common/model"
+)
+
+var (
+	exampleMatrix = model.Matrix{
+		&model.SampleStream{
+			Metric: model.Metric{
+				"le": "+Inf",
+			},
+			Values: []model.SamplePair{
+				{
+					Value:     31,
+					Timestamp: 100,
+				},
+				{
+					Value:     32,
+					Timestamp: 200,
+				},
+				{
+					Value:     40,
+					Timestamp: 300,
+				},
+			},
+		},
+		&model.SampleStream{
+			Metric: model.Metric{
+				"le": "0.5",
+			},
+			Values: []model.SamplePair{
+				{
+					Value:     10,
+					Timestamp: 100,
+				},
+				{
+					Value:     11,
+					Timestamp: 200,
+				},
+				{
+					Value:     11,
+					Timestamp: 300,
+				},
+			},
+		},
+		&model.SampleStream{
+			Metric: model.Metric{
+				"le": "10",
+			},
+			Values: []model.SamplePair{
+				{
+					Value:     30,
+					Timestamp: 100,
+				},
+				{
+					Value:     31,
+					Timestamp: 200,
+				},
+				{
+					Value:     37,
+					Timestamp: 300,
+				},
+			},
+		},
+		&model.SampleStream{
+			Metric: model.Metric{
+				"le": "2",
+			},
+			Values: []model.SamplePair{
+				{
+					Value:     25,
+					Timestamp: 100,
+				},
+				{
+					Value:     26,
+					Timestamp: 200,
+				},
+				{
+					Value:     27,
+					Timestamp: 300,
+				},
+			},
+		},
+	}
+	exampleMatrixLength = len(exampleMatrix)
+)
+
+func init() {
+	sortMatrix(exampleMatrix)
+}
+
+func TestGetBucketCountsAtTime(t *testing.T) {
+	cases := []struct {
+		matrix   model.Matrix
+		length   int
+		timeIdx  int
+		expected []int
+	}{
+		{
+			exampleMatrix,
+			exampleMatrixLength,
+			0,
+			[]int{10, 15, 5, 1},
+		},
+		{
+			exampleMatrix,
+			exampleMatrixLength,
+			1,
+			[]int{11, 15, 5, 1},
+		},
+		{
+			exampleMatrix,
+			exampleMatrixLength,
+			2,
+			[]int{11, 16, 10, 3},
+		},
+	}
+
+	for _, c := range cases {
+		t.Run(fmt.Sprintf("exampleMatrix@%d", c.timeIdx), func(t *testing.T) {
+			res, err := getBucketCountsAtTime(c.matrix, c.length, c.timeIdx)
+			require.NoError(t, err)
+			require.Equal(t, c.expected, res)
+		})
+	}
+}
+
+func TestCalcClassicBucketStatistics(t *testing.T) {
+	cases := []struct {
+		matrix   model.Matrix
+		expected *statistics
+	}{
+		{
+			exampleMatrix,
+			&statistics{
+				minPop: 4,
+				avgPop: 4,
+				maxPop: 4,
+				total:  4,
+			},
+		},
+	}
+
+	for i, c := range cases {
+		t.Run(fmt.Sprintf("case %d", i), func(t *testing.T) {
+			res, err := calcClassicBucketStatistics(c.matrix)
+			require.NoError(t, err)
+			require.Equal(t, c.expected, res)
+		})
+	}
+}

cmd/promtool/main.go

@@ -35,9 +35,7 @@ import (
 	"github.com/go-kit/log"
 	"github.com/google/pprof/profile"
 	"github.com/prometheus/client_golang/api"
-	v1 "github.com/prometheus/client_golang/api/prometheus/v1"
 	"github.com/prometheus/client_golang/prometheus"
-	"github.com/prometheus/client_golang/prometheus/promhttp"
 	"github.com/prometheus/client_golang/prometheus/testutil/promlint"
 	config_util "github.com/prometheus/common/config"
 	"github.com/prometheus/common/model"
@@ -185,6 +183,14 @@ func main() {
 	queryLabelsEnd := queryLabelsCmd.Flag("end", "End time (RFC3339 or Unix timestamp).").String()
 	queryLabelsMatch := queryLabelsCmd.Flag("match", "Series selector. Can be specified multiple times.").Strings()
 
+	queryAnalyzeCfg := &QueryAnalyzeConfig{}
+	queryAnalyzeCmd := queryCmd.Command("analyze", "Run queries against your Prometheus to analyze the usage pattern of certain metrics.")
+	queryAnalyzeCmd.Flag("server", "Prometheus server to query.").Required().URLVar(&serverURL)
+	queryAnalyzeCmd.Flag("type", "Type of metric: histogram.").Required().StringVar(&queryAnalyzeCfg.metricType)
+	queryAnalyzeCmd.Flag("duration", "Time frame to analyze.").Default("1h").DurationVar(&queryAnalyzeCfg.duration)
+	queryAnalyzeCmd.Flag("time", "Query time (RFC3339 or Unix timestamp), defaults to now.").StringVar(&queryAnalyzeCfg.time)
+	queryAnalyzeCmd.Flag("match", "Series selector. Can be specified multiple times.").Required().StringsVar(&queryAnalyzeCfg.matchers)
+
 	pushCmd := app.Command("push", "Push to a Prometheus server.")
 	pushCmd.Flag("http.config.file", "HTTP client configuration file for promtool to connect to Prometheus.").PlaceHolder("<filename>").ExistingFileVar(&httpConfigFilePath)
 	pushMetricsCmd := pushCmd.Command("metrics", "Push metrics to a prometheus remote write (for testing purpose only).")
@@ -204,6 +210,7 @@ func main() {
 		"test-rule-file",
 		"The unit test file.",
 	).Required().ExistingFiles()
+	testRulesDiff := testRulesCmd.Flag("diff", "[Experimental] Print colored differential output between expected & received output.").Default("false").Bool()
 
 	defaultDBPath := "data/"
 	tsdbCmd := app.Command("tsdb", "Run tsdb commands.")
@@ -230,7 +237,7 @@ func main() {
 	dumpPath := tsdbDumpCmd.Arg("db path", "Database path (default is "+defaultDBPath+").").Default(defaultDBPath).String()
 	dumpMinTime := tsdbDumpCmd.Flag("min-time", "Minimum timestamp to dump.").Default(strconv.FormatInt(math.MinInt64, 10)).Int64()
 	dumpMaxTime := tsdbDumpCmd.Flag("max-time", "Maximum timestamp to dump.").Default(strconv.FormatInt(math.MaxInt64, 10)).Int64()
-	dumpMatch := tsdbDumpCmd.Flag("match", "Series selector.").Default("{__name__=~'(?s:.*)'}").String()
+	dumpMatch := tsdbDumpCmd.Flag("match", "Series selector. Can be specified multiple times.").Default("{__name__=~'(?s:.*)'}").Strings()
 
 	importCmd := tsdbCmd.Command("create-blocks-from", "[Experimental] Import samples from input and produce TSDB blocks. Please refer to the storage docs for more details.")
 	importHumanReadable := importCmd.Flag("human-readable", "Print human readable values.").Short('r').Bool()
@@ -369,6 +376,7 @@ func main() {
 				EnableNegativeOffset: true,
 			},
 			*testRulesRun,
+			*testRulesDiff,
 			*testRulesFiles...),
 		)
 
@@ -390,6 +398,9 @@ func main() {
 	case importRulesCmd.FullCommand():
 		os.Exit(checkErr(importRules(serverURL, httpRoundTripper, *importRulesStart, *importRulesEnd, *importRulesOutputDir, *importRulesEvalInterval, *maxBlockDuration, *importRulesFiles...)))
 
+	case queryAnalyzeCmd.FullCommand():
+		os.Exit(checkErr(queryAnalyzeCfg.run(serverURL, httpRoundTripper)))
+
 	case documentationCmd.FullCommand():
 		os.Exit(checkErr(documentcli.GenerateMarkdown(app.Model(), os.Stdout)))
 
@@ -997,246 +1008,6 @@ func checkMetricsExtended(r io.Reader) ([]metricStat, int, error) {
 	return stats, total, nil
 }
 
-// QueryInstant performs an instant query against a Prometheus server.
-func QueryInstant(url *url.URL, roundTripper http.RoundTripper, query, evalTime string, p printer) int {
-	if url.Scheme == "" {
-		url.Scheme = "http"
-	}
-	config := api.Config{
-		Address:      url.String(),
-		RoundTripper: roundTripper,
-	}
-
-	// Create new client.
-	c, err := api.NewClient(config)
-	if err != nil {
-		fmt.Fprintln(os.Stderr, "error creating API client:", err)
-		return failureExitCode
-	}
-
-	eTime := time.Now()
-	if evalTime != "" {
-		eTime, err = parseTime(evalTime)
-		if err != nil {
-			fmt.Fprintln(os.Stderr, "error parsing evaluation time:", err)
-			return failureExitCode
-		}
-	}
-
-	// Run query against client.
-	api := v1.NewAPI(c)
-
-	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
-	val, _, err := api.Query(ctx, query, eTime) // Ignoring warnings for now.
-	cancel()
-	if err != nil {
-		return handleAPIError(err)
-	}
-
-	p.printValue(val)
-
-	return successExitCode
-}
-
-// QueryRange performs a range query against a Prometheus server.
-func QueryRange(url *url.URL, roundTripper http.RoundTripper, headers map[string]string, query, start, end string, step time.Duration, p printer) int {
-	if url.Scheme == "" {
-		url.Scheme = "http"
-	}
-	config := api.Config{
-		Address:      url.String(),
-		RoundTripper: roundTripper,
-	}
-
-	if len(headers) > 0 {
-		config.RoundTripper = promhttp.RoundTripperFunc(func(req *http.Request) (*http.Response, error) {
-			for key, value := range headers {
-				req.Header.Add(key, value)
-			}
-			return roundTripper.RoundTrip(req)
-		})
-	}
-
-	// Create new client.
-	c, err := api.NewClient(config)
-	if err != nil {
-		fmt.Fprintln(os.Stderr, "error creating API client:", err)
-		return failureExitCode
-	}
-
-	var stime, etime time.Time
-
-	if end == "" {
-		etime = time.Now()
-	} else {
-		etime, err = parseTime(end)
-		if err != nil {
-			fmt.Fprintln(os.Stderr, "error parsing end time:", err)
-			return failureExitCode
-		}
-	}
-
-	if start == "" {
-		stime = etime.Add(-5 * time.Minute)
-	} else {
-		stime, err = parseTime(start)
-		if err != nil {
-			fmt.Fprintln(os.Stderr, "error parsing start time:", err)
-			return failureExitCode
-		}
-	}
-
-	if !stime.Before(etime) {
-		fmt.Fprintln(os.Stderr, "start time is not before end time")
-		return failureExitCode
-	}
-
-	if step == 0 {
-		resolution := math.Max(math.Floor(etime.Sub(stime).Seconds()/250), 1)
-		// Convert seconds to nanoseconds such that time.Duration parses correctly.
-		step = time.Duration(resolution) * time.Second
-	}
-
-	// Run query against client.
-	api := v1.NewAPI(c)
-	r := v1.Range{Start: stime, End: etime, Step: step}
-	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
-	val, _, err := api.QueryRange(ctx, query, r) // Ignoring warnings for now.
-	cancel()
-
-	if err != nil {
-		return handleAPIError(err)
-	}
-
-	p.printValue(val)
-	return successExitCode
-}
-
-// QuerySeries queries for a series against a Prometheus server.
-func QuerySeries(url *url.URL, roundTripper http.RoundTripper, matchers []string, start, end string, p printer) int {
-	if url.Scheme == "" {
-		url.Scheme = "http"
-	}
-	config := api.Config{
-		Address:      url.String(),
-		RoundTripper: roundTripper,
-	}
-
-	// Create new client.
-	c, err := api.NewClient(config)
-	if err != nil {
-		fmt.Fprintln(os.Stderr, "error creating API client:", err)
-		return failureExitCode
-	}
-
-	stime, etime, err := parseStartTimeAndEndTime(start, end)
-	if err != nil {
-		fmt.Fprintln(os.Stderr, err)
-		return failureExitCode
-	}
-
-	// Run query against client.
-	api := v1.NewAPI(c)
-	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
-	val, _, err := api.Series(ctx, matchers, stime, etime) // Ignoring warnings for now.
-	cancel()
-
-	if err != nil {
-		return handleAPIError(err)
-	}
-
-	p.printSeries(val)
-	return successExitCode
-}
-
-// QueryLabels queries for label values against a Prometheus server.
-func QueryLabels(url *url.URL, roundTripper http.RoundTripper, matchers []string, name, start, end string, p printer) int {
-	if url.Scheme == "" {
-		url.Scheme = "http"
-	}
-	config := api.Config{
-		Address:      url.String(),
-		RoundTripper: roundTripper,
-	}
-
-	// Create new client.
-	c, err := api.NewClient(config)
-	if err != nil {
-		fmt.Fprintln(os.Stderr, "error creating API client:", err)
-		return failureExitCode
-	}
-
-	stime, etime, err := parseStartTimeAndEndTime(start, end)
-	if err != nil {
-		fmt.Fprintln(os.Stderr, err)
-		return failureExitCode
-	}
-
-	// Run query against client.
-	api := v1.NewAPI(c)
-	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
-	val, warn, err := api.LabelValues(ctx, name, matchers, stime, etime)
-	cancel()
-
-	for _, v := range warn {
-		fmt.Fprintln(os.Stderr, "query warning:", v)
-	}
-	if err != nil {
-		return handleAPIError(err)
-	}
-
-	p.printLabelValues(val)
-	return successExitCode
-}
-
-func handleAPIError(err error) int {
-	var apiErr *v1.Error
-	if errors.As(err, &apiErr) && apiErr.Detail != "" {
-		fmt.Fprintf(os.Stderr, "query error: %v (detail: %s)\n", apiErr, strings.TrimSpace(apiErr.Detail))
-	} else {
-		fmt.Fprintln(os.Stderr, "query error:", err)
-	}
-
-	return failureExitCode
-}
-
-func parseStartTimeAndEndTime(start, end string) (time.Time, time.Time, error) {
-	var (
-		minTime = time.Now().Add(-9999 * time.Hour)
-		maxTime = time.Now().Add(9999 * time.Hour)
-		err     error
-	)
-
-	stime := minTime
-	etime := maxTime
-
-	if start != "" {
-		stime, err = parseTime(start)
-		if err != nil {
-			return stime, etime, fmt.Errorf("error parsing start time: %w", err)
-		}
-	}
-
-	if end != "" {
-		etime, err = parseTime(end)
-		if err != nil {
-			return stime, etime, fmt.Errorf("error parsing end time: %w", err)
-		}
-	}
-	return stime, etime, nil
-}
-
-func parseTime(s string) (time.Time, error) {
-	if t, err := strconv.ParseFloat(s, 64); err == nil {
-		s, ns := math.Modf(t)
-		return time.Unix(int64(s), int64(ns*float64(time.Second))).UTC(), nil
-	}
-	if t, err := time.Parse(time.RFC3339Nano, s); err == nil {
-		return t, nil
-	}
-	return time.Time{}, fmt.Errorf("cannot parse %q to a valid timestamp", s)
-}
-
 type endpointsGroup struct {
 	urlToFilename map[string]string
 	postProcess   func(b []byte) ([]byte, error)
@@ -1390,15 +1161,12 @@ func importRules(url *url.URL, roundTripper http.RoundTripper, start, end, outpu
 		evalInterval:     evalInterval,
 		maxBlockDuration: maxBlockDuration,
 	}
-	client, err := api.NewClient(api.Config{
+	api, err := newAPI(url, roundTripper, nil)
-		Address:      url.String(),
-		RoundTripper: roundTripper,
-	})
 	if err != nil {
 		return fmt.Errorf("new api client error: %w", err)
 	}
 
-	ruleImporter := newRuleImporter(log.NewLogfmtLogger(log.NewSyncWriter(os.Stderr)), cfg, v1.NewAPI(client))
+	ruleImporter := newRuleImporter(log.NewLogfmtLogger(log.NewSyncWriter(os.Stderr)), cfg, api)
 	errs := ruleImporter.loadGroups(ctx, files)
 	for _, err := range errs {
 		if err != nil {

cmd/promtool/query.go

@@ -0,0 +1,251 @@
+// Copyright 2023 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package main
+
+import (
+	"context"
+	"errors"
+	"fmt"
+	"math"
+	"net/http"
+	"net/url"
+	"os"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/prometheus/client_golang/api"
+	v1 "github.com/prometheus/client_golang/api/prometheus/v1"
+	"github.com/prometheus/client_golang/prometheus/promhttp"
+
+	_ "github.com/prometheus/prometheus/plugins" // Register plugins.
+)
+
+func newAPI(url *url.URL, roundTripper http.RoundTripper, headers map[string]string) (v1.API, error) {
+	if url.Scheme == "" {
+		url.Scheme = "http"
+	}
+	config := api.Config{
+		Address:      url.String(),
+		RoundTripper: roundTripper,
+	}
+
+	if len(headers) > 0 {
+		config.RoundTripper = promhttp.RoundTripperFunc(func(req *http.Request) (*http.Response, error) {
+			for key, value := range headers {
+				req.Header.Add(key, value)
+			}
+			return roundTripper.RoundTrip(req)
+		})
+	}
+
+	// Create new client.
+	client, err := api.NewClient(config)
+	if err != nil {
+		return nil, err
+	}
+
+	api := v1.NewAPI(client)
+	return api, nil
+}
+
+// QueryInstant performs an instant query against a Prometheus server.
+func QueryInstant(url *url.URL, roundTripper http.RoundTripper, query, evalTime string, p printer) int {
+	api, err := newAPI(url, roundTripper, nil)
+	if err != nil {
+		fmt.Fprintln(os.Stderr, "error creating API client:", err)
+		return failureExitCode
+	}
+
+	eTime := time.Now()
+	if evalTime != "" {
+		eTime, err = parseTime(evalTime)
+		if err != nil {
+			fmt.Fprintln(os.Stderr, "error parsing evaluation time:", err)
+			return failureExitCode
+		}
+	}
+
+	// Run query against client.
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
+	val, _, err := api.Query(ctx, query, eTime) // Ignoring warnings for now.
+	cancel()
+	if err != nil {
+		return handleAPIError(err)
+	}
+
+	p.printValue(val)
+
+	return successExitCode
+}
+
+// QueryRange performs a range query against a Prometheus server.
+func QueryRange(url *url.URL, roundTripper http.RoundTripper, headers map[string]string, query, start, end string, step time.Duration, p printer) int {
+	api, err := newAPI(url, roundTripper, headers)
+	if err != nil {
+		fmt.Fprintln(os.Stderr, "error creating API client:", err)
+		return failureExitCode
+	}
+
+	var stime, etime time.Time
+
+	if end == "" {
+		etime = time.Now()
+	} else {
+		etime, err = parseTime(end)
+		if err != nil {
+			fmt.Fprintln(os.Stderr, "error parsing end time:", err)
+			return failureExitCode
+		}
+	}
+
+	if start == "" {
+		stime = etime.Add(-5 * time.Minute)
+	} else {
+		stime, err = parseTime(start)
+		if err != nil {
+			fmt.Fprintln(os.Stderr, "error parsing start time:", err)
+			return failureExitCode
+		}
+	}
+
+	if !stime.Before(etime) {
+		fmt.Fprintln(os.Stderr, "start time is not before end time")
+		return failureExitCode
+	}
+
+	if step == 0 {
+		resolution := math.Max(math.Floor(etime.Sub(stime).Seconds()/250), 1)
+		// Convert seconds to nanoseconds such that time.Duration parses correctly.
+		step = time.Duration(resolution) * time.Second
+	}
+
+	// Run query against client.
+	r := v1.Range{Start: stime, End: etime, Step: step}
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
+	val, _, err := api.QueryRange(ctx, query, r) // Ignoring warnings for now.
+	cancel()
+
+	if err != nil {
+		return handleAPIError(err)
+	}
+
+	p.printValue(val)
+	return successExitCode
+}
+
+// QuerySeries queries for a series against a Prometheus server.
+func QuerySeries(url *url.URL, roundTripper http.RoundTripper, matchers []string, start, end string, p printer) int {
+	api, err := newAPI(url, roundTripper, nil)
+	if err != nil {
+		fmt.Fprintln(os.Stderr, "error creating API client:", err)
+		return failureExitCode
+	}
+
+	stime, etime, err := parseStartTimeAndEndTime(start, end)
+	if err != nil {
+		fmt.Fprintln(os.Stderr, err)
+		return failureExitCode
+	}
+
+	// Run query against client.
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
+	val, _, err := api.Series(ctx, matchers, stime, etime) // Ignoring warnings for now.
+	cancel()
+
+	if err != nil {
+		return handleAPIError(err)
+	}
+
+	p.printSeries(val)
+	return successExitCode
+}
+
+// QueryLabels queries for label values against a Prometheus server.
+func QueryLabels(url *url.URL, roundTripper http.RoundTripper, matchers []string, name, start, end string, p printer) int {
+	api, err := newAPI(url, roundTripper, nil)
+	if err != nil {
+		fmt.Fprintln(os.Stderr, "error creating API client:", err)
+		return failureExitCode
+	}
+
+	stime, etime, err := parseStartTimeAndEndTime(start, end)
+	if err != nil {
+		fmt.Fprintln(os.Stderr, err)
+		return failureExitCode
+	}
+
+	// Run query against client.
+	ctx, cancel := context.WithTimeout(context.Background(), 2*time.Minute)
+	val, warn, err := api.LabelValues(ctx, name, matchers, stime, etime)
+	cancel()
+
+	for _, v := range warn {
+		fmt.Fprintln(os.Stderr, "query warning:", v)
+	}
+	if err != nil {
+		return handleAPIError(err)
+	}
+
+	p.printLabelValues(val)
+	return successExitCode
+}
+
+func handleAPIError(err error) int {
+	var apiErr *v1.Error
+	if errors.As(err, &apiErr) && apiErr.Detail != "" {
+		fmt.Fprintf(os.Stderr, "query error: %v (detail: %s)\n", apiErr, strings.TrimSpace(apiErr.Detail))
+	} else {
+		fmt.Fprintln(os.Stderr, "query error:", err)
+	}
+
+	return failureExitCode
+}
+
+func parseStartTimeAndEndTime(start, end string) (time.Time, time.Time, error) {
+	var (
+		minTime = time.Now().Add(-9999 * time.Hour)
+		maxTime = time.Now().Add(9999 * time.Hour)
+		err     error
+	)
+
+	stime := minTime
+	etime := maxTime
+
+	if start != "" {
+		stime, err = parseTime(start)
+		if err != nil {
+			return stime, etime, fmt.Errorf("error parsing start time: %w", err)
+		}
+	}
+
+	if end != "" {
+		etime, err = parseTime(end)
+		if err != nil {
+			return stime, etime, fmt.Errorf("error parsing end time: %w", err)
+		}
+	}
+	return stime, etime, nil
+}
+
+func parseTime(s string) (time.Time, error) {
+	if t, err := strconv.ParseFloat(s, 64); err == nil {
+		s, ns := math.Modf(t)
+		return time.Unix(int64(s), int64(ns*float64(time.Second))).UTC(), nil
+	}
+	if t, err := time.Parse(time.RFC3339Nano, s); err == nil {
+		return t, nil
+	}
+	return time.Time{}, fmt.Errorf("cannot parse %q to a valid timestamp", s)
+}

cmd/promtool/sd.go

@@ -78,12 +78,25 @@ func CheckSD(sdConfigFiles, sdJobName string, sdTimeout time.Duration, noDefault
 	defer cancel()
 
 	for _, cfg := range scrapeConfig.ServiceDiscoveryConfigs {
-		d, err := cfg.NewDiscoverer(discovery.DiscovererOptions{Logger: logger, Registerer: registerer})
+		reg := prometheus.NewRegistry()
+		refreshMetrics := discovery.NewRefreshMetrics(reg)
+		metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+		err := metrics.Register()
+		if err != nil {
+			fmt.Fprintln(os.Stderr, "Could not register service discovery metrics", err)
+			return failureExitCode
+		}
+
+		d, err := cfg.NewDiscoverer(discovery.DiscovererOptions{Logger: logger, Metrics: metrics})
 		if err != nil {
 			fmt.Fprintln(os.Stderr, "Could not create new discoverer", err)
 			return failureExitCode
 		}
-		go d.Run(ctx, targetGroupChan)
+		go func() {
+			d.Run(ctx, targetGroupChan)
+			metrics.Unregister()
+			refreshMetrics.Unregister()
+		}()
 	}
 
 	var targetGroups []*targetgroup.Group

cmd/promtool/testdata/dump-test-1.prom

@@ -0,0 +1,15 @@
+{__name__="heavy_metric", foo="bar"} 5 0
+{__name__="heavy_metric", foo="bar"} 4 60000
+{__name__="heavy_metric", foo="bar"} 3 120000
+{__name__="heavy_metric", foo="bar"} 2 180000
+{__name__="heavy_metric", foo="bar"} 1 240000
+{__name__="heavy_metric", foo="foo"} 5 0
+{__name__="heavy_metric", foo="foo"} 4 60000
+{__name__="heavy_metric", foo="foo"} 3 120000
+{__name__="heavy_metric", foo="foo"} 2 180000
+{__name__="heavy_metric", foo="foo"} 1 240000
+{__name__="metric", baz="abc", foo="bar"} 1 0
+{__name__="metric", baz="abc", foo="bar"} 2 60000
+{__name__="metric", baz="abc", foo="bar"} 3 120000
+{__name__="metric", baz="abc", foo="bar"} 4 180000
+{__name__="metric", baz="abc", foo="bar"} 5 240000

cmd/promtool/testdata/dump-test-2.prom

@@ -0,0 +1,10 @@
+{__name__="heavy_metric", foo="foo"} 5 0
+{__name__="heavy_metric", foo="foo"} 4 60000
+{__name__="heavy_metric", foo="foo"} 3 120000
+{__name__="heavy_metric", foo="foo"} 2 180000
+{__name__="heavy_metric", foo="foo"} 1 240000
+{__name__="metric", baz="abc", foo="bar"} 1 0
+{__name__="metric", baz="abc", foo="bar"} 2 60000
+{__name__="metric", baz="abc", foo="bar"} 3 120000
+{__name__="metric", baz="abc", foo="bar"} 4 180000
+{__name__="metric", baz="abc", foo="bar"} 5 240000

cmd/promtool/testdata/dump-test-3.prom

@@ -0,0 +1,2 @@
+{__name__="metric", baz="abc", foo="bar"} 2 60000
+{__name__="metric", baz="abc", foo="bar"} 3 120000

cmd/promtool/tsdb.go

@@ -667,7 +667,7 @@ func analyzeCompaction(ctx context.Context, block tsdb.BlockReader, indexr tsdb.
 				it := fhchk.Iterator(nil)
 				bucketCount := 0
 				for it.Next() == chunkenc.ValFloatHistogram {
-					_, f := it.AtFloatHistogram()
+					_, f := it.AtFloatHistogram(nil)
 					bucketCount += len(f.PositiveBuckets)
 					bucketCount += len(f.NegativeBuckets)
 				}
@@ -682,7 +682,7 @@ func analyzeCompaction(ctx context.Context, block tsdb.BlockReader, indexr tsdb.
 				it := hchk.Iterator(nil)
 				bucketCount := 0
 				for it.Next() == chunkenc.ValHistogram {
-					_, f := it.AtHistogram()
+					_, f := it.AtHistogram(nil)
 					bucketCount += len(f.PositiveBuckets)
 					bucketCount += len(f.NegativeBuckets)
 				}
@@ -706,7 +706,7 @@ func analyzeCompaction(ctx context.Context, block tsdb.BlockReader, indexr tsdb.
 	return nil
 }
 
-func dumpSamples(ctx context.Context, path string, mint, maxt int64, match string) (err error) {
+func dumpSamples(ctx context.Context, path string, mint, maxt int64, match []string) (err error) {
 	db, err := tsdb.OpenDBReadOnly(path, nil)
 	if err != nil {
 		return err
@@ -720,11 +720,21 @@ func dumpSamples(ctx context.Context, path string, mint, maxt int64, match strin
 	}
 	defer q.Close()
 
-	matchers, err := parser.ParseMetricSelector(match)
+	matcherSets, err := parser.ParseMetricSelectors(match)
 	if err != nil {
 		return err
 	}
-	ss := q.Select(ctx, false, nil, matchers...)
+
+	var ss storage.SeriesSet
+	if len(matcherSets) > 1 {
+		var sets []storage.SeriesSet
+		for _, mset := range matcherSets {
+			sets = append(sets, q.Select(ctx, true, nil, mset...))
+		}
+		ss = storage.NewMergeSeriesSet(sets, storage.ChainedSeriesMerge)
+	} else {
+		ss = q.Select(ctx, false, nil, matcherSets[0]...)
+	}
 
 	for ss.Next() {
 		series := ss.At()
@@ -735,11 +745,11 @@ func dumpSamples(ctx context.Context, path string, mint, maxt int64, match strin
 			fmt.Printf("%s %g %d\n", lbs, val, ts)
 		}
 		for it.Next() == chunkenc.ValFloatHistogram {
-			ts, fh := it.AtFloatHistogram()
+			ts, fh := it.AtFloatHistogram(nil)
 			fmt.Printf("%s %s %d\n", lbs, fh.String(), ts)
 		}
 		for it.Next() == chunkenc.ValHistogram {
-			ts, h := it.AtHistogram()
+			ts, h := it.AtHistogram(nil)
 			fmt.Printf("%s %s %d\n", lbs, h.String(), ts)
 		}
 		if it.Err() != nil {

cmd/promtool/tsdb_test.go

@@ -14,9 +14,18 @@
 package main
 
 import (
+	"bytes"
+	"context"
+	"io"
+	"math"
+	"os"
+	"runtime"
+	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/require"
+
+	"github.com/prometheus/prometheus/promql"
 )
 
 func TestGenerateBucket(t *testing.T) {
@@ -41,3 +50,101 @@ func TestGenerateBucket(t *testing.T) {
 		require.Equal(t, tc.step, step)
 	}
 }
+
+// getDumpedSamples dumps samples and returns them.
+func getDumpedSamples(t *testing.T, path string, mint, maxt int64, match []string) string {
+	t.Helper()
+
+	oldStdout := os.Stdout
+	r, w, _ := os.Pipe()
+	os.Stdout = w
+
+	err := dumpSamples(
+		context.Background(),
+		path,
+		mint,
+		maxt,
+		match,
+	)
+	require.NoError(t, err)
+
+	w.Close()
+	os.Stdout = oldStdout
+
+	var buf bytes.Buffer
+	io.Copy(&buf, r)
+	return buf.String()
+}
+
+func TestTSDBDump(t *testing.T) {
+	storage := promql.LoadedStorage(t, `
+		load 1m
+			metric{foo="bar", baz="abc"} 1 2 3 4 5
+			heavy_metric{foo="bar"} 5 4 3 2 1
+			heavy_metric{foo="foo"} 5 4 3 2 1
+	`)
+
+	tests := []struct {
+		name         string
+		mint         int64
+		maxt         int64
+		match        []string
+		expectedDump string
+	}{
+		{
+			name:         "default match",
+			mint:         math.MinInt64,
+			maxt:         math.MaxInt64,
+			match:        []string{"{__name__=~'(?s:.*)'}"},
+			expectedDump: "testdata/dump-test-1.prom",
+		},
+		{
+			name:         "same matcher twice",
+			mint:         math.MinInt64,
+			maxt:         math.MaxInt64,
+			match:        []string{"{foo=~'.+'}", "{foo=~'.+'}"},
+			expectedDump: "testdata/dump-test-1.prom",
+		},
+		{
+			name:         "no duplication",
+			mint:         math.MinInt64,
+			maxt:         math.MaxInt64,
+			match:        []string{"{__name__=~'(?s:.*)'}", "{baz='abc'}"},
+			expectedDump: "testdata/dump-test-1.prom",
+		},
+		{
+			name:         "well merged",
+			mint:         math.MinInt64,
+			maxt:         math.MaxInt64,
+			match:        []string{"{__name__='heavy_metric'}", "{baz='abc'}"},
+			expectedDump: "testdata/dump-test-1.prom",
+		},
+		{
+			name:         "multi matchers",
+			mint:         math.MinInt64,
+			maxt:         math.MaxInt64,
+			match:        []string{"{__name__='heavy_metric',foo='foo'}", "{__name__='metric'}"},
+			expectedDump: "testdata/dump-test-2.prom",
+		},
+		{
+			name:         "with reduced mint and maxt",
+			mint:         int64(60000),
+			maxt:         int64(120000),
+			match:        []string{"{__name__='metric'}"},
+			expectedDump: "testdata/dump-test-3.prom",
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			dumpedMetrics := getDumpedSamples(t, storage.Dir(), tt.mint, tt.maxt, tt.match)
+			expectedMetrics, err := os.ReadFile(tt.expectedDump)
+			require.NoError(t, err)
+			if strings.Contains(runtime.GOOS, "windows") {
+				// We use "/n" while dumping on windows as well.
+				expectedMetrics = bytes.ReplaceAll(expectedMetrics, []byte("\r\n"), []byte("\n"))
+			}
+			// even though in case of one matcher samples are not sorted, the order in the cases above should stay the same.
+			require.Equal(t, string(expectedMetrics), dumpedMetrics)
+		})
+	}
+}

cmd/promtool/unittest.go

@@ -15,6 +15,7 @@ package main
 
 import (
 	"context"
+	"encoding/json"
 	"errors"
 	"fmt"
 	"os"
@@ -27,6 +28,7 @@ import (
 
 	"github.com/go-kit/log"
 	"github.com/grafana/regexp"
+	"github.com/nsf/jsondiff"
 	"github.com/prometheus/common/model"
 	"gopkg.in/yaml.v2"
 
@@ -40,7 +42,7 @@ import (
 
 // RulesUnitTest does unit testing of rules based on the unit testing files provided.
 // More info about the file format can be found in the docs.
-func RulesUnitTest(queryOpts promql.LazyLoaderOpts, runStrings []string, files ...string) int {
+func RulesUnitTest(queryOpts promql.LazyLoaderOpts, runStrings []string, diffFlag bool, files ...string) int {
 	failed := false
 
 	var run *regexp.Regexp
@@ -49,7 +51,7 @@ func RulesUnitTest(queryOpts promql.LazyLoaderOpts, runStrings []string, files .
 	}
 
 	for _, f := range files {
-		if errs := ruleUnitTest(f, queryOpts, run); errs != nil {
+		if errs := ruleUnitTest(f, queryOpts, run, diffFlag); errs != nil {
 			fmt.Fprintln(os.Stderr, "  FAILED:")
 			for _, e := range errs {
 				fmt.Fprintln(os.Stderr, e.Error())
@@ -67,7 +69,7 @@ func RulesUnitTest(queryOpts promql.LazyLoaderOpts, runStrings []string, files .
 	return successExitCode
 }
 
-func ruleUnitTest(filename string, queryOpts promql.LazyLoaderOpts, run *regexp.Regexp) []error {
+func ruleUnitTest(filename string, queryOpts promql.LazyLoaderOpts, run *regexp.Regexp, diffFlag bool) []error {
 	fmt.Println("Unit Testing: ", filename)
 
 	b, err := os.ReadFile(filename)
@@ -109,7 +111,7 @@ func ruleUnitTest(filename string, queryOpts promql.LazyLoaderOpts, run *regexp.
 		if t.Interval == 0 {
 			t.Interval = unitTestInp.EvaluationInterval
 		}
-		ers := t.test(evalInterval, groupOrderMap, queryOpts, unitTestInp.RuleFiles...)
+		ers := t.test(evalInterval, groupOrderMap, queryOpts, diffFlag, unitTestInp.RuleFiles...)
 		if ers != nil {
 			errs = append(errs, ers...)
 		}
@@ -173,7 +175,7 @@ type testGroup struct {
 }
 
 // test performs the unit tests.
-func (tg *testGroup) test(evalInterval time.Duration, groupOrderMap map[string]int, queryOpts promql.LazyLoaderOpts, ruleFiles ...string) []error {
+func (tg *testGroup) test(evalInterval time.Duration, groupOrderMap map[string]int, queryOpts promql.LazyLoaderOpts, diffFlag bool, ruleFiles ...string) []error {
 	// Setup testing suite.
 	suite, err := promql.NewLazyLoader(nil, tg.seriesLoadingString(), queryOpts)
 	if err != nil {
@@ -345,10 +347,46 @@ func (tg *testGroup) test(evalInterval time.Duration, groupOrderMap map[string]i
 					}
 					expString := indentLines(expAlerts.String(), "            ")
 					gotString := indentLines(gotAlerts.String(), "            ")
+					if diffFlag {
+						// If empty, populates an empty value
+						if gotAlerts.Len() == 0 {
+							gotAlerts = append(gotAlerts, labelAndAnnotation{
+								Labels:      labels.Labels{},
+								Annotations: labels.Labels{},
+							})
+						}
+						// If empty, populates an empty value
+						if expAlerts.Len() == 0 {
+							expAlerts = append(expAlerts, labelAndAnnotation{
+								Labels:      labels.Labels{},
+								Annotations: labels.Labels{},
+							})
+						}
+
+						diffOpts := jsondiff.DefaultConsoleOptions()
+						expAlertsJSON, err := json.Marshal(expAlerts)
+						if err != nil {
+							errs = append(errs, fmt.Errorf("error marshaling expected %s alert: [%s]", tg.TestGroupName, err.Error()))
+							continue
+						}
+
+						gotAlertsJSON, err := json.Marshal(gotAlerts)
+						if err != nil {
+							errs = append(errs, fmt.Errorf("error marshaling received %s alert: [%s]", tg.TestGroupName, err.Error()))
+							continue
+						}
+
+						res, diff := jsondiff.Compare(expAlertsJSON, gotAlertsJSON, &diffOpts)
+						if res != jsondiff.FullMatch {
+							errs = append(errs, fmt.Errorf("%s    alertname: %s, time: %s, \n        diff: %v",
+								testName, testcase.Alertname, testcase.EvalTime.String(), indentLines(diff, "            ")))
+						}
+					} else {
 						errs = append(errs, fmt.Errorf("%s    alertname: %s, time: %s, \n        exp:%v, \n        got:%v",
 							testName, testcase.Alertname, testcase.EvalTime.String(), expString, gotString))
 					}
 				}
+			}
 
 			curr++
 		}

cmd/promtool/unittest_test.go

@@ -125,7 +125,7 @@ func TestRulesUnitTest(t *testing.T) {
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			if got := RulesUnitTest(tt.queryOpts, nil, tt.args.files...); got != tt.want {
+			if got := RulesUnitTest(tt.queryOpts, nil, false, tt.args.files...); got != tt.want {
 				t.Errorf("RulesUnitTest() = %v, want %v", got, tt.want)
 			}
 		})
@@ -178,7 +178,7 @@ func TestRulesUnitTestRun(t *testing.T) {
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {
-			if got := RulesUnitTest(tt.queryOpts, tt.args.run, tt.args.files...); got != tt.want {
+			if got := RulesUnitTest(tt.queryOpts, tt.args.run, false, tt.args.files...); got != tt.want {
 				t.Errorf("RulesUnitTest() = %v, want %v", got, tt.want)
 			}
 		})

config/config.go

@@ -610,9 +610,12 @@ type ScrapeConfig struct {
 	// More than this label value length post metric-relabeling will cause the
 	// scrape to fail. 0 means no limit.
 	LabelValueLengthLimit uint `yaml:"label_value_length_limit,omitempty"`
-	// More than this many buckets in a native histogram will cause the scrape to
+	// If there are more than this many buckets in a native histogram,
-	// fail.
+	// buckets will be merged to stay within the limit.
 	NativeHistogramBucketLimit uint `yaml:"native_histogram_bucket_limit,omitempty"`
+	// If the growth factor of one bucket to the next is smaller than this,
+	// buckets will be merged to increase the factor sufficiently.
+	NativeHistogramMinBucketFactor float64 `yaml:"native_histogram_min_bucket_factor,omitempty"`
 	// Keep no more than this many dropped targets per job.
 	// 0 means no limit.
 	KeepDroppedTargets uint `yaml:"keep_dropped_targets,omitempty"`
@@ -1124,6 +1127,9 @@ type QueueConfig struct {
 	MinBackoff       model.Duration `yaml:"min_backoff,omitempty"`
 	MaxBackoff       model.Duration `yaml:"max_backoff,omitempty"`
 	RetryOnRateLimit bool           `yaml:"retry_on_http_429,omitempty"`
+
+	// Samples older than the limit will be dropped.
+	SampleAgeLimit model.Duration `yaml:"sample_age_limit,omitempty"`
 }
 
 // MetadataConfig is the configuration for sending metadata to remote

config/config_default_test.go

@@ -12,7 +12,6 @@
 // limitations under the License.
 
 //go:build !windows
-// +build !windows
 
 package config
 

consoles/node-cpu.html

@@ -47,7 +47,7 @@
   <script>
   new PromConsole.Graph({
     node: document.querySelector("#cpuGraph"),
-    expr: "sum by (mode)(irate(node_cpu_seconds_total{job='node',instance='{{ .Params.instance }}',mode!='idle'}[5m]))",
+    expr: "sum by (mode)(irate(node_cpu_seconds_total{job='node',instance='{{ .Params.instance }}',mode!='idle',mode!='iowait',mode!='steal'}[5m]))",
     renderer: 'area',
     max: {{ with printf "count(count by (cpu)(node_cpu_seconds_total{job='node',instance='%s'}))" .Params.instance | query }}{{ . | first | value }}{{ else}}undefined{{end}},
     yAxisFormatter: PromConsole.NumberFormatter.humanizeNoSmallPrefix,

consoles/node-overview.html

@@ -69,7 +69,7 @@
   <script>
   new PromConsole.Graph({
     node: document.querySelector("#cpuGraph"),
-    expr: "sum by (mode)(irate(node_cpu_seconds_total{job='node',instance='{{ .Params.instance }}',mode!='idle'}[5m]))",
+    expr: "sum by (mode)(irate(node_cpu_seconds_total{job='node',instance='{{ .Params.instance }}',mode!='idle',mode!='iowait',mode!='steal'}[5m]))",
     renderer: 'area',
     max: {{ with printf "count(count by (cpu)(node_cpu_seconds_total{job='node',instance='%s'}))" .Params.instance | query }}{{ . | first | value }}{{ else}}undefined{{end}},
     yAxisFormatter: PromConsole.NumberFormatter.humanizeNoSmallPrefix,

consoles/node.html

@@ -21,7 +21,7 @@
 <tr>
   <td><a href="node-overview.html?instance={{ .Labels.instance }}">{{ reReplaceAll "(.*?://)([^:/]+?)(:\\d+)?/.*" "$2" .Labels.instance }}</a></td>
   <td{{ if eq (. | value) 1.0 }}>Yes{{ else }} class="alert-danger">No{{ end }}</td>
-  <td>{{ template "prom_query_drilldown" (args (printf "100 * (1 - avg by(instance)(irate(node_cpu_seconds_total{job='node',mode='idle',instance='%s'}[5m])))" .Labels.instance) "%" "printf.1f") }}</td>
+  <td>{{ template "prom_query_drilldown" (args (printf "100 * (1 - avg by(instance) (sum without(mode) (irate(node_cpu_seconds_total{job='node',mode=~'idle|iowait|steal',instance='%s'}[5m]))))" .Labels.instance) "%" "printf.1f") }}</td>
   <td>{{ template "prom_query_drilldown" (args (printf "node_memory_MemFree_bytes{job='node',instance='%s'} + node_memory_Cached_bytes{job='node',instance='%s'} + node_memory_Buffers_bytes{job='node',instance='%s'}" .Labels.instance .Labels.instance .Labels.instance) "B" "humanize1024") }}</td>
 </tr>
 {{ else }}

discovery/aws/ec2.go

@@ -97,12 +97,19 @@ type EC2SDConfig struct {
 	HTTPClientConfig config.HTTPClientConfig `yaml:",inline"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*EC2SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &ec2Metrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the EC2 Config.
 func (*EC2SDConfig) Name() string { return "ec2" }
 
 // NewDiscoverer returns a Discoverer for the EC2 Config.
 func (c *EC2SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewEC2Discovery(c, opts.Logger, opts.Registerer), nil
+	return NewEC2Discovery(c, opts.Logger, opts.Metrics)
 }
 
 // UnmarshalYAML implements the yaml.Unmarshaler interface for the EC2 Config.
@@ -148,7 +155,12 @@ type EC2Discovery struct {
 }
 
 // NewEC2Discovery returns a new EC2Discovery which periodically refreshes its targets.
-func NewEC2Discovery(conf *EC2SDConfig, logger log.Logger, reg prometheus.Registerer) *EC2Discovery {
+func NewEC2Discovery(conf *EC2SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*EC2Discovery, error) {
+	m, ok := metrics.(*ec2Metrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -162,10 +174,10 @@ func NewEC2Discovery(conf *EC2SDConfig, logger log.Logger, reg prometheus.Regist
 			Mech:                "ec2",
 			Interval:            time.Duration(d.cfg.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
-	return d
+	return d, nil
 }
 
 func (d *EC2Discovery) ec2Client(context.Context) (*ec2.EC2, error) {

discovery/aws/lightsail.go

@@ -80,12 +80,19 @@ type LightsailSDConfig struct {
 	HTTPClientConfig config.HTTPClientConfig `yaml:",inline"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*LightsailSDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &lightsailMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Lightsail Config.
 func (*LightsailSDConfig) Name() string { return "lightsail" }
 
 // NewDiscoverer returns a Discoverer for the Lightsail Config.
 func (c *LightsailSDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewLightsailDiscovery(c, opts.Logger, opts.Registerer), nil
+	return NewLightsailDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // UnmarshalYAML implements the yaml.Unmarshaler interface for the Lightsail Config.
@@ -122,10 +129,16 @@ type LightsailDiscovery struct {
 }
 
 // NewLightsailDiscovery returns a new LightsailDiscovery which periodically refreshes its targets.
-func NewLightsailDiscovery(conf *LightsailSDConfig, logger log.Logger, reg prometheus.Registerer) *LightsailDiscovery {
+func NewLightsailDiscovery(conf *LightsailSDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*LightsailDiscovery, error) {
+	m, ok := metrics.(*lightsailMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
+
 	d := &LightsailDiscovery{
 		cfg: conf,
 	}
@@ -135,10 +148,10 @@ func NewLightsailDiscovery(conf *LightsailSDConfig, logger log.Logger, reg prome
 			Mech:                "lightsail",
 			Interval:            time.Duration(d.cfg.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
-	return d
+	return d, nil
 }
 
 func (d *LightsailDiscovery) lightsailClient() (*lightsail.Lightsail, error) {

discovery/aws/metrics_ec2.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package aws
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+type ec2Metrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+var _ discovery.DiscovererMetrics = (*ec2Metrics)(nil)
+
+// Register implements discovery.DiscovererMetrics.
+func (m *ec2Metrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *ec2Metrics) Unregister() {}

discovery/aws/metrics_lightsail.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package aws
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+type lightsailMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+var _ discovery.DiscovererMetrics = (*lightsailMetrics)(nil)
+
+// Register implements discovery.DiscovererMetrics.
+func (m *lightsailMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *lightsailMetrics) Unregister() {}

discovery/azure/azure.go

@@ -120,12 +120,17 @@ type SDConfig struct {
 	HTTPClientConfig config_util.HTTPClientConfig `yaml:",inline"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "azure" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 func validateAuthParam(param, name string) error {
@@ -172,12 +177,16 @@ type Discovery struct {
 	cfg     *SDConfig
 	port    int
 	cache   *cache.Cache[string, *armnetwork.Interface]
-	failuresCount prometheus.Counter
+	metrics *azureMetrics
-	cacheHitCount prometheus.Counter
 }
 
 // NewDiscovery returns a new AzureDiscovery which periodically refreshes its targets.
-func NewDiscovery(cfg *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(cfg *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*azureMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -187,16 +196,7 @@ func NewDiscovery(cfg *SDConfig, logger log.Logger, reg prometheus.Registerer) (
 		port:    cfg.Port,
 		logger:  logger,
 		cache:   l,
-		failuresCount: prometheus.NewCounter(
+		metrics: m,
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_azure_failures_total",
-				Help: "Number of Azure service discovery refresh failures.",
-			}),
-		cacheHitCount: prometheus.NewCounter(
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_azure_cache_hit_total",
-				Help: "Number of cache hit during refresh.",
-			}),
 	}
 
 	d.Discovery = refresh.NewDiscovery(
@@ -205,8 +205,7 @@ func NewDiscovery(cfg *SDConfig, logger log.Logger, reg prometheus.Registerer) (
 			Mech:                "azure",
 			Interval:            time.Duration(cfg.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
-			Metrics:  []prometheus.Collector{d.failuresCount, d.cacheHitCount},
 		},
 	)
 
@@ -333,14 +332,14 @@ func (d *Discovery) refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 
 	client, err := createAzureClient(*d.cfg)
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, fmt.Errorf("could not create Azure client: %w", err)
 	}
 	client.logger = d.logger
 
 	machines, err := client.getVMs(ctx, d.cfg.ResourceGroup)
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, fmt.Errorf("could not get virtual machines: %w", err)
 	}
 
@@ -349,14 +348,14 @@ func (d *Discovery) refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 	// Load the vms managed by scale sets.
 	scaleSets, err := client.getScaleSets(ctx, d.cfg.ResourceGroup)
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, fmt.Errorf("could not get virtual machine scale sets: %w", err)
 	}
 
 	for _, scaleSet := range scaleSets {
 		scaleSetVms, err := client.getScaleSetVMs(ctx, scaleSet)
 		if err != nil {
-			d.failuresCount.Inc()
+			d.metrics.failuresCount.Inc()
 			return nil, fmt.Errorf("could not get virtual machine scale set vms: %w", err)
 		}
 		machines = append(machines, scaleSetVms...)
@@ -407,13 +406,22 @@ func (d *Discovery) refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 				var networkInterface *armnetwork.Interface
 				if v, ok := d.getFromCache(nicID); ok {
 					networkInterface = v
-					d.cacheHitCount.Add(1)
+					d.metrics.cacheHitCount.Add(1)
 				} else {
 					if vm.ScaleSet == "" {
 						networkInterface, err = client.getVMNetworkInterfaceByID(ctx, nicID)
 					} else {
 						networkInterface, err = client.getVMScaleSetVMNetworkInterfaceByID(ctx, nicID, vm.ScaleSet, vm.InstanceID)
 					}
+					if err != nil {
+						if errors.Is(err, errorNotFound) {
+							level.Warn(d.logger).Log("msg", "Network interface does not exist", "name", nicID, "err", err)
+						} else {
+							ch <- target{labelSet: nil, err: err}
+						}
+						d.addToCache(nicID, networkInterface)
+					} else {
+						networkInterface, err = client.getVMScaleSetVMNetworkInterfaceByID(ctx, nicID, vm.ScaleSet, vm.InstanceID)
 						if err != nil {
 							if errors.Is(err, errorNotFound) {
 								level.Warn(d.logger).Log("msg", "Network interface does not exist", "name", nicID, "err", err)
@@ -425,6 +433,7 @@ func (d *Discovery) refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 						}
 						d.addToCache(nicID, networkInterface)
 					}
+				}
 
 				if networkInterface.Properties == nil {
 					continue
@@ -470,7 +479,7 @@ func (d *Discovery) refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 	var tg targetgroup.Group
 	for tgt := range ch {
 		if tgt.err != nil {
-			d.failuresCount.Inc()
+			d.metrics.failuresCount.Inc()
 			return nil, fmt.Errorf("unable to complete Azure service discovery: %w", tgt.err)
 		}
 		if tgt.labelSet != nil {

discovery/azure/metrics.go

@@ -0,0 +1,64 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package azure
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*azureMetrics)(nil)
+
+type azureMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+
+	failuresCount prometheus.Counter
+	cacheHitCount prometheus.Counter
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &azureMetrics{
+		refreshMetrics: rmi,
+		failuresCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_azure_failures_total",
+				Help: "Number of Azure service discovery refresh failures.",
+			}),
+		cacheHitCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_azure_cache_hit_total",
+				Help: "Number of cache hit during refresh.",
+			}),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.failuresCount,
+		m.cacheHitCount,
+	})
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *azureMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *azureMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/consul/consul.go

@@ -119,12 +119,17 @@ type SDConfig struct {
 	HTTPClientConfig config.HTTPClientConfig `yaml:",inline"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "consul" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -173,15 +178,16 @@ type Discovery struct {
 	refreshInterval  time.Duration
 	finalizer        func()
 	logger           log.Logger
-	rpcFailuresCount    prometheus.Counter
+	metrics          *consulMetrics
-	rpcDuration         *prometheus.SummaryVec
-	servicesRPCDuration prometheus.Observer
-	serviceRPCDuration  prometheus.Observer
-	metricRegisterer    discovery.MetricRegisterer
 }
 
 // NewDiscovery returns a new Discovery for the given config.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*consulMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -219,35 +225,9 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 		clientPartition:  conf.Partition,
 		finalizer:        wrapper.CloseIdleConnections,
 		logger:           logger,
-		rpcFailuresCount: prometheus.NewCounter(
+		metrics:          m,
-			prometheus.CounterOpts{
-				Namespace: namespace,
-				Name:      "sd_consul_rpc_failures_total",
-				Help:      "The number of Consul RPC call failures.",
-			}),
-		rpcDuration: prometheus.NewSummaryVec(
-			prometheus.SummaryOpts{
-				Namespace:  namespace,
-				Name:       "sd_consul_rpc_duration_seconds",
-				Help:       "The duration of a Consul RPC call in seconds.",
-				Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
-			},
-			[]string{"endpoint", "call"},
-		),
 	}
 
-	cd.metricRegisterer = discovery.NewMetricRegisterer(
-		reg,
-		[]prometheus.Collector{
-			cd.rpcFailuresCount,
-			cd.rpcDuration,
-		},
-	)
-
-	// Initialize metric vectors.
-	cd.servicesRPCDuration = cd.rpcDuration.WithLabelValues("catalog", "services")
-	cd.serviceRPCDuration = cd.rpcDuration.WithLabelValues("catalog", "service")
-
 	return cd, nil
 }
 
@@ -303,7 +283,7 @@ func (d *Discovery) getDatacenter() error {
 	info, err := d.client.Agent().Self()
 	if err != nil {
 		level.Error(d.logger).Log("msg", "Error retrieving datacenter name", "err", err)
-		d.rpcFailuresCount.Inc()
+		d.metrics.rpcFailuresCount.Inc()
 		return err
 	}
 
@@ -344,13 +324,6 @@ func (d *Discovery) initialize(ctx context.Context) {
 
 // Run implements the Discoverer interface.
 func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
-	err := d.metricRegisterer.RegisterMetrics()
-	if err != nil {
-		level.Error(d.logger).Log("msg", "Unable to register metrics", "err", err.Error())
-		return
-	}
-	defer d.metricRegisterer.UnregisterMetrics()
-
 	if d.finalizer != nil {
 		defer d.finalizer()
 	}
@@ -399,7 +372,7 @@ func (d *Discovery) watchServices(ctx context.Context, ch chan<- []*targetgroup.
 	t0 := time.Now()
 	srvs, meta, err := catalog.Services(opts.WithContext(ctx))
 	elapsed := time.Since(t0)
-	d.servicesRPCDuration.Observe(elapsed.Seconds())
+	d.metrics.servicesRPCDuration.Observe(elapsed.Seconds())
 
 	// Check the context before in order to exit early.
 	select {
@@ -410,7 +383,7 @@ func (d *Discovery) watchServices(ctx context.Context, ch chan<- []*targetgroup.
 
 	if err != nil {
 		level.Error(d.logger).Log("msg", "Error refreshing service list", "err", err)
-		d.rpcFailuresCount.Inc()
+		d.metrics.rpcFailuresCount.Inc()
 		time.Sleep(retryInterval)
 		return
 	}
@@ -490,8 +463,8 @@ func (d *Discovery) watchService(ctx context.Context, ch chan<- []*targetgroup.G
 		},
 		tagSeparator:       d.tagSeparator,
 		logger:             d.logger,
-		rpcFailuresCount:   d.rpcFailuresCount,
+		rpcFailuresCount:   d.metrics.rpcFailuresCount,
-		serviceRPCDuration: d.serviceRPCDuration,
+		serviceRPCDuration: d.metrics.serviceRPCDuration,
 	}
 
 	go func() {

discovery/consul/consul_test.go

@@ -29,6 +29,7 @@ import (
 	"go.uber.org/goleak"
 	"gopkg.in/yaml.v2"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -36,11 +37,25 @@ func TestMain(m *testing.M) {
 	goleak.VerifyTestMain(m)
 }
 
+// TODO: Add ability to unregister metrics?
+func NewTestMetrics(t *testing.T, conf discovery.Config, reg prometheus.Registerer) discovery.DiscovererMetrics {
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	require.NoError(t, refreshMetrics.Register())
+
+	metrics := conf.NewDiscovererMetrics(prometheus.NewRegistry(), refreshMetrics)
+	require.NoError(t, metrics.Register())
+
+	return metrics
+}
+
 func TestConfiguredService(t *testing.T) {
 	conf := &SDConfig{
 		Services: []string{"configuredServiceName"},
 	}
-	consulDiscovery, err := NewDiscovery(conf, nil, prometheus.NewRegistry())
+
+	metrics := NewTestMetrics(t, conf, prometheus.NewRegistry())
+
+	consulDiscovery, err := NewDiscovery(conf, nil, metrics)
 	if err != nil {
 		t.Errorf("Unexpected error when initializing discovery %v", err)
 	}
@@ -57,7 +72,10 @@ func TestConfiguredServiceWithTag(t *testing.T) {
 		Services:    []string{"configuredServiceName"},
 		ServiceTags: []string{"http"},
 	}
-	consulDiscovery, err := NewDiscovery(conf, nil, prometheus.NewRegistry())
+
+	metrics := NewTestMetrics(t, conf, prometheus.NewRegistry())
+
+	consulDiscovery, err := NewDiscovery(conf, nil, metrics)
 	if err != nil {
 		t.Errorf("Unexpected error when initializing discovery %v", err)
 	}
@@ -152,7 +170,9 @@ func TestConfiguredServiceWithTags(t *testing.T) {
 	}
 
 	for _, tc := range cases {
-		consulDiscovery, err := NewDiscovery(tc.conf, nil, prometheus.NewRegistry())
+		metrics := NewTestMetrics(t, tc.conf, prometheus.NewRegistry())
+
+		consulDiscovery, err := NewDiscovery(tc.conf, nil, metrics)
 		if err != nil {
 			t.Errorf("Unexpected error when initializing discovery %v", err)
 		}
@@ -160,13 +180,15 @@ func TestConfiguredServiceWithTags(t *testing.T) {
 		if ret != tc.shouldWatch {
 			t.Errorf("Expected should watch? %t, got %t. Watched service and tags: %s %+v, input was %s %+v", tc.shouldWatch, ret, tc.conf.Services, tc.conf.ServiceTags, tc.serviceName, tc.serviceTags)
 		}
-
 	}
 }
 
 func TestNonConfiguredService(t *testing.T) {
 	conf := &SDConfig{}
-	consulDiscovery, err := NewDiscovery(conf, nil, prometheus.NewRegistry())
+
+	metrics := NewTestMetrics(t, conf, prometheus.NewRegistry())
+
+	consulDiscovery, err := NewDiscovery(conf, nil, metrics)
 	if err != nil {
 		t.Errorf("Unexpected error when initializing discovery %v", err)
 	}
@@ -263,7 +285,10 @@ func newServer(t *testing.T) (*httptest.Server, *SDConfig) {
 
 func newDiscovery(t *testing.T, config *SDConfig) *Discovery {
 	logger := log.NewNopLogger()
-	d, err := NewDiscovery(config, logger, prometheus.NewRegistry())
+
+	metrics := NewTestMetrics(t, config, prometheus.NewRegistry())
+
+	d, err := NewDiscovery(config, logger, metrics)
 	require.NoError(t, err)
 	return d
 }

discovery/consul/metrics.go

@@ -0,0 +1,73 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package consul
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*consulMetrics)(nil)
+
+type consulMetrics struct {
+	rpcFailuresCount prometheus.Counter
+	rpcDuration      *prometheus.SummaryVec
+
+	servicesRPCDuration prometheus.Observer
+	serviceRPCDuration  prometheus.Observer
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &consulMetrics{
+		rpcFailuresCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Namespace: namespace,
+				Name:      "sd_consul_rpc_failures_total",
+				Help:      "The number of Consul RPC call failures.",
+			}),
+		rpcDuration: prometheus.NewSummaryVec(
+			prometheus.SummaryOpts{
+				Namespace:  namespace,
+				Name:       "sd_consul_rpc_duration_seconds",
+				Help:       "The duration of a Consul RPC call in seconds.",
+				Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
+			},
+			[]string{"endpoint", "call"},
+		),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.rpcFailuresCount,
+		m.rpcDuration,
+	})
+
+	// Initialize metric vectors.
+	m.servicesRPCDuration = m.rpcDuration.WithLabelValues("catalog", "services")
+	m.serviceRPCDuration = m.rpcDuration.WithLabelValues("catalog", "service")
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *consulMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *consulMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/digitalocean/digitalocean.go

@@ -63,6 +63,13 @@ func init() {
 	discovery.RegisterConfig(&SDConfig{})
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &digitaloceanMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // SDConfig is the configuration for DigitalOcean based service discovery.
 type SDConfig struct {
 	HTTPClientConfig config.HTTPClientConfig `yaml:",inline"`
@@ -76,7 +83,7 @@ func (*SDConfig) Name() string { return "digitalocean" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -104,7 +111,12 @@ type Discovery struct {
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*digitaloceanMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	d := &Discovery{
 		port: conf.Port,
 	}
@@ -131,7 +143,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "digitalocean",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil

discovery/digitalocean/digitalocean_test.go

@@ -23,6 +23,8 @@ import (
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 type DigitalOceanSDTestSuite struct {
@@ -47,7 +49,15 @@ func TestDigitalOceanSDRefresh(t *testing.T) {
 
 	cfg := DefaultSDConfig
 	cfg.HTTPClientConfig.BearerToken = tokenID
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 	endpoint, err := url.Parse(sdmock.Mock.Endpoint())
 	require.NoError(t, err)

discovery/digitalocean/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package digitalocean
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*digitaloceanMetrics)(nil)
+
+type digitaloceanMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *digitaloceanMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *digitaloceanMetrics) Unregister() {}

discovery/discoverer_metrics_noop.go

@@ -0,0 +1,28 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package discovery
+
+// Create a dummy metrics struct, because this SD doesn't have any metrics.
+type NoopDiscovererMetrics struct{}
+
+var _ DiscovererMetrics = (*NoopDiscovererMetrics)(nil)
+
+// Register implements discovery.DiscovererMetrics.
+func (*NoopDiscovererMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (*NoopDiscovererMetrics) Unregister() {
+}

discovery/discovery.go

@@ -39,24 +39,47 @@ type Discoverer interface {
 	Run(ctx context.Context, up chan<- []*targetgroup.Group)
 }
 
+// Internal metrics of service discovery mechanisms.
+type DiscovererMetrics interface {
+	Register() error
+	Unregister()
+}
+
 // DiscovererOptions provides options for a Discoverer.
 type DiscovererOptions struct {
 	Logger log.Logger
 
-	// A registerer for the Discoverer's metrics.
+	Metrics DiscovererMetrics
-	// Some Discoverers may ignore this registerer and use the global one instead.
-	// For now this will work, because the Prometheus `main` function uses the global registry.
-	// However, in the future the Prometheus `main` function will be updated to not use the global registry.
-	// Hence, if a discoverer wants its metrics to be visible via the Prometheus executable's
-	// `/metrics` endpoint, it should use this explicit registerer.
-	// TODO(ptodev): Update this comment once the Prometheus `main` function does not use the global registry.
-	Registerer prometheus.Registerer
 
 	// Extra HTTP client options to expose to Discoverers. This field may be
 	// ignored; Discoverer implementations must opt-in to reading it.
 	HTTPClientOptions []config.HTTPClientOption
 }
 
+// Metrics used by the "refresh" package.
+// We define them here in the "discovery" package in order to avoid a cyclic dependency between
+// "discovery" and "refresh".
+type RefreshMetrics struct {
+	Failures prometheus.Counter
+	Duration prometheus.Observer
+}
+
+// Instantiate the metrics used by the "refresh" package.
+type RefreshMetricsInstantiator interface {
+	Instantiate(mech string) *RefreshMetrics
+}
+
+// An interface for registering, unregistering, and instantiating metrics for the "refresh" package.
+// Refresh metrics are registered and unregistered outside of the service discovery mechanism.
+// This is so that the same metrics can be reused across different service discovery mechanisms.
+// To manage refresh metrics inside the SD mechanism, we'd need to use const labels which are
+// specific to that SD. However, doing so would also expose too many unused metrics on
+// the Prometheus /metrics endpoint.
+type RefreshMetricsManager interface {
+	DiscovererMetrics
+	RefreshMetricsInstantiator
+}
+
 // A Config provides the configuration and constructor for a Discoverer.
 type Config interface {
 	// Name returns the name of the discovery mechanism.
@@ -65,6 +88,9 @@ type Config interface {
 	// NewDiscoverer returns a Discoverer for the Config
 	// with the given DiscovererOptions.
 	NewDiscoverer(DiscovererOptions) (Discoverer, error)
+
+	// NewDiscovererMetrics returns the metrics used by the service discovery.
+	NewDiscovererMetrics(prometheus.Registerer, RefreshMetricsInstantiator) DiscovererMetrics
 }
 
 // Configs is a slice of Config values that uses custom YAML marshaling and unmarshaling
@@ -119,6 +145,11 @@ func (c StaticConfig) NewDiscoverer(DiscovererOptions) (Discoverer, error) {
 	return staticDiscoverer(c), nil
 }
 
+// No metrics are needed for this service discovery mechanism.
+func (c StaticConfig) NewDiscovererMetrics(prometheus.Registerer, RefreshMetricsInstantiator) DiscovererMetrics {
+	return &NoopDiscovererMetrics{}
+}
+
 type staticDiscoverer []*targetgroup.Group
 
 func (c staticDiscoverer) Run(ctx context.Context, up chan<- []*targetgroup.Group) {

discovery/dns/dns.go

@@ -67,12 +67,17 @@ type SDConfig struct {
 	Port            int            `yaml:"port"` // Ignored for SRV records
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "dns" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(*c, opts.Logger, opts.Registerer)
+	return NewDiscovery(*c, opts.Logger, opts.Metrics)
 }
 
 // UnmarshalYAML implements the yaml.Unmarshaler interface.
@@ -106,14 +111,18 @@ type Discovery struct {
 	port    int
 	qtype   uint16
 	logger  log.Logger
-	dnsSDLookupsCount        prometheus.Counter
+	metrics *dnsMetrics
-	dnsSDLookupFailuresCount prometheus.Counter
 
 	lookupFn func(name string, qtype uint16, logger log.Logger) (*dns.Msg, error)
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*dnsMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -137,18 +146,7 @@ func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (
 		port:     conf.Port,
 		logger:   logger,
 		lookupFn: lookupWithSearchPath,
-		dnsSDLookupsCount: prometheus.NewCounter(
+		metrics:  m,
-			prometheus.CounterOpts{
-				Namespace: namespace,
-				Name:      "sd_dns_lookups_total",
-				Help:      "The number of DNS-SD lookups.",
-			}),
-		dnsSDLookupFailuresCount: prometheus.NewCounter(
-			prometheus.CounterOpts{
-				Namespace: namespace,
-				Name:      "sd_dns_lookup_failures_total",
-				Help:      "The number of DNS-SD lookup failures.",
-			}),
 	}
 
 	d.Discovery = refresh.NewDiscovery(
@@ -157,8 +155,7 @@ func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (
 			Mech:                "dns",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: prometheus.NewRegistry(),
+			MetricsInstantiator: m.refreshMetrics,
-			Metrics:  []prometheus.Collector{d.dnsSDLookupsCount, d.dnsSDLookupFailuresCount},
 		},
 	)
 
@@ -195,9 +192,9 @@ func (d *Discovery) refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 
 func (d *Discovery) refreshOne(ctx context.Context, name string, ch chan<- *targetgroup.Group) error {
 	response, err := d.lookupFn(name, d.qtype, d.logger)
-	d.dnsSDLookupsCount.Inc()
+	d.metrics.dnsSDLookupsCount.Inc()
 	if err != nil {
-		d.dnsSDLookupFailuresCount.Inc()
+		d.metrics.dnsSDLookupFailuresCount.Inc()
 		return err
 	}
 

discovery/dns/dns_test.go

@@ -28,6 +28,7 @@ import (
 	"go.uber.org/goleak"
 	"gopkg.in/yaml.v2"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -253,13 +254,21 @@ func TestDNS(t *testing.T) {
 		tc := tc
 		t.Run(tc.name, func(t *testing.T) {
 			t.Parallel()
-			sd, err := NewDiscovery(tc.config, nil, prometheus.NewRegistry())
+
+			reg := prometheus.NewRegistry()
+			refreshMetrics := discovery.NewRefreshMetrics(reg)
+			metrics := tc.config.NewDiscovererMetrics(reg, refreshMetrics)
+			require.NoError(t, metrics.Register())
+
+			sd, err := NewDiscovery(tc.config, nil, metrics)
 			require.NoError(t, err)
 			sd.lookupFn = tc.lookup
 
 			tgs, err := sd.refresh(context.Background())
 			require.NoError(t, err)
 			require.Equal(t, tc.expected, tgs)
+
+			metrics.Unregister()
 		})
 	}
 }

discovery/dns/metrics.go

@@ -0,0 +1,66 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package dns
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*dnsMetrics)(nil)
+
+type dnsMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+
+	dnsSDLookupsCount        prometheus.Counter
+	dnsSDLookupFailuresCount prometheus.Counter
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &dnsMetrics{
+		refreshMetrics: rmi,
+		dnsSDLookupsCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Namespace: namespace,
+				Name:      "sd_dns_lookups_total",
+				Help:      "The number of DNS-SD lookups.",
+			}),
+		dnsSDLookupFailuresCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Namespace: namespace,
+				Name:      "sd_dns_lookup_failures_total",
+				Help:      "The number of DNS-SD lookup failures.",
+			}),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.dnsSDLookupsCount,
+		m.dnsSDLookupFailuresCount,
+	})
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *dnsMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *dnsMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/eureka/eureka.go

@@ -16,6 +16,7 @@ package eureka
 import (
 	"context"
 	"errors"
+	"fmt"
 	"net"
 	"net/http"
 	"net/url"
@@ -76,12 +77,19 @@ type SDConfig struct {
 	RefreshInterval  model.Duration          `yaml:"refresh_interval,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &eurekaMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "eureka" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -118,7 +126,12 @@ type Discovery struct {
 }
 
 // NewDiscovery creates a new Eureka discovery for the given role.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*eurekaMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	rt, err := config.NewRoundTripperFromConfig(conf.HTTPClientConfig, "eureka_sd")
 	if err != nil {
 		return nil, err
@@ -134,7 +147,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "eureka",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil

discovery/eureka/eureka_test.go

@@ -24,6 +24,7 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -36,7 +37,17 @@ func testUpdateServices(respHandler http.HandlerFunc) ([]*targetgroup.Group, err
 		Server: ts.URL,
 	}
 
-	md, err := NewDiscovery(&conf, nil, prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := conf.NewDiscovererMetrics(reg, refreshMetrics)
+	err := metrics.Register()
+	if err != nil {
+		return nil, err
+	}
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	md, err := NewDiscovery(&conf, nil, metrics)
 	if err != nil {
 		return nil, err
 	}

discovery/eureka/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package eureka
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*eurekaMetrics)(nil)
+
+type eurekaMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *eurekaMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *eurekaMetrics) Unregister() {}

discovery/file/file.go

@@ -57,12 +57,17 @@ type SDConfig struct {
 	RefreshInterval model.Duration `yaml:"refresh_interval,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "file" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -94,6 +99,9 @@ func (c *SDConfig) UnmarshalYAML(unmarshal func(interface{}) error) error {
 const fileSDFilepathLabel = model.MetaLabelPrefix + "filepath"
 
 // TimestampCollector is a Custom Collector for Timestamps of the files.
+// TODO(ptodev): Now that each file SD has its own TimestampCollector
+// inside discovery/file/metrics.go, we can refactor this collector
+// (or get rid of it) as each TimestampCollector instance will only use one discoverer.
 type TimestampCollector struct {
 	Description *prometheus.Desc
 	discoverers map[*Discovery]struct{}
@@ -169,16 +177,16 @@ type Discovery struct {
 	lastRefresh map[string]int
 	logger      log.Logger
 
-	fileSDReadErrorsCount  prometheus.Counter
+	metrics *fileMetrics
-	fileSDScanDuration     prometheus.Summary
-	fileWatcherErrorsCount prometheus.Counter
-	fileSDTimeStamp        *TimestampCollector
-
-	metricRegisterer discovery.MetricRegisterer
 }
 
 // NewDiscovery returns a new file discovery for the given paths.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	fm, ok := metrics.(*fileMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -188,33 +196,10 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 		interval:   time.Duration(conf.RefreshInterval),
 		timestamps: make(map[string]float64),
 		logger:     logger,
-		fileSDReadErrorsCount: prometheus.NewCounter(
+		metrics:    fm,
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_file_read_errors_total",
-				Help: "The number of File-SD read errors.",
-			}),
-		fileSDScanDuration: prometheus.NewSummary(
-			prometheus.SummaryOpts{
-				Name:       "prometheus_sd_file_scan_duration_seconds",
-				Help:       "The duration of the File-SD scan in seconds.",
-				Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
-			}),
-		fileWatcherErrorsCount: prometheus.NewCounter(
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_file_watcher_errors_total",
-				Help: "The number of File-SD errors caused by filesystem watch failures.",
-			}),
-		fileSDTimeStamp: NewTimestampCollector(),
 	}
 
-	disc.fileSDTimeStamp.addDiscoverer(disc)
+	fm.init(disc)
-
-	disc.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
-		disc.fileSDReadErrorsCount,
-		disc.fileSDScanDuration,
-		disc.fileWatcherErrorsCount,
-		disc.fileSDTimeStamp,
-	})
 
 	return disc, nil
 }
@@ -253,17 +238,10 @@ func (d *Discovery) watchFiles() {
 
 // Run implements the Discoverer interface.
 func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
-	err := d.metricRegisterer.RegisterMetrics()
-	if err != nil {
-		level.Error(d.logger).Log("msg", "Unable to register metrics", "err", err.Error())
-		return
-	}
-	defer d.metricRegisterer.UnregisterMetrics()
-
 	watcher, err := fsnotify.NewWatcher()
 	if err != nil {
 		level.Error(d.logger).Log("msg", "Error adding file watcher", "err", err)
-		d.fileWatcherErrorsCount.Inc()
+		d.metrics.fileWatcherErrorsCount.Inc()
 		return
 	}
 	d.watcher = watcher
@@ -327,7 +305,7 @@ func (d *Discovery) stop() {
 	done := make(chan struct{})
 	defer close(done)
 
-	d.fileSDTimeStamp.removeDiscoverer(d)
+	d.metrics.fileSDTimeStamp.removeDiscoverer(d)
 
 	// Closing the watcher will deadlock unless all events and errors are drained.
 	go func() {
@@ -353,13 +331,13 @@ func (d *Discovery) stop() {
 func (d *Discovery) refresh(ctx context.Context, ch chan<- []*targetgroup.Group) {
 	t0 := time.Now()
 	defer func() {
-		d.fileSDScanDuration.Observe(time.Since(t0).Seconds())
+		d.metrics.fileSDScanDuration.Observe(time.Since(t0).Seconds())
 	}()
 	ref := map[string]int{}
 	for _, p := range d.listFiles() {
 		tgroups, err := d.readFile(p)
 		if err != nil {
-			d.fileSDReadErrorsCount.Inc()
+			d.metrics.fileSDReadErrorsCount.Inc()
 
 			level.Error(d.logger).Log("msg", "Error reading file", "path", p, "err", err)
 			// Prevent deletion down below.

discovery/file/file_test.go

@@ -29,6 +29,7 @@ import (
 	"github.com/stretchr/testify/require"
 	"go.uber.org/goleak"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -144,19 +145,28 @@ func (t *testRunner) run(files ...string) {
 	ctx, cancel := context.WithCancel(context.Background())
 	t.cancelSD = cancel
 	go func() {
-		d, err := NewDiscovery(
+		conf := &SDConfig{
-			&SDConfig{
 			Files: files,
 			// Setting a high refresh interval to make sure that the tests only
 			// rely on file watches.
 			RefreshInterval: model.Duration(1 * time.Hour),
-			},
+		}
+
+		reg := prometheus.NewRegistry()
+		refreshMetrics := discovery.NewRefreshMetrics(reg)
+		metrics := conf.NewDiscovererMetrics(reg, refreshMetrics)
+		require.NoError(t, metrics.Register())
+
+		d, err := NewDiscovery(
+			conf,
 			nil,
-			prometheus.NewRegistry(),
+			metrics,
 		)
 		require.NoError(t, err)
 
 		d.Run(ctx, t.ch)
+
+		metrics.Unregister()
 	}()
 }
 
@@ -193,9 +203,10 @@ func (t *testRunner) targets() []*targetgroup.Group {
 func (t *testRunner) requireUpdate(ref time.Time, expected []*targetgroup.Group) {
 	t.Helper()
 
+	timeout := time.After(defaultWait)
 	for {
 		select {
-		case <-time.After(defaultWait):
+		case <-timeout:
 			t.Fatalf("Expected update but got none")
 			return
 		case <-time.After(defaultWait / 10):

discovery/file/metrics.go

@@ -0,0 +1,76 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package file
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*fileMetrics)(nil)
+
+type fileMetrics struct {
+	fileSDReadErrorsCount  prometheus.Counter
+	fileSDScanDuration     prometheus.Summary
+	fileWatcherErrorsCount prometheus.Counter
+	fileSDTimeStamp        *TimestampCollector
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	fm := &fileMetrics{
+		fileSDReadErrorsCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_file_read_errors_total",
+				Help: "The number of File-SD read errors.",
+			}),
+		fileSDScanDuration: prometheus.NewSummary(
+			prometheus.SummaryOpts{
+				Name:       "prometheus_sd_file_scan_duration_seconds",
+				Help:       "The duration of the File-SD scan in seconds.",
+				Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
+			}),
+		fileWatcherErrorsCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_file_watcher_errors_total",
+				Help: "The number of File-SD errors caused by filesystem watch failures.",
+			}),
+		fileSDTimeStamp: NewTimestampCollector(),
+	}
+
+	fm.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		fm.fileSDReadErrorsCount,
+		fm.fileSDScanDuration,
+		fm.fileWatcherErrorsCount,
+		fm.fileSDTimeStamp,
+	})
+
+	return fm
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (fm *fileMetrics) Register() error {
+	return fm.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (fm *fileMetrics) Unregister() {
+	fm.metricRegisterer.UnregisterMetrics()
+}
+
+func (fm *fileMetrics) init(disc *Discovery) {
+	fm.fileSDTimeStamp.addDiscoverer(disc)
+}

discovery/gce/gce.go

@@ -82,12 +82,19 @@ type SDConfig struct {
 	TagSeparator    string         `yaml:"tag_separator,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &gceMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "gce" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(*c, opts.Logger, opts.Registerer)
+	return NewDiscovery(*c, opts.Logger, opts.Metrics)
 }
 
 // UnmarshalYAML implements the yaml.Unmarshaler interface.
@@ -122,7 +129,12 @@ type Discovery struct {
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*gceMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	d := &Discovery{
 		project:      conf.Project,
 		zone:         conf.Zone,
@@ -147,7 +159,7 @@ func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (
 			Mech:                "gce",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil

discovery/gce/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package gce
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*gceMetrics)(nil)
+
+type gceMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *gceMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *gceMetrics) Unregister() {}

discovery/hetzner/hetzner.go

@@ -63,12 +63,19 @@ type SDConfig struct {
 	robotEndpoint   string         // For tests only.
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &hetznerMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "hetzner" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 type refresher interface {
@@ -128,7 +135,12 @@ type Discovery struct {
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*refresh.Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*refresh.Discovery, error) {
+	m, ok := metrics.(*hetznerMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	r, err := newRefresher(conf, logger)
 	if err != nil {
 		return nil, err
@@ -140,7 +152,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "hetzner",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            r.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	), nil
 }

discovery/hetzner/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package hetzner
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*hetznerMetrics)(nil)
+
+type hetznerMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *hetznerMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *hetznerMetrics) Unregister() {}

discovery/http/http.go

@@ -58,12 +58,17 @@ type SDConfig struct {
 	URL              string                  `yaml:"url"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "http" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.HTTPClientOptions, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.HTTPClientOptions, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -105,11 +110,16 @@ type Discovery struct {
 	client          *http.Client
 	refreshInterval time.Duration
 	tgLastLength    int
-	failuresCount   prometheus.Counter
+	metrics         *httpMetrics
 }
 
 // NewDiscovery returns a new HTTP discovery for the given config.
-func NewDiscovery(conf *SDConfig, logger log.Logger, clientOpts []config.HTTPClientOption, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, clientOpts []config.HTTPClientOption, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*httpMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -124,11 +134,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, clientOpts []config.HTTPCli
 		url:             conf.URL,
 		client:          client,
 		refreshInterval: time.Duration(conf.RefreshInterval), // Stored to be sent as headers.
-		failuresCount: prometheus.NewCounter(
+		metrics:         m,
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_http_failures_total",
-				Help: "Number of HTTP service discovery refresh failures.",
-			}),
 	}
 
 	d.Discovery = refresh.NewDiscovery(
@@ -137,8 +143,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, clientOpts []config.HTTPCli
 			Mech:                "http",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.Refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
-			Metrics:  []prometheus.Collector{d.failuresCount},
 		},
 	)
 	return d, nil
@@ -155,7 +160,7 @@ func (d *Discovery) Refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 
 	resp, err := d.client.Do(req.WithContext(ctx))
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, err
 	}
 	defer func() {
@@ -164,31 +169,31 @@ func (d *Discovery) Refresh(ctx context.Context) ([]*targetgroup.Group, error) {
 	}()
 
 	if resp.StatusCode != http.StatusOK {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, fmt.Errorf("server returned HTTP status %s", resp.Status)
 	}
 
 	if !matchContentType.MatchString(strings.TrimSpace(resp.Header.Get("Content-Type"))) {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, fmt.Errorf("unsupported content type %q", resp.Header.Get("Content-Type"))
 	}
 
 	b, err := io.ReadAll(resp.Body)
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, err
 	}
 
 	var targetGroups []*targetgroup.Group
 
 	if err := json.Unmarshal(b, &targetGroups); err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, err
 	}
 
 	for i, tg := range targetGroups {
 		if tg == nil {
-			d.failuresCount.Inc()
+			d.metrics.failuresCount.Inc()
 			err = errors.New("nil target group item found")
 			return nil, err
 		}

discovery/http/http_test.go

@@ -28,6 +28,7 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -41,7 +42,14 @@ func TestHTTPValidRefresh(t *testing.T) {
 		RefreshInterval:  model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	defer refreshMetrics.Unregister()
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
@@ -63,7 +71,7 @@ func TestHTTPValidRefresh(t *testing.T) {
 		},
 	}
 	require.Equal(t, expectedTargets, tgs)
-	require.Equal(t, 0.0, getFailureCount(d.failuresCount))
+	require.Equal(t, 0.0, getFailureCount(d.metrics.failuresCount))
 }
 
 func TestHTTPInvalidCode(t *testing.T) {
@@ -79,13 +87,20 @@ func TestHTTPInvalidCode(t *testing.T) {
 		RefreshInterval:  model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	defer refreshMetrics.Unregister()
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
 	_, err = d.Refresh(ctx)
 	require.EqualError(t, err, "server returned HTTP status 400 Bad Request")
-	require.Equal(t, 1.0, getFailureCount(d.failuresCount))
+	require.Equal(t, 1.0, getFailureCount(d.metrics.failuresCount))
 }
 
 func TestHTTPInvalidFormat(t *testing.T) {
@@ -101,13 +116,20 @@ func TestHTTPInvalidFormat(t *testing.T) {
 		RefreshInterval:  model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	defer refreshMetrics.Unregister()
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
 	_, err = d.Refresh(ctx)
 	require.EqualError(t, err, `unsupported content type "text/plain; charset=utf-8"`)
-	require.Equal(t, 1.0, getFailureCount(d.failuresCount))
+	require.Equal(t, 1.0, getFailureCount(d.metrics.failuresCount))
 }
 
 func getFailureCount(failuresCount prometheus.Counter) float64 {
@@ -412,7 +434,15 @@ func TestSourceDisappeared(t *testing.T) {
 		URL:              ts.URL,
 		RefreshInterval:  model.Duration(1 * time.Second),
 	}
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	defer refreshMetrics.Unregister()
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), nil, metrics)
 	require.NoError(t, err)
 	for _, test := range cases {
 		ctx := context.Background()

discovery/http/metrics.go

@@ -0,0 +1,57 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package http
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*httpMetrics)(nil)
+
+type httpMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+
+	failuresCount prometheus.Counter
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &httpMetrics{
+		refreshMetrics: rmi,
+		failuresCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_http_failures_total",
+				Help: "Number of HTTP service discovery refresh failures.",
+			}),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.failuresCount,
+	})
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *httpMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *httpMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/ionos/ionos.go

@@ -15,16 +15,16 @@ package ionos
 
 import (
 	"errors"
+	"fmt"
 	"time"
 
 	"github.com/go-kit/log"
+	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/common/config"
 	"github.com/prometheus/common/model"
 
 	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/refresh"
-
-	"github.com/prometheus/client_golang/prometheus"
 )
 
 const (
@@ -43,7 +43,12 @@ func init() {
 type Discovery struct{}
 
 // NewDiscovery returns a new refresh.Discovery for IONOS Cloud.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*refresh.Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*refresh.Discovery, error) {
+	m, ok := metrics.(*ionosMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if conf.ionosEndpoint == "" {
 		conf.ionosEndpoint = "https://api.ionos.com"
 	}
@@ -59,7 +64,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "ionos",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	), nil
 }
@@ -84,6 +89,13 @@ type SDConfig struct {
 	ionosEndpoint string // For tests only.
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &ionosMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the IONOS Cloud service discovery.
 func (c SDConfig) Name() string {
 	return "ionos"
@@ -91,7 +103,7 @@ func (c SDConfig) Name() string {
 
 // NewDiscoverer returns a new discovery.Discoverer for IONOS Cloud.
 func (c SDConfig) NewDiscoverer(options discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(&c, options.Logger, options.Registerer)
+	return NewDiscovery(&c, options.Logger, options.Metrics)
 }
 
 // UnmarshalYAML implements the yaml.Unmarshaler interface.

discovery/ionos/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package ionos
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*ionosMetrics)(nil)
+
+type ionosMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *ionosMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *ionosMetrics) Unregister() {}

discovery/kubernetes/endpoints.go

@@ -62,6 +62,8 @@ func NewEndpoints(l log.Logger, eps cache.SharedIndexInformer, svc, pod, node ca
 	svcUpdateCount := eventCount.WithLabelValues(RoleService.String(), MetricLabelRoleUpdate)
 	svcDeleteCount := eventCount.WithLabelValues(RoleService.String(), MetricLabelRoleDelete)
 
+	podUpdateCount := eventCount.WithLabelValues(RolePod.String(), MetricLabelRoleUpdate)
+
 	e := &Endpoints{
 		logger:           l,
 		endpointsInf:     eps,
@@ -131,6 +133,29 @@ func NewEndpoints(l log.Logger, eps cache.SharedIndexInformer, svc, pod, node ca
 	if err != nil {
 		level.Error(l).Log("msg", "Error adding services event handler.", "err", err)
 	}
+	_, err = e.podInf.AddEventHandler(cache.ResourceEventHandlerFuncs{
+		UpdateFunc: func(old, cur interface{}) {
+			podUpdateCount.Inc()
+			oldPod, ok := old.(*apiv1.Pod)
+			if !ok {
+				return
+			}
+
+			curPod, ok := cur.(*apiv1.Pod)
+			if !ok {
+				return
+			}
+
+			// the Pod's phase may change without triggering an update on the Endpoints/Service.
+			// https://github.com/prometheus/prometheus/issues/11305.
+			if curPod.Status.Phase != oldPod.Status.Phase {
+				e.enqueuePod(namespacedName(curPod.Namespace, curPod.Name))
+			}
+		},
+	})
+	if err != nil {
+		level.Error(l).Log("msg", "Error adding pods event handler.", "err", err)
+	}
 	if e.withNodeMetadata {
 		_, err = e.nodeInf.AddEventHandler(cache.ResourceEventHandlerFuncs{
 			AddFunc: func(o interface{}) {
@@ -166,6 +191,18 @@ func (e *Endpoints) enqueueNode(nodeName string) {
 	}
 }
 
+func (e *Endpoints) enqueuePod(podNamespacedName string) {
+	endpoints, err := e.endpointsInf.GetIndexer().ByIndex(podIndex, podNamespacedName)
+	if err != nil {
+		level.Error(e.logger).Log("msg", "Error getting endpoints for pod", "pod", podNamespacedName, "err", err)
+		return
+	}
+
+	for _, endpoint := range endpoints {
+		e.enqueue(endpoint)
+	}
+}
+
 func (e *Endpoints) enqueue(obj interface{}) {
 	key, err := cache.DeletionHandlingMetaNamespaceKeyFunc(obj)
 	if err != nil {
@@ -312,7 +349,7 @@ func (e *Endpoints) buildEndpoints(eps *apiv1.Endpoints) *targetgroup.Group {
 			tg.Targets = append(tg.Targets, target)
 			return
 		}
-		s := pod.Namespace + "/" + pod.Name
+		s := namespacedName(pod.Namespace, pod.Name)
 
 		sp, ok := seenPods[s]
 		if !ok {
@@ -369,6 +406,11 @@ func (e *Endpoints) buildEndpoints(eps *apiv1.Endpoints) *targetgroup.Group {
 	// For all seen pods, check all container ports. If they were not covered
 	// by one of the service endpoints, generate targets for them.
 	for _, pe := range seenPods {
+		// PodIP can be empty when a pod is starting or has been evicted.
+		if len(pe.pod.Status.PodIP) == 0 {
+			continue
+		}
+
 		for _, c := range pe.pod.Spec.Containers {
 			for _, cport := range c.Ports {
 				hasSeenPort := func() bool {
@@ -383,8 +425,6 @@ func (e *Endpoints) buildEndpoints(eps *apiv1.Endpoints) *targetgroup.Group {
 					continue
 				}
 
-				// PodIP can be empty when a pod is starting or has been evicted.
-				if len(pe.pod.Status.PodIP) != 0 {
 				a := net.JoinHostPort(pe.pod.Status.PodIP, strconv.FormatUint(uint64(cport.ContainerPort), 10))
 				ports := strconv.FormatUint(uint64(cport.ContainerPort), 10)
 
@@ -400,7 +440,6 @@ func (e *Endpoints) buildEndpoints(eps *apiv1.Endpoints) *targetgroup.Group {
 			}
 		}
 	}
-	}
 
 	return tg
 }

discovery/kubernetes/endpoints_test.go

@@ -969,3 +969,123 @@ func TestEndpointsDiscoveryEmptyPodStatus(t *testing.T) {
 		expectedRes:      map[string]*targetgroup.Group{},
 	}.Run(t)
 }
+
+// TestEndpointsUpdatePod makes sure that Endpoints discovery detects underlying Pods changes.
+// See https://github.com/prometheus/prometheus/issues/11305 for more details.
+func TestEndpointsDiscoveryUpdatePod(t *testing.T) {
+	pod := &v1.Pod{
+		ObjectMeta: metav1.ObjectMeta{
+			Name:      "testpod",
+			Namespace: "default",
+			UID:       types.UID("deadbeef"),
+		},
+		Spec: v1.PodSpec{
+			NodeName: "testnode",
+			Containers: []v1.Container{
+				{
+					Name:  "c1",
+					Image: "c1:latest",
+					Ports: []v1.ContainerPort{
+						{
+							Name:          "mainport",
+							ContainerPort: 9000,
+							Protocol:      v1.ProtocolTCP,
+						},
+					},
+				},
+			},
+		},
+		Status: v1.PodStatus{
+			// Pod is in Pending phase when discovered for first time.
+			Phase: "Pending",
+			Conditions: []v1.PodCondition{
+				{
+					Type:   v1.PodReady,
+					Status: v1.ConditionFalse,
+				},
+			},
+			HostIP: "2.3.4.5",
+			PodIP:  "4.3.2.1",
+		},
+	}
+	objs := []runtime.Object{
+		&v1.Endpoints{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:      "testendpoints",
+				Namespace: "default",
+			},
+			Subsets: []v1.EndpointSubset{
+				{
+					Addresses: []v1.EndpointAddress{
+						{
+							IP: "4.3.2.1",
+							// The Pending Pod may be included because the Endpoints was created manually.
+							// Or because the corresponding service has ".spec.publishNotReadyAddresses: true".
+							TargetRef: &v1.ObjectReference{
+								Kind:      "Pod",
+								Name:      "testpod",
+								Namespace: "default",
+							},
+						},
+					},
+					Ports: []v1.EndpointPort{
+						{
+							Name:     "mainport",
+							Port:     9000,
+							Protocol: v1.ProtocolTCP,
+						},
+					},
+				},
+			},
+		},
+		pod,
+	}
+	n, c := makeDiscovery(RoleEndpoint, NamespaceDiscovery{}, objs...)
+
+	k8sDiscoveryTest{
+		discovery: n,
+		afterStart: func() {
+			// the Pod becomes Ready.
+			pod.Status.Phase = "Running"
+			pod.Status.Conditions = []v1.PodCondition{
+				{
+					Type:   v1.PodReady,
+					Status: v1.ConditionTrue,
+				},
+			}
+			c.CoreV1().Pods(pod.Namespace).Update(context.Background(), pod, metav1.UpdateOptions{})
+		},
+		expectedMaxItems: 2,
+		expectedRes: map[string]*targetgroup.Group{
+			"endpoints/default/testendpoints": {
+				Targets: []model.LabelSet{
+					{
+						"__address__":                                    "4.3.2.1:9000",
+						"__meta_kubernetes_endpoint_port_name":           "mainport",
+						"__meta_kubernetes_endpoint_port_protocol":       "TCP",
+						"__meta_kubernetes_endpoint_ready":               "true",
+						"__meta_kubernetes_endpoint_address_target_kind": "Pod",
+						"__meta_kubernetes_endpoint_address_target_name": "testpod",
+						"__meta_kubernetes_pod_name":                     "testpod",
+						"__meta_kubernetes_pod_ip":                       "4.3.2.1",
+						"__meta_kubernetes_pod_ready":                    "true",
+						"__meta_kubernetes_pod_phase":                    "Running",
+						"__meta_kubernetes_pod_node_name":                "testnode",
+						"__meta_kubernetes_pod_host_ip":                  "2.3.4.5",
+						"__meta_kubernetes_pod_container_name":           "c1",
+						"__meta_kubernetes_pod_container_image":          "c1:latest",
+						"__meta_kubernetes_pod_container_port_name":      "mainport",
+						"__meta_kubernetes_pod_container_port_number":    "9000",
+						"__meta_kubernetes_pod_container_port_protocol":  "TCP",
+						"__meta_kubernetes_pod_uid":                      "deadbeef",
+					},
+				},
+				Labels: model.LabelSet{
+					"__meta_kubernetes_namespace":      "default",
+					"__meta_kubernetes_endpoints_name": "testendpoints",
+				},
+				Source: "endpoints/default/testendpoints",
+			},
+		},
+	}.Run(t)
+}

discovery/kubernetes/endpointslice.go

@@ -358,7 +358,7 @@ func (e *EndpointSlice) buildEndpointSlice(eps endpointSliceAdaptor) *targetgrou
 			tg.Targets = append(tg.Targets, target)
 			return
 		}
-		s := pod.Namespace + "/" + pod.Name
+		s := namespacedName(pod.Namespace, pod.Name)
 
 		sp, ok := seenPods[s]
 		if !ok {
@@ -405,6 +405,11 @@ func (e *EndpointSlice) buildEndpointSlice(eps endpointSliceAdaptor) *targetgrou
 	// For all seen pods, check all container ports. If they were not covered
 	// by one of the service endpoints, generate targets for them.
 	for _, pe := range seenPods {
+		// PodIP can be empty when a pod is starting or has been evicted.
+		if len(pe.pod.Status.PodIP) == 0 {
+			continue
+		}
+
 		for _, c := range pe.pod.Spec.Containers {
 			for _, cport := range c.Ports {
 				hasSeenPort := func() bool {
@@ -422,8 +427,6 @@ func (e *EndpointSlice) buildEndpointSlice(eps endpointSliceAdaptor) *targetgrou
 					continue
 				}
 
-				// PodIP can be empty when a pod is starting or has been evicted.
-				if len(pe.pod.Status.PodIP) != 0 {
 				a := net.JoinHostPort(pe.pod.Status.PodIP, strconv.FormatUint(uint64(cport.ContainerPort), 10))
 				ports := strconv.FormatUint(uint64(cport.ContainerPort), 10)
 
@@ -439,7 +442,6 @@ func (e *EndpointSlice) buildEndpointSlice(eps endpointSliceAdaptor) *targetgrou
 			}
 		}
 	}
-	}
 
 	return tg
 }

discovery/kubernetes/kubernetes.go

@@ -123,12 +123,17 @@ type SDConfig struct {
 	AttachMetadata     AttachMetadataConfig    `yaml:"attach_metadata,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "kubernetes" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return New(opts.Logger, opts.Registerer, c)
+	return New(opts.Logger, opts.Metrics, c)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -265,8 +270,7 @@ type Discovery struct {
 	selectors          roleSelector
 	ownNamespace       string
 	attachMetadata     AttachMetadataConfig
-	eventCount         *prometheus.CounterVec
+	metrics            *kubernetesMetrics
-	metricRegisterer   discovery.MetricRegisterer
 }
 
 func (d *Discovery) getNamespaces() []string {
@@ -285,7 +289,12 @@ func (d *Discovery) getNamespaces() []string {
 }
 
 // New creates a new Kubernetes discovery for the given role.
-func New(l log.Logger, reg prometheus.Registerer, conf *SDConfig) (*Discovery, error) {
+func New(l log.Logger, metrics discovery.DiscovererMetrics, conf *SDConfig) (*Discovery, error) {
+	m, ok := metrics.(*kubernetesMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if l == nil {
 		l = log.NewNopLogger()
 	}
@@ -348,34 +357,7 @@ func New(l log.Logger, reg prometheus.Registerer, conf *SDConfig) (*Discovery, e
 		selectors:          mapSelector(conf.Selectors),
 		ownNamespace:       ownNamespace,
 		attachMetadata:     conf.AttachMetadata,
-		eventCount: prometheus.NewCounterVec(
+		metrics:            m,
-			prometheus.CounterOpts{
-				Namespace: discovery.KubernetesMetricsNamespace,
-				Name:      "events_total",
-				Help:      "The number of Kubernetes events handled.",
-			},
-			[]string{"role", "event"},
-		),
-	}
-
-	d.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{d.eventCount})
-
-	// Initialize metric vectors.
-	for _, role := range []string{
-		RoleEndpointSlice.String(),
-		RoleEndpoint.String(),
-		RoleNode.String(),
-		RolePod.String(),
-		RoleService.String(),
-		RoleIngress.String(),
-	} {
-		for _, evt := range []string{
-			MetricLabelRoleAdd,
-			MetricLabelRoleDelete,
-			MetricLabelRoleUpdate,
-		} {
-			d.eventCount.WithLabelValues(role, evt)
-		}
 	}
 
 	return d, nil
@@ -415,13 +397,6 @@ const resyncDisabled = 0
 func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
 	d.Lock()
 
-	err := d.metricRegisterer.RegisterMetrics()
-	if err != nil {
-		level.Error(d.logger).Log("msg", "Unable to register metrics", "err", err.Error())
-		return
-	}
-	defer d.metricRegisterer.UnregisterMetrics()
-
 	namespaces := d.getNamespaces()
 
 	switch d.role {
@@ -513,7 +488,7 @@ func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
 				cache.NewSharedInformer(slw, &apiv1.Service{}, resyncDisabled),
 				cache.NewSharedInformer(plw, &apiv1.Pod{}, resyncDisabled),
 				nodeInf,
-				d.eventCount,
+				d.metrics.eventCount,
 			)
 			d.discoverers = append(d.discoverers, eps)
 			go eps.endpointSliceInf.Run(ctx.Done())
@@ -573,7 +548,7 @@ func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
 				cache.NewSharedInformer(slw, &apiv1.Service{}, resyncDisabled),
 				cache.NewSharedInformer(plw, &apiv1.Pod{}, resyncDisabled),
 				nodeInf,
-				d.eventCount,
+				d.metrics.eventCount,
 			)
 			d.discoverers = append(d.discoverers, eps)
 			go eps.endpointsInf.Run(ctx.Done())
@@ -605,7 +580,7 @@ func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
 				log.With(d.logger, "role", "pod"),
 				d.newPodsByNodeInformer(plw),
 				nodeInformer,
-				d.eventCount,
+				d.metrics.eventCount,
 			)
 			d.discoverers = append(d.discoverers, pod)
 			go pod.podInf.Run(ctx.Done())
@@ -628,7 +603,7 @@ func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
 			svc := NewService(
 				log.With(d.logger, "role", "service"),
 				cache.NewSharedInformer(slw, &apiv1.Service{}, resyncDisabled),
-				d.eventCount,
+				d.metrics.eventCount,
 			)
 			d.discoverers = append(d.discoverers, svc)
 			go svc.informer.Run(ctx.Done())
@@ -686,14 +661,14 @@ func (d *Discovery) Run(ctx context.Context, ch chan<- []*targetgroup.Group) {
 			ingress := NewIngress(
 				log.With(d.logger, "role", "ingress"),
 				informer,
-				d.eventCount,
+				d.metrics.eventCount,
 			)
 			d.discoverers = append(d.discoverers, ingress)
 			go ingress.informer.Run(ctx.Done())
 		}
 	case RoleNode:
 		nodeInformer := d.newNodeInformer(ctx)
-		node := NewNode(log.With(d.logger, "role", "node"), nodeInformer, d.eventCount)
+		node := NewNode(log.With(d.logger, "role", "node"), nodeInformer, d.metrics.eventCount)
 		d.discoverers = append(d.discoverers, node)
 		go node.informer.Run(ctx.Done())
 	default:
@@ -792,6 +767,21 @@ func (d *Discovery) newPodsByNodeInformer(plw *cache.ListWatch) cache.SharedInde
 
 func (d *Discovery) newEndpointsByNodeInformer(plw *cache.ListWatch) cache.SharedIndexInformer {
 	indexers := make(map[string]cache.IndexFunc)
+	indexers[podIndex] = func(obj interface{}) ([]string, error) {
+		e, ok := obj.(*apiv1.Endpoints)
+		if !ok {
+			return nil, fmt.Errorf("object is not endpoints")
+		}
+		var pods []string
+		for _, target := range e.Subsets {
+			for _, addr := range target.Addresses {
+				if addr.TargetRef != nil && addr.TargetRef.Kind == "Pod" {
+					pods = append(pods, namespacedName(addr.TargetRef.Namespace, addr.TargetRef.Name))
+				}
+			}
+		}
+		return pods, nil
+	}
 	if !d.attachMetadata.Node {
 		return cache.NewSharedIndexInformer(plw, &apiv1.Endpoints{}, resyncDisabled, indexers)
 	}
@@ -897,3 +887,7 @@ func addObjectMetaLabels(labelSet model.LabelSet, objectMeta metav1.ObjectMeta,
 		labelSet[model.LabelName(metaLabelPrefix+string(role)+"_annotationpresent_"+ln)] = presentValue
 	}
 }
+
+func namespacedName(namespace, name string) string {
+	return namespace + "/" + name
+}

discovery/kubernetes/kubernetes_test.go

@@ -51,24 +51,29 @@ func makeDiscoveryWithVersion(role Role, nsDiscovery NamespaceDiscovery, k8sVer
 	fakeDiscovery, _ := clientset.Discovery().(*fakediscovery.FakeDiscovery)
 	fakeDiscovery.FakedServerVersion = &version.Info{GitVersion: k8sVer}
 
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := newDiscovererMetrics(reg, refreshMetrics)
+	err := metrics.Register()
+	if err != nil {
+		panic(err)
+	}
+	// TODO(ptodev): Unregister the metrics at the end of the test.
+
+	kubeMetrics, ok := metrics.(*kubernetesMetrics)
+	if !ok {
+		panic("invalid discovery metrics type")
+	}
+
 	d := &Discovery{
 		client:             clientset,
 		logger:             log.NewNopLogger(),
 		role:               role,
 		namespaceDiscovery: &nsDiscovery,
 		ownNamespace:       "own-ns",
-		eventCount: prometheus.NewCounterVec(
+		metrics:            kubeMetrics,
-			prometheus.CounterOpts{
-				Namespace: discovery.KubernetesMetricsNamespace,
-				Name:      "events_total",
-				Help:      "The number of Kubernetes events handled.",
-			},
-			[]string{"role", "event"},
-		),
 	}
 
-	d.metricRegisterer = discovery.NewMetricRegisterer(prometheus.NewRegistry(), []prometheus.Collector{d.eventCount})
-
 	return d, clientset
 }
 

discovery/kubernetes/metrics.go

@@ -0,0 +1,75 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package kubernetes
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*kubernetesMetrics)(nil)
+
+type kubernetesMetrics struct {
+	eventCount *prometheus.CounterVec
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &kubernetesMetrics{
+		eventCount: prometheus.NewCounterVec(
+			prometheus.CounterOpts{
+				Namespace: discovery.KubernetesMetricsNamespace,
+				Name:      "events_total",
+				Help:      "The number of Kubernetes events handled.",
+			},
+			[]string{"role", "event"},
+		),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.eventCount,
+	})
+
+	// Initialize metric vectors.
+	for _, role := range []string{
+		RoleEndpointSlice.String(),
+		RoleEndpoint.String(),
+		RoleNode.String(),
+		RolePod.String(),
+		RoleService.String(),
+		RoleIngress.String(),
+	} {
+		for _, evt := range []string{
+			MetricLabelRoleAdd,
+			MetricLabelRoleDelete,
+			MetricLabelRoleUpdate,
+		} {
+			m.eventCount.WithLabelValues(role, evt)
+		}
+	}
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *kubernetesMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *kubernetesMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/kubernetes/pod.go

@@ -33,7 +33,10 @@ import (
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
-const nodeIndex = "node"
+const (
+	nodeIndex = "node"
+	podIndex  = "pod"
+)
 
 // Pod discovers new pod targets.
 type Pod struct {
@@ -326,7 +329,7 @@ func podSource(pod *apiv1.Pod) string {
 }
 
 func podSourceFromNamespaceAndName(namespace, name string) string {
-	return "pod/" + namespace + "/" + name
+	return "pod/" + namespacedName(namespace, name)
 }
 
 func podReady(pod *apiv1.Pod) model.LabelValue {

discovery/legacymanager/manager.go

@@ -42,7 +42,7 @@ type provider struct {
 }
 
 // NewManager is the Discovery Manager constructor.
-func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Registerer, options ...func(*Manager)) *Manager {
+func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Registerer, sdMetrics map[string]discovery.DiscovererMetrics, options ...func(*Manager)) *Manager {
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -55,6 +55,7 @@ func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Re
 		updatert:       5 * time.Second,
 		triggerSend:    make(chan struct{}, 1),
 		registerer:     registerer,
+		sdMetrics:      sdMetrics,
 	}
 	for _, option := range options {
 		option(mgr)
@@ -62,7 +63,7 @@ func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Re
 
 	// Register the metrics.
 	// We have to do this after setting all options, so that the name of the Manager is set.
-	if metrics, err := discovery.NewMetrics(registerer, mgr.name); err == nil {
+	if metrics, err := discovery.NewManagerMetrics(registerer, mgr.name); err == nil {
 		mgr.metrics = metrics
 	} else {
 		level.Error(logger).Log("msg", "Failed to create discovery manager metrics", "manager", mgr.name, "err", err)
@@ -109,6 +110,7 @@ type Manager struct {
 	registerer prometheus.Registerer
 
 	metrics   *discovery.Metrics
+	sdMetrics map[string]discovery.DiscovererMetrics
 }
 
 // Run starts the background processing.
@@ -284,7 +286,7 @@ func (m *Manager) registerProviders(cfgs discovery.Configs, setName string) int
 		typ := cfg.Name()
 		d, err := cfg.NewDiscoverer(discovery.DiscovererOptions{
 			Logger:  log.With(m.logger, "discovery", typ, "config", setName),
-			Registerer: m.registerer,
+			Metrics: m.sdMetrics[typ],
 		})
 		if err != nil {
 			level.Error(m.logger).Log("msg", "Cannot create service discovery", "err", err, "type", typ, "config", setName)

discovery/legacymanager/manager_test.go

@@ -36,6 +36,13 @@ func TestMain(m *testing.M) {
 	testutil.TolerantVerifyLeak(m)
 }
 
+func newTestMetrics(t *testing.T, reg prometheus.Registerer) (*discovery.RefreshMetricsManager, map[string]discovery.DiscovererMetrics) {
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	sdMetrics, err := discovery.RegisterSDMetrics(reg, refreshMetrics)
+	require.NoError(t, err)
+	return &refreshMetrics, sdMetrics
+}
+
 // TestTargetUpdatesOrder checks that the target updates are received in the expected order.
 func TestTargetUpdatesOrder(t *testing.T) {
 	// The order by which the updates are send is determined by the interval passed to the mock discovery adapter
@@ -665,7 +672,10 @@ func TestTargetUpdatesOrder(t *testing.T) {
 			ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 			defer cancel()
 
-			discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+			reg := prometheus.NewRegistry()
+			_, sdMetrics := newTestMetrics(t, reg)
+
+			discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 			require.NotNil(t, discoveryManager)
 			discoveryManager.updatert = 100 * time.Millisecond
 
@@ -748,7 +758,11 @@ func verifyPresence(t *testing.T, tSets map[poolKey]map[string]*targetgroup.Grou
 func TestTargetSetRecreatesTargetGroupsEveryRun(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := newTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -777,7 +791,11 @@ func TestTargetSetRecreatesTargetGroupsEveryRun(t *testing.T) {
 func TestDiscovererConfigs(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := newTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -802,7 +820,11 @@ func TestDiscovererConfigs(t *testing.T) {
 func TestTargetSetRecreatesEmptyStaticConfigs(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := newTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -842,7 +864,11 @@ func TestTargetSetRecreatesEmptyStaticConfigs(t *testing.T) {
 func TestIdenticalConfigurationsAreCoalesced(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, nil, prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := newTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, nil, reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -874,7 +900,11 @@ func TestApplyConfigDoesNotModifyStaticTargets(t *testing.T) {
 	}
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := newTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -897,10 +927,19 @@ func (e errorConfig) NewDiscoverer(discovery.DiscovererOptions) (discovery.Disco
 	return nil, e.err
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (errorConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &discovery.NoopDiscovererMetrics{}
+}
+
 func TestGaugeFailedConfigs(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := newTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1057,7 +1096,10 @@ func TestCoordinationWithReceiver(t *testing.T) {
 			ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 			defer cancel()
 
-			mgr := NewManager(ctx, nil, prometheus.NewRegistry())
+			reg := prometheus.NewRegistry()
+			_, sdMetrics := newTestMetrics(t, reg)
+
+			mgr := NewManager(ctx, nil, reg, sdMetrics)
 			require.NotNil(t, mgr)
 			mgr.updatert = updateDelay
 			go mgr.Run()

discovery/linode/linode.go

@@ -87,12 +87,17 @@ type SDConfig struct {
 	TagSeparator    string         `yaml:"tag_separator,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "linode" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -122,22 +127,23 @@ type Discovery struct {
 	pollCount            int
 	lastResults          []*targetgroup.Group
 	eventPollingEnabled  bool
-	failuresCount        prometheus.Counter
+	metrics              *linodeMetrics
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*linodeMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	d := &Discovery{
 		port:                 conf.Port,
 		tagSeparator:         conf.TagSeparator,
 		pollCount:            0,
 		lastRefreshTimestamp: time.Now().UTC(),
 		eventPollingEnabled:  true,
-		failuresCount: prometheus.NewCounter(
+		metrics:              m,
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_linode_failures_total",
-				Help: "Number of Linode service discovery refresh failures.",
-			}),
 	}
 
 	rt, err := config.NewRoundTripperFromConfig(conf.HTTPClientConfig, "linode_sd")
@@ -160,8 +166,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "linode",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
-			Metrics:  []prometheus.Collector{d.failuresCount},
 		},
 	)
 	return d, nil
@@ -223,14 +228,14 @@ func (d *Discovery) refreshData(ctx context.Context) ([]*targetgroup.Group, erro
 	// Gather all linode instances.
 	instances, err := d.client.ListInstances(ctx, &linodego.ListOptions{PageSize: 500})
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, err
 	}
 
 	// Gather detailed IP address info for all IPs on all linode instances.
 	detailedIPs, err := d.client.ListIPAddresses(ctx, &linodego.ListOptions{PageSize: 500})
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, err
 	}
 

discovery/linode/linode_test.go

@@ -24,6 +24,8 @@ import (
 	"github.com/prometheus/common/config"
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 type LinodeSDTestSuite struct {
@@ -53,7 +55,15 @@ func TestLinodeSDRefresh(t *testing.T) {
 		Credentials: tokenID,
 		Type:        "Bearer",
 	}
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 	endpoint, err := url.Parse(sdmock.Mock.Endpoint())
 	require.NoError(t, err)

discovery/linode/metrics.go

@@ -0,0 +1,57 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package linode
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*linodeMetrics)(nil)
+
+type linodeMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+
+	failuresCount prometheus.Counter
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &linodeMetrics{
+		refreshMetrics: rmi,
+		failuresCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_linode_failures_total",
+				Help: "Number of Linode service discovery refresh failures.",
+			}),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.failuresCount,
+	})
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *linodeMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *linodeMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/manager.go

@@ -64,8 +64,24 @@ func (p *Provider) Config() interface{} {
 	return p.config
 }
 
+// Registers the metrics needed for SD mechanisms.
+// Does not register the metrics for the Discovery Manager.
+// TODO(ptodev): Add ability to unregister the metrics?
+func CreateAndRegisterSDMetrics(reg prometheus.Registerer) (map[string]DiscovererMetrics, error) {
+	// Some SD mechanisms use the "refresh" package, which has its own metrics.
+	refreshSdMetrics := NewRefreshMetrics(reg)
+
+	// Register the metrics specific for each SD mechanism, and the ones for the refresh package.
+	sdMetrics, err := RegisterSDMetrics(reg, refreshSdMetrics)
+	if err != nil {
+		return nil, fmt.Errorf("failed to register service discovery metrics: %w", err)
+	}
+
+	return sdMetrics, nil
+}
+
 // NewManager is the Discovery Manager constructor.
-func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Registerer, options ...func(*Manager)) *Manager {
+func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Registerer, sdMetrics map[string]DiscovererMetrics, options ...func(*Manager)) *Manager {
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -77,6 +93,7 @@ func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Re
 		updatert:    5 * time.Second,
 		triggerSend: make(chan struct{}, 1),
 		registerer:  registerer,
+		sdMetrics:   sdMetrics,
 	}
 	for _, option := range options {
 		option(mgr)
@@ -84,7 +101,7 @@ func NewManager(ctx context.Context, logger log.Logger, registerer prometheus.Re
 
 	// Register the metrics.
 	// We have to do this after setting all options, so that the name of the Manager is set.
-	if metrics, err := NewMetrics(registerer, mgr.name); err == nil {
+	if metrics, err := NewManagerMetrics(registerer, mgr.name); err == nil {
 		mgr.metrics = metrics
 	} else {
 		level.Error(logger).Log("msg", "Failed to create discovery manager metrics", "manager", mgr.name, "err", err)
@@ -144,6 +161,7 @@ type Manager struct {
 	registerer prometheus.Registerer
 
 	metrics   *Metrics
+	sdMetrics map[string]DiscovererMetrics
 }
 
 // Providers returns the currently configured SD providers.
@@ -402,7 +420,7 @@ func (m *Manager) registerProviders(cfgs Configs, setName string) int {
 		d, err := cfg.NewDiscoverer(DiscovererOptions{
 			Logger:            log.With(m.logger, "discovery", typ, "config", setName),
 			HTTPClientOptions: m.httpOpts,
-			Registerer:        m.registerer,
+			Metrics:           m.sdMetrics[typ],
 		})
 		if err != nil {
 			level.Error(m.logger).Log("msg", "Cannot create service discovery", "err", err, "type", typ, "config", setName)

discovery/manager_test.go

@@ -36,6 +36,13 @@ func TestMain(m *testing.M) {
 	testutil.TolerantVerifyLeak(m)
 }
 
+func NewTestMetrics(t *testing.T, reg prometheus.Registerer) (*RefreshMetricsManager, map[string]DiscovererMetrics) {
+	refreshMetrics := NewRefreshMetrics(reg)
+	sdMetrics, err := RegisterSDMetrics(reg, refreshMetrics)
+	require.NoError(t, err)
+	return &refreshMetrics, sdMetrics
+}
+
 // TestTargetUpdatesOrder checks that the target updates are received in the expected order.
 func TestTargetUpdatesOrder(t *testing.T) {
 	// The order by which the updates are send is determined by the interval passed to the mock discovery adapter
@@ -665,7 +672,10 @@ func TestTargetUpdatesOrder(t *testing.T) {
 			ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 			defer cancel()
 
-			discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+			reg := prometheus.NewRegistry()
+			_, sdMetrics := NewTestMetrics(t, reg)
+
+			discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 			require.NotNil(t, discoveryManager)
 			discoveryManager.updatert = 100 * time.Millisecond
 
@@ -780,7 +790,11 @@ func pk(provider, setName string, n int) poolKey {
 func TestTargetSetTargetGroupsPresentOnConfigReload(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -813,7 +827,11 @@ func TestTargetSetTargetGroupsPresentOnConfigReload(t *testing.T) {
 func TestTargetSetTargetGroupsPresentOnConfigRename(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -849,7 +867,11 @@ func TestTargetSetTargetGroupsPresentOnConfigRename(t *testing.T) {
 func TestTargetSetTargetGroupsPresentOnConfigDuplicateAndDeleteOriginal(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -888,7 +910,11 @@ func TestTargetSetTargetGroupsPresentOnConfigDuplicateAndDeleteOriginal(t *testi
 func TestTargetSetTargetGroupsPresentOnConfigChange(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -950,7 +976,11 @@ func TestTargetSetTargetGroupsPresentOnConfigChange(t *testing.T) {
 func TestTargetSetRecreatesTargetGroupsOnConfigChange(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -990,7 +1020,11 @@ func TestTargetSetRecreatesTargetGroupsOnConfigChange(t *testing.T) {
 func TestDiscovererConfigs(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1023,7 +1057,11 @@ func TestDiscovererConfigs(t *testing.T) {
 func TestTargetSetRecreatesEmptyStaticConfigs(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1071,7 +1109,11 @@ func TestTargetSetRecreatesEmptyStaticConfigs(t *testing.T) {
 func TestIdenticalConfigurationsAreCoalesced(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, nil, prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, nil, reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1108,7 +1150,11 @@ func TestApplyConfigDoesNotModifyStaticTargets(t *testing.T) {
 	}
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1129,11 +1175,21 @@ type errorConfig struct{ err error }
 func (e errorConfig) Name() string                                        { return "error" }
 func (e errorConfig) NewDiscoverer(DiscovererOptions) (Discoverer, error) { return nil, e.err }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (errorConfig) NewDiscovererMetrics(prometheus.Registerer, RefreshMetricsInstantiator) DiscovererMetrics {
+	return &NoopDiscovererMetrics{}
+}
+
 type lockStaticConfig struct {
 	mu     *sync.Mutex
 	config StaticConfig
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (lockStaticConfig) NewDiscovererMetrics(prometheus.Registerer, RefreshMetricsInstantiator) DiscovererMetrics {
+	return &NoopDiscovererMetrics{}
+}
+
 func (s lockStaticConfig) Name() string { return "lockstatic" }
 func (s lockStaticConfig) NewDiscoverer(DiscovererOptions) (Discoverer, error) {
 	return (lockStaticDiscoverer)(s), nil
@@ -1155,7 +1211,11 @@ func (s lockStaticDiscoverer) Run(ctx context.Context, up chan<- []*targetgroup.
 func TestGaugeFailedConfigs(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1312,7 +1372,10 @@ func TestCoordinationWithReceiver(t *testing.T) {
 			ctx, cancel := context.WithTimeout(context.Background(), 5*time.Second)
 			defer cancel()
 
-			mgr := NewManager(ctx, nil, prometheus.NewRegistry())
+			reg := prometheus.NewRegistry()
+			_, sdMetrics := NewTestMetrics(t, reg)
+
+			mgr := NewManager(ctx, nil, reg, sdMetrics)
 			require.NotNil(t, mgr)
 			mgr.updatert = updateDelay
 			go mgr.Run()
@@ -1408,7 +1471,11 @@ func (o onceProvider) Run(_ context.Context, ch chan<- []*targetgroup.Group) {
 func TestTargetSetTargetGroupsUpdateDuringApplyConfig(t *testing.T) {
 	ctx, cancel := context.WithCancel(context.Background())
 	defer cancel()
-	discoveryManager := NewManager(ctx, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	_, sdMetrics := NewTestMetrics(t, reg)
+
+	discoveryManager := NewManager(ctx, log.NewNopLogger(), reg, sdMetrics)
 	require.NotNil(t, discoveryManager)
 	discoveryManager.updatert = 100 * time.Millisecond
 	go discoveryManager.Run()
@@ -1470,6 +1537,11 @@ func newTestDiscoverer() *testDiscoverer {
 	}
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*testDiscoverer) NewDiscovererMetrics(prometheus.Registerer, RefreshMetricsInstantiator) DiscovererMetrics {
+	return &NoopDiscovererMetrics{}
+}
+
 // Name implements Config.
 func (t *testDiscoverer) Name() string {
 	return "test"

discovery/marathon/marathon.go

@@ -79,12 +79,19 @@ type SDConfig struct {
 	HTTPClientConfig config.HTTPClientConfig `yaml:",inline"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &marathonMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "marathon" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(*c, opts.Logger, opts.Registerer)
+	return NewDiscovery(*c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -133,7 +140,12 @@ type Discovery struct {
 }
 
 // NewDiscovery returns a new Marathon Discovery.
-func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*marathonMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	rt, err := config.NewRoundTripperFromConfig(conf.HTTPClientConfig, "marathon_sd")
 	if err != nil {
 		return nil, err
@@ -160,7 +172,7 @@ func NewDiscovery(conf SDConfig, logger log.Logger, reg prometheus.Registerer) (
 			Mech:                "marathon",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil

discovery/marathon/marathon_test.go

@@ -23,7 +23,9 @@ import (
 
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/common/model"
+	"github.com/stretchr/testify/require"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -37,7 +39,19 @@ func testConfig() SDConfig {
 }
 
 func testUpdateServices(client appListClient) ([]*targetgroup.Group, error) {
-	md, err := NewDiscovery(testConfig(), nil, prometheus.NewRegistry())
+	cfg := testConfig()
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	err := metrics.Register()
+	if err != nil {
+		return nil, err
+	}
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	md, err := NewDiscovery(cfg, nil, metrics)
 	if err != nil {
 		return nil, err
 	}
@@ -130,7 +144,15 @@ func TestMarathonSDSendGroup(t *testing.T) {
 }
 
 func TestMarathonSDRemoveApp(t *testing.T) {
-	md, err := NewDiscovery(testConfig(), nil, prometheus.NewRegistry())
+	cfg := testConfig()
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	md, err := NewDiscovery(cfg, nil, metrics)
 	if err != nil {
 		t.Fatalf("%s", err)
 	}

discovery/marathon/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package marathon
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*marathonMetrics)(nil)
+
+type marathonMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *marathonMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *marathonMetrics) Unregister() {}

discovery/metrics.go

@@ -38,7 +38,7 @@ type Metrics struct {
 	SentUpdates       prometheus.Counter
 }
 
-func NewMetrics(registerer prometheus.Registerer, sdManagerName string) (*Metrics, error) {
+func NewManagerMetrics(registerer prometheus.Registerer, sdManagerName string) (*Metrics, error) {
 	m := &Metrics{}
 
 	m.FailedConfigs = prometheus.NewGauge(

discovery/metrics_refresh.go

@@ -0,0 +1,75 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package discovery
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+)
+
+// Metric vectors for the "refresh" package.
+// We define them here in the "discovery" package in order to avoid a cyclic dependency between
+// "discovery" and "refresh".
+type RefreshMetricsVecs struct {
+	failuresVec *prometheus.CounterVec
+	durationVec *prometheus.SummaryVec
+
+	metricRegisterer MetricRegisterer
+}
+
+var _ RefreshMetricsManager = (*RefreshMetricsVecs)(nil)
+
+func NewRefreshMetrics(reg prometheus.Registerer) RefreshMetricsManager {
+	m := &RefreshMetricsVecs{
+		failuresVec: prometheus.NewCounterVec(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_refresh_failures_total",
+				Help: "Number of refresh failures for the given SD mechanism.",
+			},
+			[]string{"mechanism"}),
+		durationVec: prometheus.NewSummaryVec(
+			prometheus.SummaryOpts{
+				Name:       "prometheus_sd_refresh_duration_seconds",
+				Help:       "The duration of a refresh in seconds for the given SD mechanism.",
+				Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
+			},
+			[]string{"mechanism"}),
+	}
+
+	// The reason we register metric vectors instead of metrics is so that
+	// the metrics are not visible until they are recorded.
+	m.metricRegisterer = NewMetricRegisterer(reg, []prometheus.Collector{
+		m.failuresVec,
+		m.durationVec,
+	})
+
+	return m
+}
+
+// Instantiate returns metrics out of metric vectors.
+func (m *RefreshMetricsVecs) Instantiate(mech string) *RefreshMetrics {
+	return &RefreshMetrics{
+		Failures: m.failuresVec.WithLabelValues(mech),
+		Duration: m.durationVec.WithLabelValues(mech),
+	}
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *RefreshMetricsVecs) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *RefreshMetricsVecs) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/moby/docker.go

@@ -22,7 +22,7 @@ import (
 	"strconv"
 	"time"
 
-	"github.com/docker/docker/api/types"
+	"github.com/docker/docker/api/types/container"
 	"github.com/docker/docker/api/types/filters"
 	"github.com/docker/docker/client"
 	"github.com/go-kit/log"
@@ -76,12 +76,19 @@ type DockerSDConfig struct {
 	RefreshInterval model.Duration `yaml:"refresh_interval"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*DockerSDConfig) NewDiscovererMetrics(_ prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &dockerMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*DockerSDConfig) Name() string { return "docker" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *DockerSDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDockerDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDockerDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -115,8 +122,11 @@ type DockerDiscovery struct {
 }
 
 // NewDockerDiscovery returns a new DockerDiscovery which periodically refreshes its targets.
-func NewDockerDiscovery(conf *DockerSDConfig, logger log.Logger, reg prometheus.Registerer) (*DockerDiscovery, error) {
+func NewDockerDiscovery(conf *DockerSDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*DockerDiscovery, error) {
-	var err error
+	m, ok := metrics.(*dockerMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
 
 	d := &DockerDiscovery{
 		port:               conf.Port,
@@ -171,7 +181,7 @@ func NewDockerDiscovery(conf *DockerSDConfig, logger log.Logger, reg prometheus.
 			Mech:                "docker",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil
@@ -182,7 +192,7 @@ func (d *DockerDiscovery) refresh(ctx context.Context) ([]*targetgroup.Group, er
 		Source: "Docker",
 	}
 
-	containers, err := d.client.ContainerList(ctx, types.ContainerListOptions{Filters: d.filters})
+	containers, err := d.client.ContainerList(ctx, container.ListOptions{Filters: d.filters})
 	if err != nil {
 		return nil, fmt.Errorf("error while listing containers: %w", err)
 	}

discovery/moby/docker_test.go

@@ -23,6 +23,8 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 	"gopkg.in/yaml.v2"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 func TestDockerSDRefresh(t *testing.T) {
@@ -38,7 +40,14 @@ host: %s
 	var cfg DockerSDConfig
 	require.NoError(t, yaml.Unmarshal([]byte(cfgString), &cfg))
 
-	d, err := NewDockerDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDockerDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()

discovery/moby/dockerswarm.go

@@ -70,12 +70,19 @@ type Filter struct {
 	Values []string `yaml:"values"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*DockerSwarmSDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &dockerswarmMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*DockerSwarmSDConfig) Name() string { return "dockerswarm" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *DockerSwarmSDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -118,8 +125,11 @@ type Discovery struct {
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf *DockerSwarmSDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *DockerSwarmSDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
-	var err error
+	m, ok := metrics.(*dockerswarmMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
 
 	d := &Discovery{
 		port: conf.Port,
@@ -174,7 +184,7 @@ func NewDiscovery(conf *DockerSwarmSDConfig, logger log.Logger, reg prometheus.R
 			Mech:                "dockerswarm",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil

discovery/moby/metrics_docker.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package moby
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*dockerMetrics)(nil)
+
+type dockerMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *dockerMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *dockerMetrics) Unregister() {}

discovery/moby/metrics_dockerswarm.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package moby
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*dockerswarmMetrics)(nil)
+
+type dockerswarmMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *dockerswarmMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *dockerswarmMetrics) Unregister() {}

discovery/moby/nodes_test.go

@@ -23,6 +23,8 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 	"gopkg.in/yaml.v2"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 func TestDockerSwarmNodesSDRefresh(t *testing.T) {
@@ -39,7 +41,14 @@ host: %s
 	var cfg DockerSwarmSDConfig
 	require.NoError(t, yaml.Unmarshal([]byte(cfgString), &cfg))
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()

discovery/moby/services_test.go

@@ -23,6 +23,8 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 	"gopkg.in/yaml.v2"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 func TestDockerSwarmSDServicesRefresh(t *testing.T) {
@@ -39,7 +41,14 @@ host: %s
 	var cfg DockerSwarmSDConfig
 	require.NoError(t, yaml.Unmarshal([]byte(cfgString), &cfg))
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
@@ -333,7 +342,14 @@ filters:
 	var cfg DockerSwarmSDConfig
 	require.NoError(t, yaml.Unmarshal([]byte(cfgString), &cfg))
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()

discovery/moby/tasks_test.go

@@ -23,6 +23,8 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 	"gopkg.in/yaml.v2"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 func TestDockerSwarmTasksSDRefresh(t *testing.T) {
@@ -39,7 +41,14 @@ host: %s
 	var cfg DockerSwarmSDConfig
 	require.NoError(t, yaml.Unmarshal([]byte(cfgString), &cfg))
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()

discovery/nomad/metrics.go

@@ -0,0 +1,57 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package nomad
+
+import (
+	"github.com/prometheus/client_golang/prometheus"
+
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*nomadMetrics)(nil)
+
+type nomadMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+
+	failuresCount prometheus.Counter
+
+	metricRegisterer discovery.MetricRegisterer
+}
+
+func newDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	m := &nomadMetrics{
+		refreshMetrics: rmi,
+		failuresCount: prometheus.NewCounter(
+			prometheus.CounterOpts{
+				Name: "prometheus_sd_nomad_failures_total",
+				Help: "Number of nomad service discovery refresh failures.",
+			}),
+	}
+
+	m.metricRegisterer = discovery.NewMetricRegisterer(reg, []prometheus.Collector{
+		m.failuresCount,
+	})
+
+	return m
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *nomadMetrics) Register() error {
+	return m.metricRegisterer.RegisterMetrics()
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *nomadMetrics) Unregister() {
+	m.metricRegisterer.UnregisterMetrics()
+}

discovery/nomad/nomad.go

@@ -74,12 +74,17 @@ type SDConfig struct {
 	TagSeparator     string                  `yaml:"tag_separator,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return newDiscovererMetrics(reg, rmi)
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "nomad" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -112,11 +117,16 @@ type Discovery struct {
 	region          string
 	server          string
 	tagSeparator    string
-	failuresCount   prometheus.Counter
+	metrics         *nomadMetrics
 }
 
 // NewDiscovery returns a new Discovery which periodically refreshes its targets.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*nomadMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	d := &Discovery{
 		allowStale:      conf.AllowStale,
 		namespace:       conf.Namespace,
@@ -124,11 +134,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 		region:          conf.Region,
 		server:          conf.Server,
 		tagSeparator:    conf.TagSeparator,
-		failuresCount: prometheus.NewCounter(
+		metrics:         m,
-			prometheus.CounterOpts{
-				Name: "prometheus_sd_nomad_failures_total",
-				Help: "Number of nomad service discovery refresh failures.",
-			}),
 	}
 
 	HTTPClient, err := config.NewClientFromConfig(conf.HTTPClientConfig, "nomad_sd")
@@ -155,8 +161,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "nomad",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
-			Metrics:  []prometheus.Collector{d.failuresCount},
 		},
 	)
 	return d, nil
@@ -168,7 +173,7 @@ func (d *Discovery) refresh(context.Context) ([]*targetgroup.Group, error) {
 	}
 	stubs, _, err := d.client.Services().List(opts)
 	if err != nil {
-		d.failuresCount.Inc()
+		d.metrics.failuresCount.Inc()
 		return nil, err
 	}
 
@@ -180,7 +185,7 @@ func (d *Discovery) refresh(context.Context) ([]*targetgroup.Group, error) {
 		for _, service := range stub.Services {
 			instances, _, err := d.client.Services().Get(service.ServiceName, opts)
 			if err != nil {
-				d.failuresCount.Inc()
+				d.metrics.failuresCount.Inc()
 				return nil, fmt.Errorf("failed to fetch services: %w", err)
 			}
 

discovery/nomad/nomad_test.go

@@ -25,6 +25,8 @@ import (
 	"github.com/prometheus/client_golang/prometheus"
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
+
+	"github.com/prometheus/prometheus/discovery"
 )
 
 type NomadSDTestSuite struct {
@@ -128,8 +130,16 @@ func TestConfiguredService(t *testing.T) {
 	conf := &SDConfig{
 		Server: "http://localhost:4646",
 	}
-	_, err := NewDiscovery(conf, nil, prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := conf.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+
+	_, err := NewDiscovery(conf, nil, metrics)
 	require.NoError(t, err)
+
+	metrics.Unregister()
 }
 
 func TestNomadSDRefresh(t *testing.T) {
@@ -142,7 +152,15 @@ func TestNomadSDRefresh(t *testing.T) {
 
 	cfg := DefaultSDConfig
 	cfg.Server = endpoint.String()
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	tgs, err := d.refresh(context.Background())

discovery/openstack/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package openstack
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+type openstackMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+var _ discovery.DiscovererMetrics = (*openstackMetrics)(nil)
+
+// Register implements discovery.DiscovererMetrics.
+func (m *openstackMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *openstackMetrics) Unregister() {}

discovery/openstack/openstack.go

@@ -66,12 +66,19 @@ type SDConfig struct {
 	Availability                string           `yaml:"availability,omitempty"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &openstackMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "openstack" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -135,7 +142,12 @@ type refresher interface {
 }
 
 // NewDiscovery returns a new OpenStack Discoverer which periodically refreshes its targets.
-func NewDiscovery(conf *SDConfig, l log.Logger, reg prometheus.Registerer) (*refresh.Discovery, error) {
+func NewDiscovery(conf *SDConfig, l log.Logger, metrics discovery.DiscovererMetrics) (*refresh.Discovery, error) {
+	m, ok := metrics.(*openstackMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	r, err := newRefresher(conf, l)
 	if err != nil {
 		return nil, err
@@ -146,7 +158,7 @@ func NewDiscovery(conf *SDConfig, l log.Logger, reg prometheus.Registerer) (*ref
 			Mech:                "openstack",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            r.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	), nil
 }

discovery/ovhcloud/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package ovhcloud
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*ovhcloudMetrics)(nil)
+
+type ovhcloudMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *ovhcloudMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *ovhcloudMetrics) Unregister() {}

discovery/ovhcloud/ovhcloud.go

@@ -53,6 +53,13 @@ type SDConfig struct {
 	Service           string         `yaml:"service"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &ovhcloudMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name implements the Discoverer interface.
 func (c SDConfig) Name() string {
 	return "ovhcloud"
@@ -94,7 +101,7 @@ func createClient(config *SDConfig) (*ovh.Client, error) {
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(options discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, options.Logger, options.Registerer)
+	return NewDiscovery(c, options.Logger, options.Metrics)
 }
 
 func init() {
@@ -141,7 +148,12 @@ func newRefresher(conf *SDConfig, logger log.Logger) (refresher, error) {
 }
 
 // NewDiscovery returns a new OVHcloud Discoverer which periodically refreshes its targets.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*refresh.Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*refresh.Discovery, error) {
+	m, ok := metrics.(*ovhcloudMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	r, err := newRefresher(conf, logger)
 	if err != nil {
 		return nil, err
@@ -153,7 +165,7 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 			Mech:                "ovhcloud",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            r.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	), nil
 }

discovery/ovhcloud/ovhcloud_test.go

@@ -122,9 +122,17 @@ func TestParseIPs(t *testing.T) {
 func TestDiscoverer(t *testing.T) {
 	conf, _ := getMockConf("vps")
 	logger := testutil.NewLogger(t)
+
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := conf.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+	defer metrics.Unregister()
+	defer refreshMetrics.Unregister()
+
 	_, err := conf.NewDiscoverer(discovery.DiscovererOptions{
 		Logger:  logger,
-		Registerer: prometheus.NewRegistry(),
+		Metrics: metrics,
 	})
 
 	require.NoError(t, err)

discovery/puppetdb/metrics.go

@@ -0,0 +1,32 @@
+// Copyright 2015 The Prometheus Authors
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package puppetdb
+
+import (
+	"github.com/prometheus/prometheus/discovery"
+)
+
+var _ discovery.DiscovererMetrics = (*puppetdbMetrics)(nil)
+
+type puppetdbMetrics struct {
+	refreshMetrics discovery.RefreshMetricsInstantiator
+}
+
+// Register implements discovery.DiscovererMetrics.
+func (m *puppetdbMetrics) Register() error {
+	return nil
+}
+
+// Unregister implements discovery.DiscovererMetrics.
+func (m *puppetdbMetrics) Unregister() {}

discovery/puppetdb/puppetdb.go

@@ -79,12 +79,19 @@ type SDConfig struct {
 	Port              int                     `yaml:"port"`
 }
 
+// NewDiscovererMetrics implements discovery.Config.
+func (*SDConfig) NewDiscovererMetrics(reg prometheus.Registerer, rmi discovery.RefreshMetricsInstantiator) discovery.DiscovererMetrics {
+	return &puppetdbMetrics{
+		refreshMetrics: rmi,
+	}
+}
+
 // Name returns the name of the Config.
 func (*SDConfig) Name() string { return "puppetdb" }
 
 // NewDiscoverer returns a Discoverer for the Config.
 func (c *SDConfig) NewDiscoverer(opts discovery.DiscovererOptions) (discovery.Discoverer, error) {
-	return NewDiscovery(c, opts.Logger, opts.Registerer)
+	return NewDiscovery(c, opts.Logger, opts.Metrics)
 }
 
 // SetDirectory joins any relative file paths with dir.
@@ -131,7 +138,12 @@ type Discovery struct {
 }
 
 // NewDiscovery returns a new PuppetDB discovery for the given config.
-func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer) (*Discovery, error) {
+func NewDiscovery(conf *SDConfig, logger log.Logger, metrics discovery.DiscovererMetrics) (*Discovery, error) {
+	m, ok := metrics.(*puppetdbMetrics)
+	if !ok {
+		return nil, fmt.Errorf("invalid discovery metrics type")
+	}
+
 	if logger == nil {
 		logger = log.NewNopLogger()
 	}
@@ -159,10 +171,10 @@ func NewDiscovery(conf *SDConfig, logger log.Logger, reg prometheus.Registerer)
 	d.Discovery = refresh.NewDiscovery(
 		refresh.Options{
 			Logger:              logger,
-			Mech:     "http",
+			Mech:                "puppetdb",
 			Interval:            time.Duration(conf.RefreshInterval),
 			RefreshF:            d.refresh,
-			Registry: reg,
+			MetricsInstantiator: m.refreshMetrics,
 		},
 	)
 	return d, nil

discovery/puppetdb/puppetdb_test.go

@@ -28,6 +28,7 @@ import (
 	"github.com/prometheus/common/model"
 	"github.com/stretchr/testify/require"
 
+	"github.com/prometheus/prometheus/discovery"
 	"github.com/prometheus/prometheus/discovery/targetgroup"
 )
 
@@ -63,9 +64,17 @@ func TestPuppetSlashInURL(t *testing.T) {
 			Port:             80,
 			RefreshInterval:  model.Duration(30 * time.Second),
 		}
-		d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+
+		reg := prometheus.NewRegistry()
+		refreshMetrics := discovery.NewRefreshMetrics(reg)
+		metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+		require.NoError(t, metrics.Register())
+
+		d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 		require.NoError(t, err)
 		require.Equal(t, apiURL, d.url)
+
+		metrics.Unregister()
 	}
 }
 
@@ -80,7 +89,12 @@ func TestPuppetDBRefresh(t *testing.T) {
 		RefreshInterval:  model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
@@ -107,6 +121,8 @@ func TestPuppetDBRefresh(t *testing.T) {
 		},
 	}
 	require.Equal(t, expectedTargets, tgs)
+
+	metrics.Unregister()
 }
 
 func TestPuppetDBRefreshWithParameters(t *testing.T) {
@@ -121,7 +137,12 @@ func TestPuppetDBRefreshWithParameters(t *testing.T) {
 		RefreshInterval:   model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
@@ -158,6 +179,8 @@ func TestPuppetDBRefreshWithParameters(t *testing.T) {
 		},
 	}
 	require.Equal(t, expectedTargets, tgs)
+
+	metrics.Unregister()
 }
 
 func TestPuppetDBInvalidCode(t *testing.T) {
@@ -173,12 +196,19 @@ func TestPuppetDBInvalidCode(t *testing.T) {
 		RefreshInterval:  model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
 	_, err = d.refresh(ctx)
 	require.EqualError(t, err, "server returned HTTP status 400 Bad Request")
+
+	metrics.Unregister()
 }
 
 func TestPuppetDBInvalidFormat(t *testing.T) {
@@ -194,10 +224,17 @@ func TestPuppetDBInvalidFormat(t *testing.T) {
 		RefreshInterval:  model.Duration(30 * time.Second),
 	}
 
-	d, err := NewDiscovery(&cfg, log.NewNopLogger(), prometheus.NewRegistry())
+	reg := prometheus.NewRegistry()
+	refreshMetrics := discovery.NewRefreshMetrics(reg)
+	metrics := cfg.NewDiscovererMetrics(reg, refreshMetrics)
+	require.NoError(t, metrics.Register())
+
+	d, err := NewDiscovery(&cfg, log.NewNopLogger(), metrics)
 	require.NoError(t, err)
 
 	ctx := context.Background()
 	_, err = d.refresh(ctx)
 	require.EqualError(t, err, "unsupported content type text/plain; charset=utf-8")
+
+	metrics.Unregister()
 }