Merge branch 'main' into hackathon-xray-web-stream

.github/workflows/integration-tests.yml

@@ -16,7 +16,7 @@ jobs:
     - name: Setup Go
       uses: actions/setup-go@v5
       with:
-        go-version: "1.21"
+        go-version: "1.22"
     - name: Set OTEL Exporter Endpoint
       run: echo "OTEL_EXPORTER_ENDPOINT=172.17.0.1:4318" >> $GITHUB_ENV
     - name: Run tests

.github/workflows/test.yml

@@ -17,10 +17,10 @@ jobs:
     steps:
     - name: Checkout code
       uses: actions/checkout@v4
-    - name: Set up Go 1.21
+    - name: Set up Go 1.22
       uses: actions/setup-go@v5
       with:
-        go-version: "1.21"
+        go-version: "1.22"
         cache: true
     - name: Test
       run: make GO_TAGS="nodocker" test

CHANGELOG.md

@@ -10,23 +10,51 @@ internal API changes are not present.
 Main (unreleased)
 -----------------
 
-### Security fixes
+### Breaking changes
 
-- Fixes following vulnerabilities (@hainenber)
-  - [GO-2023-2409](https://github.com/advisories/GHSA-mhpq-9638-x6pw)
-  - [GO-2023-2412](https://github.com/advisories/GHSA-7ww5-4wqc-m92c)
-  - [CVE-2023-49568](https://github.com/advisories/GHSA-mw99-9chc-xw7r)
+- Prohibit the configuration of services within modules. (@wildum)
+
+- For `otelcol.exporter` components, change the default value of `disable_high_cardinality_metrics` to `true`. (@ptodev)
+
+### Features
+
+- A new `discovery.process` component for discovering Linux OS processes on the current host. (@korniltsev)
+
+- A new `pyroscope.java` component for profiling Java processes using async-profiler. (@korniltsev)
+
+- A new `otelcol.processor.resourcedetection` component which inserts resource attributes
+  to OTLP telemetry based on the host on which Grafana Agent is running. (@ptodev)
+
+- Expose track_timestamps_staleness on Prometheus scraping, to fix the issue where container metrics live for 5 minutes after the container disappears. (@ptodev)
 
 ### Enhancements
 
+- Include line numbers in profiles produced by `pyrsocope.java` component. (@korniltsev)
 - Add an option to the windows static mode installer for expanding environment vars in the yaml config. (@erikbaranowski)
 - Add authentication support to `loki.source.awsfirehose` (@sberz)
 
 - Sort kubelet endpoint to reduce pressure on K8s's API server and watcher endpoints. (@hainenber)
 
-- Expose `physical_disk` collector from `windows_exporter` v0.24.0 to 
+- Expose `physical_disk` collector from `windows_exporter` v0.24.0 to
   Flow configuration. (@hainenber)
 
+- Renamed Grafana Agent Mixin's "prometheus.remote_write" dashboard to
+  "Prometheus Components" and added charts for `prometheus.scrape` success rate
+  and duration metrics. (@thampiotr)
+
+- Removed `ClusterLamportClockDrift` and `ClusterLamportClockStuck` alerts from
+  Grafana Agent Mixin to focus on alerting on symptoms. (@thampiotr)
+
+- Increased clustering alert periods to 10 minutes to improve the
+  signal-to-noise ratio in Grafana Agent Mixin. (@thampiotr)
+
+- `mimir.rules.kubernetes` has a new `prometheus_http_prefix` argument to configure
+  the HTTP endpoint on which to connect to Mimir's API. (@hainenber)
+
+- `service_name` label is inferred from discovery meta labels in `pyroscope.java` (@korniltsev)
+
+- Mutex and block pprofs are now available via the pprof endpoint. (@mattdurham)
+
 ### Bugfixes
 
 - Fix an issue in `remote.s3` where the exported content of an object would be an empty string if `remote.s3` failed to fully retrieve
@@ -37,6 +65,16 @@ Main (unreleased)
 - Fix a duplicate metrics registration panic when sending metrics to an static
   mode metric instance's write handler. (@tpaschalis)
 
+- Fix issue causing duplicate logs when a docker target is restarted. (@captncraig)
+
+- Fix an issue where blocks having the same type and the same label across
+  modules could result in missed updates. (@thampiotr)
+
+- Fix an issue with static integrations-next marshaling where non singletons
+  would cause `/-/config` to fail to marshal. (@erikbaranowski)
+
+- Fix divide-by-zero issue when sharding targets. (@hainenber) 
+
 ### Other changes
 
 - Removed support for Windows 2012 in line with Microsoft end of life. (@mattdurham)
@@ -45,6 +83,34 @@ Main (unreleased)
 
 - Updated dependency to add support for Go 1.22 (@stefanb)
 
+- Use Go 1.22 for builds. (@rfratto)
+
+v0.39.2 (2024-1-31)
+--------------------
+
+### Bugfixes
+
+- Fix error introduced in v0.39.0 preventing remote write to Amazon Managed Prometheus. (@captncraig)
+
+- An error will be returned in the converter from Static to Flow when `scrape_integration` is set
+  to `true` but no `remote_write` is defined. (@erikbaranowski)
+
+
+v0.39.1 (2024-01-19)
+--------------------
+
+### Security fixes
+
+- Fixes following vulnerabilities (@hainenber)
+  - [GO-2023-2409](https://github.com/advisories/GHSA-mhpq-9638-x6pw)
+  - [GO-2023-2412](https://github.com/advisories/GHSA-7ww5-4wqc-m92c)
+  - [CVE-2023-49568](https://github.com/advisories/GHSA-mw99-9chc-xw7r)
+
+### Bugfixes
+
+- Fix issue where installing the Windows Agent Flow installer would hang then crash. (@mattdurham)
+
+
 v0.39.0 (2024-01-09)
 --------------------
 
@@ -63,7 +129,7 @@ v0.39.0 (2024-01-09)
   - This change will not break any existing configurations and you can opt in to validation via the `validate_dimensions` configuration option.
   - Before this change, pulling metrics for azure resources with variable dimensions required one configuration per metric + dimension combination to avoid an error.
   - After this change, you can include all metrics and dimensions in a single configuration and the Azure APIs will only return dimensions which are valid for the various metrics.
-  
+
 ### Features
 
 - A new `discovery.ovhcloud` component for discovering scrape targets on OVHcloud. (@ptodev)
@@ -160,7 +226,7 @@ v0.39.0 (2024-01-09)
 - Attach unique Agent ID header to remote-write requests. (@captncraig)
 
 - Update to v2.48.1 of `github.com/prometheus/prometheus`.
-  Previously, a custom fork of v2.47.2 was used. 
+  Previously, a custom fork of v2.47.2 was used.
   The custom fork of v2.47.2 also contained prometheus#12729 and prometheus#12677.
 
 v0.38.1 (2023-11-30)

Makefile

@@ -17,7 +17,7 @@
 ##
 ##   test              Run tests
 ##   lint              Lint code
-##   integration-tests Run integration tests
+##   integration-test Run integration tests
 ##
 ## Targets for building binaries:
 ##
@@ -167,7 +167,7 @@ test-packages:
 	docker pull $(BUILD_IMAGE)
 	go test -tags=packaging  ./packaging
 
-.PHONY: integration-tests
+.PHONY: integration-test
 integration-test:
 	cd integration-tests && $(GO_ENV) go run .
 

build-image/Dockerfile

@@ -23,7 +23,7 @@ FROM alpine:3.17 as helm
 RUN apk add --no-cache helm
 
 # Dependency: Go and Go dependencies
-FROM golang:1.21.4-bullseye as golang
+FROM golang:1.22.0-bullseye as golang
 
 # Keep in sync with cmd/grafana-agent-operator/DEVELOPERS.md
 ENV CONTROLLER_GEN_VERSION v0.9.2

build-image/windows/Dockerfile

@@ -1,4 +1,4 @@
-FROM library/golang:1.21.4-windowsservercore-1809
+FROM library/golang:1.22.0-windowsservercore-1809
 
 SHELL ["powershell", "-command"]
 

cmd/grafana-agent-operator/DEVELOPERS.md

@@ -74,7 +74,7 @@ running.
 ### Apply the CRDs
 
 Generated CRDs used by the operator can be found in [the Production
-folder](../../production/operator/crds). Deploy them from the root of the
+folder](../../operations/agent-static-operator/crds). Deploy them from the root of the
 repository with:
 
 ```

cmd/internal/flowmode/cmd_run.go

@@ -8,6 +8,8 @@ import (
 	"os"
 	"os/signal"
 	"path/filepath"
+	"runtime"
+	"strconv"
 	"strings"
 	"sync"
 	"syscall"
@@ -178,6 +180,9 @@ func (fr *flowRun) Run(configPath string) error {
 
 	level.Info(l).Log("boringcrypto enabled", boringcrypto.Enabled)
 
+	// Enable the profiling.
+	setMutexBlockProfiling(l)
+
 	// Immediately start the tracer.
 	go func() {
 		err := t.Run(ctx)
@@ -365,7 +370,7 @@ func getEnabledComponentsFunc(f *flow.Flow) func() map[string]interface{} {
 		components := component.GetAllComponents(f, component.InfoOptions{})
 		componentNames := map[string]struct{}{}
 		for _, c := range components {
-			componentNames[c.Registration.Name] = struct{}{}
+			componentNames[c.ComponentName] = struct{}{}
 		}
 		return map[string]interface{}{"enabled-components": maps.Keys(componentNames)}
 	}
@@ -455,3 +460,34 @@ func splitPeers(s, sep string) []string {
 	}
 	return strings.Split(s, sep)
 }
+
+func setMutexBlockProfiling(l log.Logger) {
+	mutexPercent := os.Getenv("PPROF_MUTEX_PROFILING_PERCENT")
+	if mutexPercent != "" {
+		rate, err := strconv.Atoi(mutexPercent)
+		if err == nil && rate > 0 {
+			// The 100/rate is because the value is interpreted as 1/rate. So 50 would be 100/50 = 2 and become 1/2 or 50%.
+			runtime.SetMutexProfileFraction(100 / rate)
+		} else {
+			level.Error(l).Log("msg", "error setting PPROF_MUTEX_PROFILING_PERCENT", "err", err, "value", mutexPercent)
+			runtime.SetMutexProfileFraction(1000)
+		}
+	} else {
+		// Why 1000 because that is what istio defaults to and that seemed reasonable to start with. This is 00.1% sampling.
+		runtime.SetMutexProfileFraction(1000)
+	}
+	blockRate := os.Getenv("PPROF_BLOCK_PROFILING_RATE")
+	if blockRate != "" {
+		rate, err := strconv.Atoi(blockRate)
+		if err == nil && rate > 0 {
+			runtime.SetBlockProfileRate(rate)
+		} else {
+			level.Error(l).Log("msg", "error setting PPROF_BLOCK_PROFILING_RATE", "err", err, "value", blockRate)
+			runtime.SetBlockProfileRate(10_000)
+		}
+	} else {
+		// This should have a negligible impact. This will track anything over 10_000ns, and will randomly sample shorter durations.
+		// Default taken from https://github.com/DataDog/go-profiler-notes/blob/main/block.md
+		runtime.SetBlockProfileRate(10_000)
+	}
+}

component/all/all.go

@@ -25,6 +25,7 @@ import (
 	_ "github.com/grafana/agent/component/discovery/nomad"                          // Import discovery.nomad
 	_ "github.com/grafana/agent/component/discovery/openstack"                      // Import discovery.openstack
 	_ "github.com/grafana/agent/component/discovery/ovhcloud"                       // Import discovery.ovhcloud
+	_ "github.com/grafana/agent/component/discovery/process"                        // Import discovery.process
 	_ "github.com/grafana/agent/component/discovery/puppetdb"                       // Import discovery.puppetdb
 	_ "github.com/grafana/agent/component/discovery/relabel"                        // Import discovery.relabel
 	_ "github.com/grafana/agent/component/discovery/scaleway"                       // Import discovery.scaleway
@@ -81,6 +82,7 @@ import (
 	_ "github.com/grafana/agent/component/otelcol/processor/k8sattributes"          // Import otelcol.processor.k8sattributes
 	_ "github.com/grafana/agent/component/otelcol/processor/memorylimiter"          // Import otelcol.processor.memory_limiter
 	_ "github.com/grafana/agent/component/otelcol/processor/probabilistic_sampler"  // Import otelcol.processor.probabilistic_sampler
+	_ "github.com/grafana/agent/component/otelcol/processor/resourcedetection"      // Import otelcol.processor.resourcedetection
 	_ "github.com/grafana/agent/component/otelcol/processor/span"                   // Import otelcol.processor.span
 	_ "github.com/grafana/agent/component/otelcol/processor/tail_sampling"          // Import otelcol.processor.tail_sampling
 	_ "github.com/grafana/agent/component/otelcol/processor/transform"              // Import otelcol.processor.transform
@@ -127,6 +129,7 @@ import (
 	_ "github.com/grafana/agent/component/prometheus/remotewrite"                   // Import prometheus.remote_write
 	_ "github.com/grafana/agent/component/prometheus/scrape"                        // Import prometheus.scrape
 	_ "github.com/grafana/agent/component/pyroscope/ebpf"                           // Import pyroscope.ebpf
+	_ "github.com/grafana/agent/component/pyroscope/java"                           // Import pyroscope.java
 	_ "github.com/grafana/agent/component/pyroscope/scrape"                         // Import pyroscope.scrape
 	_ "github.com/grafana/agent/component/pyroscope/write"                          // Import pyroscope.write
 	_ "github.com/grafana/agent/component/remote/http"                              // Import remote.http

component/component_provider.go

@@ -93,8 +93,8 @@ type Info struct {
 	// this component depends on, or is depended on by, respectively.
 	References, ReferencedBy []string
 
-	Registration Registration // Component registration.
-	Health       Health       // Current component health.
+	ComponentName string // Name of the component.
+	Health        Health // Current component health.
 
 	Arguments Arguments   // Current arguments value of the component.
 	Exports   Exports     // Current exports value of the component.
@@ -157,7 +157,7 @@ func (info *Info) MarshalJSON() ([]byte, error) {
 	}
 
 	return json.Marshal(&componentDetailJSON{
-		Name:         info.Registration.Name,
+		Name:         info.ComponentName,
 		Type:         "block",
 		ModuleID:     info.ID.ModuleID,
 		LocalID:      info.ID.LocalID,

component/discovery/discovery.go

@@ -45,7 +45,13 @@ func (t *DistributedTargets) Get() []Target {
 		return t.targets
 	}
 
-	res := make([]Target, 0, (len(t.targets)+1)/len(t.cluster.Peers()))
+	peerCount := len(t.cluster.Peers())
+	resCap := (len(t.targets) + 1)
+	if peerCount != 0 {
+		resCap = (len(t.targets) + 1) / peerCount
+	}
+
+	res := make([]Target, 0, resCap)
 
 	for _, tgt := range t.targets {
 		peers, err := t.cluster.Lookup(shard.StringKey(tgt.NonMetaLabels().String()), 1, shard.OpReadWrite)
@@ -55,7 +61,7 @@ func (t *DistributedTargets) Get() []Target {
 			// back to owning the target ourselves.
 			res = append(res, tgt)
 		}
-		if peers[0].Self {
+		if len(peers) == 0 || peers[0].Self {
 			res = append(res, tgt)
 		}
 	}

component/discovery/process/args.go

@@ -0,0 +1,37 @@
+package process
+
+import (
+	"time"
+
+	"github.com/grafana/agent/component/discovery"
+)
+
+type Arguments struct {
+	Join            []discovery.Target `river:"join,attr,optional"`
+	RefreshInterval time.Duration      `river:"refresh_interval,attr,optional"`
+	DiscoverConfig  DiscoverConfig     `river:"discover_config,block,optional"`
+}
+
+type DiscoverConfig struct {
+	Cwd         bool `river:"cwd,attr,optional"`
+	Exe         bool `river:"exe,attr,optional"`
+	Commandline bool `river:"commandline,attr,optional"`
+	Username    bool `river:"username,attr,optional"`
+	UID         bool `river:"uid,attr,optional"`
+	ContainerID bool `river:"container_id,attr,optional"`
+}
+
+var DefaultConfig = Arguments{
+	Join:            nil,
+	RefreshInterval: 60 * time.Second,
+	DiscoverConfig: DiscoverConfig{
+		Cwd:         true,
+		Exe:         true,
+		Commandline: true,
+		ContainerID: true,
+	},
+}
+
+func (args *Arguments) SetToDefault() {
+	*args = DefaultConfig
+}