metrics, pprof: support reloading services with SIGHUP

CHANGELOG.md

@@ -22,6 +22,7 @@ attribute, which is used for container domain name in NNS contracts (#2954)
 - New `peapod-to-fstree` tool providing peapod-to-fstree data migration (#3013)
 - Reloading node attributes with SIGHUP (#3005)
 - Reloading pool sizes (#3018)
+- Reloading pprof/metrics services with SIGHUP (#3016)
 
 ### Fixed
 - Do not search for tombstones when handling their expiration, use local indexes instead (#2929)

cmd/neofs-node/config.go

@@ -24,7 +24,6 @@
 	shardconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/engine/shard"
 	fstreeconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/engine/shard/blobstor/fstree"
 	loggerconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/logger"
-	metricsconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/metrics"
 	morphconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/morph"
 	nodeconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/node"
 	objectconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/object"
@@ -73,6 +72,11 @@
 // for each contract listener.
 const notificationHandlerPoolSize = 10
 
+const (
+	metricName   = "prometheus"
+	profilerName = "pprof"
+)
+
 // applicationConfiguration reads and stores component-specific configuration
 // values. It should not store any application helpers structs (pointers to shared
 // structs).
@@ -291,7 +295,8 @@
 	closers []func()
 	// services that are useful for debug (e.g. when a regular closer does not
 	// close), must be close at the very end of application life cycle
-	veryLastClosers []func()
+	veryLastClosersLock sync.RWMutex
+	veryLastClosers     map[string]func()
 
 	apiVersion   version.Version
 	healthStatus atomic.Int32
@@ -639,10 +644,10 @@
 
 	c.ownerIDFromKey = user.NewFromECDSAPublicKey(key.PrivateKey.PublicKey)
 
-	if metricsconfig.Enabled(c.cfgReader) {
-		c.metricsCollector = metrics.NewNodeMetrics(misc.Version)
-		c.basics.networkState.metrics = c.metricsCollector
-	}
+	c.metricsCollector = metrics.NewNodeMetrics(misc.Version)
+	c.basics.networkState.metrics = c.metricsCollector
+
+	c.veryLastClosers = make(map[string]func())
 
 	c.onShutdown(c.clientCache.CloseAll)    // clean up connections
 	c.onShutdown(c.bgClientCache.CloseAll)  // clean up connections
@@ -867,6 +872,9 @@
 		case <-ch:
 			c.log.Info("SIGHUP has been received, rereading configuration...")
 
+			oldMetrics := writeMetricConfig(c.cfgReader)
+			oldProfiler := writeProfilerConfig(c.cfgReader)
+
 			err := c.readConfig(c.cfgReader)
 			if err != nil {
 				c.log.Error("configuration reading", zap.Error(err))
@@ -877,6 +885,11 @@
 
 			c.reloadObjectPoolSizes()
 
+			// Prometheus and pprof
+
+			// nolint:contextcheck
+			c.reloadMetricsAndPprof(oldMetrics, oldProfiler)
+
 			// Logger
 
 			err = c.internals.logLevel.UnmarshalText([]byte(c.logger.level))
@@ -961,3 +974,30 @@
 
 	return nil
 }
+
+func (c *cfg) reloadMetricsAndPprof(oldMetrics metricConfig, oldProfiler profilerConfig) {
+	c.veryLastClosersLock.Lock()
+	defer c.veryLastClosersLock.Unlock()
+
+	// Metrics
+
+	if oldMetrics.isUpdated(c.cfgReader) {
+		if closer, ok := c.veryLastClosers[metricName]; ok {
+			closer()
+		}
+		delete(c.veryLastClosers, metricName)
+
+		preRunAndLog(c, metricName, initMetrics(c))
+	}
+
+	//Profiler
+
+	if oldProfiler.isUpdated(c.cfgReader) {
+		if closer, ok := c.veryLastClosers[profilerName]; ok {
+			closer()
+		}
+		delete(c.veryLastClosers, profilerName)
+
+		preRunAndLog(c, profilerName, initProfiler(c))
+	}
+}

cmd/neofs-node/control.go

@@ -67,9 +67,7 @@
 func (c *cfg) setHealthStatus(st control.HealthStatus) {
 	c.healthStatus.Store(int32(st))
 
-	if c.metricsCollector != nil {
-		c.metricsCollector.SetHealth(int32(st))
-	}
+	c.metricsCollector.SetHealth(int32(st))
 }
 
 func (c *cfg) HealthStatus() control.HealthStatus {

cmd/neofs-node/main.go

@@ -55,9 +55,9 @@
 
 	c := initCfg(appCfg)
 
-	preRunAndLog(c, "prometheus", initMetrics(c))
+	preRunAndLog(c, metricName, initMetrics(c))
 
-	preRunAndLog(c, "pprof", initProfiler(c))
+	preRunAndLog(c, profilerName, initProfiler(c))
 
 	initApp(c)
 
@@ -91,13 +91,13 @@
 	c.log.Info(fmt.Sprintf("%s service is initialized", name))
 	c.wg.Add(1)
 	go func() {
-		runAndLog(c, name, true, func(c *cfg) {
+		runAndLog(c, name, false, func(c *cfg) {
 			fatalOnErr(srv.Serve(ln))
 			c.wg.Done()
 		})
 	}()
 
-	c.veryLastClosers = append(c.veryLastClosers, func() {
+	c.veryLastClosers[name] = func() {
 		c.log.Debug(fmt.Sprintf("shutting down %s service", name))
 
 		err := srv.Shutdown()
@@ -108,7 +108,7 @@
 		}
 
 		c.log.Debug(fmt.Sprintf("%s service has been stopped", name))
-	})
+	}
 }
 
 func initAndLog(c *cfg, name string, initializer func(*cfg)) {
@@ -184,9 +184,12 @@
 	for _, closer := range c.closers {
 		closer()
 	}
+
+	c.veryLastClosersLock.RLock()
 	for _, lastCloser := range c.veryLastClosers {
 		lastCloser()
 	}
+	c.veryLastClosersLock.RUnlock()
 
 	c.log.Debug("waiting for all processes to stop")
 

cmd/neofs-node/metrics.go

@@ -1,6 +1,9 @@
 package main
 
 import (
+	"time"
+
+	"github.com/nspcc-dev/neofs-node/cmd/neofs-node/config"
 	metricsconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/metrics"
 	httputil "github.com/nspcc-dev/neofs-node/pkg/util/http"
 	"github.com/prometheus/client_golang/prometheus/promhttp"
@@ -25,3 +28,23 @@
 
 	return srv
 }
+
+type metricConfig struct {
+	enabled         bool
+	shutdownTimeout time.Duration
+	address         string
+}
+
+func writeMetricConfig(c *config.Config) metricConfig {
+	return metricConfig{
+		enabled:         metricsconfig.Enabled(c),
+		shutdownTimeout: metricsconfig.ShutdownTimeout(c),
+		address:         metricsconfig.Address(c),
+	}
+}
+
+func (m1 metricConfig) isUpdated(c *config.Config) bool {
+	return m1.enabled != metricsconfig.Enabled(c) ||
+		m1.shutdownTimeout != metricsconfig.ShutdownTimeout(c) ||
+		m1.address != metricsconfig.Address(c)
+}

cmd/neofs-node/netmap.go

@@ -48,9 +48,8 @@
 
 func (s *networkState) setCurrentEpoch(v uint64) {
 	s.epoch.Store(v)
-	if s.metrics != nil {
-		s.metrics.SetEpoch(v)
-	}
+
+	s.metrics.SetEpoch(v)
 }
 
 func (s *networkState) setNodeInfo(ni *netmapSDK.NodeInfo) {

cmd/neofs-node/object.go

@@ -347,10 +347,7 @@
 		respSvc,
 	)
 
-	var firstSvc objectService.ServiceServer = signSvc
-	if c.metricsCollector != nil {
-		firstSvc = objectService.NewMetricCollector(signSvc, c.metricsCollector)
-	}
+	firstSvc := objectService.NewMetricCollector(signSvc, c.metricsCollector)
 
 	server := objectTransportGRPC.New(firstSvc, mNumber, objNode, neofsecdsa.SignerRFC6979(c.shared.basics.key.PrivateKey))
 

cmd/neofs-node/pprof.go

@@ -1,6 +1,9 @@
 package main
 
 import (
+	"time"
+
+	"github.com/nspcc-dev/neofs-node/cmd/neofs-node/config"
 	profilerconfig "github.com/nspcc-dev/neofs-node/cmd/neofs-node/config/profiler"
 	httputil "github.com/nspcc-dev/neofs-node/pkg/util/http"
 )
@@ -24,3 +27,23 @@
 
 	return srv
 }
+
+type profilerConfig struct {
+	enabled         bool
+	shutdownTimeout time.Duration
+	address         string
+}
+
+func writeProfilerConfig(c *config.Config) profilerConfig {
+	return profilerConfig{
+		enabled:         profilerconfig.Enabled(c),
+		shutdownTimeout: profilerconfig.ShutdownTimeout(c),
+		address:         profilerconfig.Address(c),
+	}
+}
+
+func (m1 profilerConfig) isUpdated(c *config.Config) bool {
+	return m1.enabled != profilerconfig.Enabled(c) ||
+		m1.shutdownTimeout != profilerconfig.ShutdownTimeout(c) ||
+		m1.address != profilerconfig.Address(c)
+}

cmd/neofs-node/storage.go

@@ -93,9 +93,7 @@
 		opts = append(opts, engine.WithContainersSource(cntClient.AsContainerSource(c.shared.basics.cCli)))
 	}
 
-	if c.metricsCollector != nil {
-		opts = append(opts, engine.WithMetrics(c.metricsCollector))
-	}
+	opts = append(opts, engine.WithMetrics(c.metricsCollector))
 
 	return opts
 }