diff --git a/cmd/frostfs-ir/config.go b/cmd/frostfs-ir/config.go
index 09af08525..19b7f05d6 100644
--- a/cmd/frostfs-ir/config.go
+++ b/cmd/frostfs-ir/config.go
@@ -9,6 +9,7 @@ import (
 	configViper "git.frostfs.info/TrueCloudLab/frostfs-node/cmd/internal/common/config"
 	"git.frostfs.info/TrueCloudLab/frostfs-node/internal/logs"
 	control "git.frostfs.info/TrueCloudLab/frostfs-node/pkg/services/control/ir"
+	"git.frostfs.info/TrueCloudLab/frostfs-node/pkg/util/logger"
 	"github.com/spf13/viper"
 	"go.uber.org/zap"
 )
@@ -38,13 +39,14 @@ func reloadConfig() error {
 	}
 	cmode.Store(cfg.GetBool("node.kludge_compatibility_mode"))
 	audit.Store(cfg.GetBool("audit.enabled"))
+	var logPrm logger.Prm
 	err = logPrm.SetLevelString(cfg.GetString("logger.level"))
 	if err != nil {
 		return err
 	}
-	logPrm.PrependTimestamp = cfg.GetBool("logger.timestamp")
+	log.Reload(logPrm)
 
-	return logPrm.Reload()
+	return nil
 }
 
 func watchForSignal(ctx context.Context, cancel func()) {
diff --git a/cmd/frostfs-ir/main.go b/cmd/frostfs-ir/main.go
index ade64ba84..114d8e4de 100644
--- a/cmd/frostfs-ir/main.go
+++ b/cmd/frostfs-ir/main.go
@@ -31,7 +31,6 @@ const (
 var (
 	wg         = new(sync.WaitGroup)
 	intErr     = make(chan error) // internal inner ring errors
-	logPrm     = new(logger.Prm)
 	innerRing  *innerring.Server
 	pprofCmp   *pprofComponent
 	metricsCmp *httpComponent
@@ -70,6 +69,7 @@ func main() {
 
 	metrics := irMetrics.NewInnerRingMetrics()
 
+	var logPrm logger.Prm
 	err = logPrm.SetLevelString(
 		cfg.GetString("logger.level"),
 	)
diff --git a/cmd/frostfs-node/config.go b/cmd/frostfs-node/config.go
index 2531e9173..c3c687763 100644
--- a/cmd/frostfs-node/config.go
+++ b/cmd/frostfs-node/config.go
@@ -473,7 +473,6 @@ type shared struct {
 // dynamicConfiguration stores parameters of the
 // components that supports runtime reconfigurations.
 type dynamicConfiguration struct {
-	logger  *logger.Prm
 	pprof   *httpComponent
 	metrics *httpComponent
 }
@@ -714,7 +713,8 @@ func initCfg(appCfg *config.Config) *cfg {
 
 	netState.metrics = c.metricsCollector
 
-	logPrm := c.loggerPrm()
+	logPrm, err := c.loggerPrm()
+	fatalOnErr(err)
 	logPrm.SamplingHook = c.metricsCollector.LogMetrics().GetSamplingHook()
 	log, err := logger.NewLogger(logPrm)
 	fatalOnErr(err)
@@ -1076,26 +1076,22 @@ func (c *cfg) getShardOpts(ctx context.Context, shCfg shardCfg) shardOptsWithID
 	return sh
 }
 
-func (c *cfg) loggerPrm() *logger.Prm {
-	// check if it has been inited before
-	if c.dynamicConfiguration.logger == nil {
-		c.dynamicConfiguration.logger = new(logger.Prm)
-	}
-
+func (c *cfg) loggerPrm() (logger.Prm, error) {
+	var prm logger.Prm
 	// (re)init read configuration
-	err := c.dynamicConfiguration.logger.SetLevelString(c.LoggerCfg.level)
+	err := prm.SetLevelString(c.LoggerCfg.level)
 	if err != nil {
 		// not expected since validation should be performed before
-		panic("incorrect log level format: " + c.LoggerCfg.level)
+		return logger.Prm{}, errors.New("incorrect log level format: " + c.LoggerCfg.level)
 	}
-	err = c.dynamicConfiguration.logger.SetDestination(c.LoggerCfg.destination)
+	err = prm.SetDestination(c.LoggerCfg.destination)
 	if err != nil {
 		// not expected since validation should be performed before
-		panic("incorrect log destination format: " + c.LoggerCfg.destination)
+		return logger.Prm{}, errors.New("incorrect log destination format: " + c.LoggerCfg.destination)
 	}
-	c.dynamicConfiguration.logger.PrependTimestamp = c.LoggerCfg.timestamp
+	prm.PrependTimestamp = c.LoggerCfg.timestamp
 
-	return c.dynamicConfiguration.logger
+	return prm, nil
 }
 
 func (c *cfg) LocalAddress() network.AddressGroup {
@@ -1335,11 +1331,7 @@ func (c *cfg) reloadConfig(ctx context.Context) {
 	// all the components are expected to support
 	// Logger's dynamic reconfiguration approach
 
-	// Logger
-
-	logPrm := c.loggerPrm()
-
-	components := c.getComponents(ctx, logPrm)
+	components := c.getComponents(ctx)
 
 	// Object
 	c.cfgObject.tombstoneLifetime.Store(c.ObjectCfg.tombstoneLifetime)
@@ -1377,10 +1369,17 @@ func (c *cfg) reloadConfig(ctx context.Context) {
 	c.log.Info(ctx, logs.FrostFSNodeConfigurationHasBeenReloadedSuccessfully)
 }
 
-func (c *cfg) getComponents(ctx context.Context, logPrm *logger.Prm) []dCmp {
+func (c *cfg) getComponents(ctx context.Context) []dCmp {
 	var components []dCmp
 
-	components = append(components, dCmp{"logger", logPrm.Reload})
+	components = append(components, dCmp{"logger", func() error {
+		prm, err := c.loggerPrm()
+		if err != nil {
+			return err
+		}
+		c.log.Reload(prm)
+		return nil
+	}})
 	components = append(components, dCmp{"runtime", func() error {
 		setRuntimeParameters(ctx, c)
 		return nil
diff --git a/pkg/services/object/common/writer/ec_test.go b/pkg/services/object/common/writer/ec_test.go
index 2458e352f..d5eeddf21 100644
--- a/pkg/services/object/common/writer/ec_test.go
+++ b/pkg/services/object/common/writer/ec_test.go
@@ -130,7 +130,7 @@ func TestECWriter(t *testing.T) {
 	nodeKey, err := keys.NewPrivateKey()
 	require.NoError(t, err)
 
-	log, err := logger.NewLogger(nil)
+	log, err := logger.NewLogger(logger.Prm{})
 	require.NoError(t, err)
 
 	var n nmKeys
diff --git a/pkg/util/logger/logger.go b/pkg/util/logger/logger.go
index 2eb5e5538..952a6f2dc 100644
--- a/pkg/util/logger/logger.go
+++ b/pkg/util/logger/logger.go
@@ -23,16 +23,8 @@ type Logger struct {
 // Parameters that have been connected to the Logger support its
 // configuration changing.
 //
-// Passing Prm after a successful connection via the NewLogger, connects
-// the Prm to a new instance of the Logger.
-//
-// See also Reload, SetLevelString.
+// See also Logger.Reload, SetLevelString.
 type Prm struct {
-	// link to the created Logger
-	// instance; used for a runtime
-	// reconfiguration
-	_log *Logger
-
 	// support runtime rereading
 	level zapcore.Level
 
@@ -73,22 +65,6 @@ func (p *Prm) SetDestination(d string) error {
 	return nil
 }
 
-// Reload reloads configuration of a connected instance of the Logger.
-// Returns ErrLoggerNotConnected if no connection has been performed.
-// Returns any reconfiguration error from the Logger directly.
-func (p Prm) Reload() error {
-	if p._log == nil {
-		// incorrect logger usage
-		panic("parameters are not connected to any Logger")
-	}
-
-	return p._log.reload(p)
-}
-
-func defaultPrm() *Prm {
-	return new(Prm)
-}
-
 // NewLogger constructs a new zap logger instance. Constructing with nil
 // parameters is safe: default values will be used then.
 // Passing non-nil parameters after a successful creation (non-error) allows
@@ -100,10 +76,7 @@ func defaultPrm() *Prm {
 //   - ISO8601 time encoding.
 //
 // Logger records a stack trace for all messages at or above fatal level.
-func NewLogger(prm *Prm) (*Logger, error) {
-	if prm == nil {
-		prm = defaultPrm()
-	}
+func NewLogger(prm Prm) (*Logger, error) {
 	switch prm.dest {
 	case DestinationUndefined, DestinationStdout:
 		return newConsoleLogger(prm)
@@ -114,7 +87,7 @@ func NewLogger(prm *Prm) (*Logger, error) {
 	}
 }
 
-func newConsoleLogger(prm *Prm) (*Logger, error) {
+func newConsoleLogger(prm Prm) (*Logger, error) {
 	lvl := zap.NewAtomicLevelAt(prm.level)
 
 	c := zap.NewProductionConfig()
@@ -139,12 +112,11 @@ func newConsoleLogger(prm *Prm) (*Logger, error) {
 	}
 
 	l := &Logger{z: lZap, lvl: lvl}
-	prm._log = l
 
 	return l, nil
 }
 
-func newJournaldLogger(prm *Prm) (*Logger, error) {
+func newJournaldLogger(prm Prm) (*Logger, error) {
 	lvl := zap.NewAtomicLevelAt(prm.level)
 
 	c := zap.NewProductionConfig()
@@ -181,14 +153,12 @@ func newJournaldLogger(prm *Prm) (*Logger, error) {
 	lZap := zap.New(samplingCore, zap.AddStacktrace(zap.NewAtomicLevelAt(zap.FatalLevel)), zap.AddCallerSkip(1))
 
 	l := &Logger{z: lZap, lvl: lvl}
-	prm._log = l
 
 	return l, nil
 }
 
-func (l *Logger) reload(prm Prm) error {
+func (l *Logger) Reload(prm Prm) {
 	l.lvl.SetLevel(prm.level)
-	return nil
 }
 
 func (l *Logger) WithOptions(options ...zap.Option) {