Merge pull request #7111 from ethereum-optimism/runtime-config-reloading

op-node: implement runtime config reloading

op-e2e/setup.go

@@ -120,8 +120,9 @@ func DefaultSystemConfig(t *testing.T) SystemConfig {
 					ListenPort:  0,
 					EnableAdmin: true,
 				},
-				L1EpochPollInterval: time.Second * 2,
-				ConfigPersistence:   &rollupNode.DisabledConfigPersistence{},
+				L1EpochPollInterval:         time.Second * 2,
+				RuntimeConfigReloadInterval: time.Minute * 10,
+				ConfigPersistence:           &rollupNode.DisabledConfigPersistence{},
 			},
 			"verifier": {
 				Driver: driver.Config{
@@ -129,8 +130,9 @@ func DefaultSystemConfig(t *testing.T) SystemConfig {
 					SequencerConfDepth: 0,
 					SequencerEnabled:   false,
 				},
-				L1EpochPollInterval: time.Second * 4,
-				ConfigPersistence:   &rollupNode.DisabledConfigPersistence{},
+				L1EpochPollInterval:         time.Second * 4,
+				RuntimeConfigReloadInterval: time.Minute * 10,
+				ConfigPersistence:           &rollupNode.DisabledConfigPersistence{},
 			},
 		},
 		Loggers: map[string]log.Logger{

op-e2e/system_test.go

@@ -30,6 +30,7 @@ import (
 	"github.com/ethereum-optimism/optimism/op-bindings/bindings"
 	"github.com/ethereum-optimism/optimism/op-bindings/predeploys"
 	"github.com/ethereum-optimism/optimism/op-e2e/config"
+	"github.com/ethereum-optimism/optimism/op-e2e/e2eutils/wait"
 	"github.com/ethereum-optimism/optimism/op-node/client"
 	"github.com/ethereum-optimism/optimism/op-node/metrics"
 	rollupNode "github.com/ethereum-optimism/optimism/op-node/node"
@@ -1389,3 +1390,47 @@ func TestPendingBlockIsLatest(t *testing.T) {
 		t.Fatal("failed to get pending header with same number as latest header")
 	})
 }
+
+func TestRuntimeConfigReload(t *testing.T) {
+	InitParallel(t)
+
+	cfg := DefaultSystemConfig(t)
+	// to speed up the test, make it reload the config more often, and do not impose a long conf depth
+	cfg.Nodes["verifier"].RuntimeConfigReloadInterval = time.Second * 5
+	cfg.Nodes["verifier"].Driver.VerifierConfDepth = 1
+
+	sys, err := cfg.Start(t)
+	require.Nil(t, err, "Error starting up system")
+	defer sys.Close()
+	initialRuntimeConfig := sys.RollupNodes["verifier"].RuntimeConfig()
+
+	// close the EL node, since we want to block derivation, to solely rely on the reloading mechanism for updates.
+	sys.EthInstances["verifier"].Close()
+
+	l1 := sys.Clients["l1"]
+
+	// Change the system-config via L1
+	sysCfgContract, err := bindings.NewSystemConfig(cfg.L1Deployments.SystemConfigProxy, l1)
+	require.NoError(t, err)
+	newUnsafeBlocksSigner := common.Address{0x12, 0x23, 0x45}
+	require.NotEqual(t, initialRuntimeConfig.P2PSequencerAddress(), newUnsafeBlocksSigner, "changing to a different address")
+	opts, err := bind.NewKeyedTransactorWithChainID(cfg.Secrets.SysCfgOwner, cfg.L1ChainIDBig())
+	require.Nil(t, err)
+	// the unsafe signer address is part of the runtime config
+	tx, err := sysCfgContract.SetUnsafeBlockSigner(opts, newUnsafeBlocksSigner)
+	require.NoError(t, err)
+
+	// wait for the change to confirm
+	_, err = wait.ForReceiptOK(context.Background(), l1, tx.Hash())
+	require.NoError(t, err)
+
+	// wait for the address to change
+	_, err = retry.Do(context.Background(), 10, retry.Fixed(time.Second*10), func() (struct{}, error) {
+		v := sys.RollupNodes["verifier"].RuntimeConfig().P2PSequencerAddress()
+		if v == newUnsafeBlocksSigner {
+			return struct{}{}, nil
+		}
+		return struct{}{}, fmt.Errorf("no change yet, seeing %s but looking for %s", v, newUnsafeBlocksSigner)
+	})
+	require.NoError(t, err)
+}

op-node/flags/flags.go

@@ -146,6 +146,13 @@ var (
 		Required: false,
 		Value:    time.Second * 12 * 32,
 	}
+	RuntimeConfigReloadIntervalFlag = &cli.DurationFlag{
+		Name:     "l1.runtime-config-reload-interval",
+		Usage:    "Poll interval for reloading the runtime config, useful when config events are not being picked up. Disabled if 0 or negative.",
+		EnvVars:  prefixEnvVars("L1_RUNTIME_CONFIG_RELOAD_INTERVAL"),
+		Required: false,
+		Value:    time.Minute * 10,
+	}
 	MetricsEnabledFlag = &cli.BoolFlag{
 		Name:    "metrics.enabled",
 		Usage:   "Enable the metrics server",
@@ -261,6 +268,7 @@ var optionalFlags = []cli.Flag{
 	SequencerMaxSafeLagFlag,
 	SequencerL1Confs,
 	L1EpochPollIntervalFlag,
+	RuntimeConfigReloadIntervalFlag,
 	RPCEnableAdmin,
 	RPCAdminPersistence,
 	MetricsEnabledFlag,

op-node/node/config.go

@@ -41,6 +41,12 @@ type Config struct {
 
 	ConfigPersistence ConfigPersistence
 
+	// RuntimeConfigReloadInterval defines the interval between runtime config reloads.
+	// Disabled if 0.
+	// Runtime config changes should be picked up from log-events,
+	// but if log-events are not coming in (e.g. not syncing blocks) then the reload ensures the config stays accurate.
+	RuntimeConfigReloadInterval time.Duration
+
 	// Optional
 	Tracer    Tracer
 	Heartbeat HeartbeatConfig

op-node/node/node.go

@@ -2,7 +2,6 @@ package node
 
 import (
 	"context"
-	"errors"
 	"fmt"
 	"time"
 
@@ -19,6 +18,7 @@ import (
 	"github.com/ethereum-optimism/optimism/op-node/rollup/driver"
 	"github.com/ethereum-optimism/optimism/op-node/sources"
 	"github.com/ethereum-optimism/optimism/op-service/eth"
+	"github.com/ethereum-optimism/optimism/op-service/retry"
 )
 
 type OpNode struct {
@@ -159,27 +159,70 @@ func (n *OpNode) initRuntimeConfig(ctx context.Context, cfg *Config) error {
 	// attempt to load runtime config, repeat N times
 	n.runCfg = NewRuntimeConfig(n.log, n.l1Source, &cfg.Rollup)
 
-	for i := 0; i < 5; i++ {
+	confDepth := cfg.Driver.VerifierConfDepth
+	reload := func(ctx context.Context) (eth.L1BlockRef, error) {
 		fetchCtx, fetchCancel := context.WithTimeout(ctx, time.Second*10)
 		l1Head, err := n.l1Source.L1BlockRefByLabel(fetchCtx, eth.Unsafe)
 		fetchCancel()
 		if err != nil {
 			n.log.Error("failed to fetch L1 head for runtime config initialization", "err", err)
-			continue
+			return eth.L1BlockRef{}, err
+		}
+
+		// Apply confirmation-distance
+		blNum := l1Head.Number
+		if blNum >= confDepth {
+			blNum -= confDepth
+		}
+		fetchCtx, fetchCancel = context.WithTimeout(ctx, time.Second*10)
+		confirmed, err := n.l1Source.L1BlockRefByNumber(fetchCtx, blNum)
+		fetchCancel()
+		if err != nil {
+			n.log.Error("failed to fetch confirmed L1 block for runtime config loading", "err", err, "number", blNum)
+			return eth.L1BlockRef{}, err
 		}
 
 		fetchCtx, fetchCancel = context.WithTimeout(ctx, time.Second*10)
-		err = n.runCfg.Load(fetchCtx, l1Head)
+		err = n.runCfg.Load(fetchCtx, confirmed)
 		fetchCancel()
 		if err != nil {
 			n.log.Error("failed to fetch runtime config data", "err", err)
-			continue
+			return l1Head, err
 		}
+		return l1Head, nil
+	}
 
-		return nil
+	// initialize the runtime config before unblocking
+	if _, err := retry.Do(ctx, 5, retry.Fixed(time.Second*10), func() (eth.L1BlockRef, error) {
+		return reload(ctx)
+	}); err != nil {
+		return fmt.Errorf("failed to load runtime configuration repeatedly, last error: %w", err)
 	}
 
-	return errors.New("failed to load runtime configuration repeatedly")
+	// start a background loop, to keep reloading it at the configured reload interval
+	go func(ctx context.Context, reloadInterval time.Duration) {
+		if reloadInterval <= 0 {
+			n.log.Debug("not running runtime-config reloading background loop")
+			return
+		}
+		ticker := time.NewTicker(reloadInterval)
+		defer ticker.Stop()
+		for {
+			select {
+			case <-ticker.C:
+				// If the reload fails, we will try again the next interval.
+				// Missing a runtime-config update is not critical, and we do not want to overwhelm the L1 RPC.
+				if l1Head, err := reload(ctx); err != nil {
+					n.log.Warn("failed to reload runtime config", "err", err)
+				} else {
+					n.log.Debug("reloaded runtime config", "l1_head", l1Head)
+				}
+			case <-ctx.Done():
+				return
+			}
+		}
+	}(n.resourcesCtx, cfg.RuntimeConfigReloadInterval) // this keeps running after initialization
+	return nil
 }
 
 func (n *OpNode) initL2(ctx context.Context, cfg *Config, snapshotLog log.Logger) error {
@@ -397,6 +440,10 @@ func (n *OpNode) P2P() p2p.Node {
 	return n.p2pNode
 }
 
+func (n *OpNode) RuntimeConfig() ReadonlyRuntimeConfig {
+	return n.runCfg
+}
+
 // Close closes all resources.
 func (n *OpNode) Close() error {
 	var result *multierror.Error

op-node/node/runtime_config.go

@@ -23,6 +23,10 @@ type RuntimeCfgL1Source interface {
 	ReadStorageAt(ctx context.Context, address common.Address, storageSlot common.Hash, blockHash common.Hash) (common.Hash, error)
 }
 
+type ReadonlyRuntimeConfig interface {
+	P2PSequencerAddress() common.Address
+}
+
 // RuntimeConfig maintains runtime-configurable options.
 // These options are loaded based on initial loading + updates for every subsequent L1 block.
 // Only the *latest* values are maintained however, the runtime config has no concept of chain history,

op-node/service.go

@@ -82,9 +82,10 @@ func NewConfig(ctx *cli.Context, log log.Logger) (*node.Config, error) {
 			ListenAddr: ctx.String(flags.PprofAddrFlag.Name),
 			ListenPort: ctx.Int(flags.PprofPortFlag.Name),
 		},
-		P2P:                 p2pConfig,
-		P2PSigner:           p2pSignerSetup,
-		L1EpochPollInterval: ctx.Duration(flags.L1EpochPollIntervalFlag.Name),
+		P2P:                         p2pConfig,
+		P2PSigner:                   p2pSignerSetup,
+		L1EpochPollInterval:         ctx.Duration(flags.L1EpochPollIntervalFlag.Name),
+		RuntimeConfigReloadInterval: ctx.Duration(flags.RuntimeConfigReloadIntervalFlag.Name),
 		Heartbeat: node.HeartbeatConfig{
 			Enabled: ctx.Bool(flags.HeartbeatEnabledFlag.Name),
 			Moniker: ctx.String(flags.HeartbeatMonikerFlag.Name),