Documentation Index
Fetch the complete documentation index at: https://docs.optimism.io/llms.txt
Use this file to discover all available pages before exploring further.
The Kona node can operate in sequencer mode to build and produce new L2 blocks. In this mode, the node acts as the sequencer for an OP Stack rollup, building L2 blocks on top of the current unsafe head and extending the L2 chain.
:::info
Sequencer mode is an advanced configuration primarily used by rollup operators. Most users will run nodes in the default validator mode.
:::
Overview
When running in sequencer mode, the Kona node:
- Builds L2 blocks by collecting transactions from the mempool and constructing new blocks
- Selects L1 origins for new L2 blocks based on finalized L1 data
- Manages block production timing and ensures proper sequencing constraints
- Integrates with conductor services for leader election in multi-sequencer setups
- Handles recovery scenarios when the sequencer needs to catch up with L1
The sequencer uses the same core derivation pipeline as validator nodes but operates in reverse - instead of deriving L2 blocks from L1 data, it produces L2 blocks that will later be derivable from L1.
Trait Abstractions
Core Interfaces
The sequencer functionality is built around several key trait abstractions:
RollupNodeService
The main service trait that defines the node’s operational mode and actor types:
pub trait RollupNodeService {
type SequencerActor: NodeActor<
Error: Display,
OutboundData = SequencerContext,
Builder: AttributesBuilderConfig<AB = Self::AttributesBuilder>,
InboundData = SequencerInboundData,
>;
fn mode(&self) -> NodeMode;
// ... other methods
}
AttributesBuilderConfig
Configures how L2 block attributes are constructed:
pub trait AttributesBuilderConfig {
type AB: AttributesBuilder;
fn build(self) -> Self::AB;
}
SequencerActor
The core actor responsible for block production:
- Builds L2 blocks using the
AttributesBuilder
- Manages timing and L1 origin selection
- Handles admin RPC commands for sequencer control
- Coordinates with conductor services for leader election
Programmatic Configuration
Using the RollupNodeBuilder
To configure a Kona node programmatically for sequencer mode:
use kona_node_service::{NodeMode, RollupNode, SequencerConfig};
use url::Url;
// Configure sequencer settings
let sequencer_config = SequencerConfig {
sequencer_stopped: false, // Start sequencer immediately
sequencer_recovery_mode: false, // Normal operation mode
conductor_rpc_url: Some( // Optional conductor integration
Url::parse("http://conductor:8080").unwrap()
),
};
// Build and start the sequencer node
let node = RollupNode::builder(rollup_config)
.with_mode(NodeMode::Sequencer) // Enable sequencer mode
.with_sequencer_config(sequencer_config) // Apply sequencer settings
.with_l1_provider_rpc_url(l1_rpc_url) // L1 data source
.with_l2_engine_rpc_url(l2_engine_url) // L2 execution engine
.with_jwt_secret(jwt_secret) // Engine API authentication
// ... other configuration
.build()
.start()
.await?;
Configuration Options
| Field | Description | Default |
|---|
sequencer_stopped | Start sequencer in stopped state | false |
sequencer_recovery_mode | Enable recovery mode for catch-up | false |
conductor_rpc_url | Conductor service endpoint for leader election | None |
CLI Usage
Basic Sequencer Setup
To run a Kona node in sequencer mode:
kona-node node \
--mode=Sequencer \
--l1-eth-rpc=http://l1-node:8545 \
--l1-beacon=http://l1-beacon:5052 \
--l2-engine-rpc=http://l2-execution:8551 \
--l2.jwt-secret=./jwt.hex \
--chain=123456
Required Arguments
:::warning
Required Configuration
Sequencer mode requires all standard node arguments plus the --mode=Sequencer flag. Missing any required argument will prevent the node from starting.
:::
| Argument | Flag | Environment Variable | Description |
|---|
| Mode | --mode | KONA_NODE_MODE | Must be set to Sequencer |
| L1 RPC | --l1-eth-rpc | KONA_NODE_L1_ETH_RPC | L1 execution client RPC URL |
| L1 Beacon | --l1-beacon | KONA_NODE_L1_BEACON | L1 beacon API URL |
| L2 Engine | --l2-engine-rpc | KONA_NODE_L2_ENGINE_RPC | L2 engine API endpoint |
| JWT Secret | --l2.jwt-secret | KONA_NODE_L2_ENGINE_AUTH | Path to JWT secret file |
| Chain ID | --chain | KONA_NODE_L2_CHAIN_ID | L2 chain identifier |
Sequencer-Specific Flags
| Flag | Environment Variable | Default | Description |
|---|
--sequencer.stopped | KONA_NODE_SEQUENCER_STOPPED | false | Start sequencer in stopped state |
--sequencer.max-safe-lag | KONA_NODE_SEQUENCER_MAX_SAFE_LAG | 0 | Max L2 blocks between safe and unsafe heads |
--sequencer.l1-confs | KONA_NODE_SEQUENCER_L1_CONFS | 4 | L1 confirmations for origin selection |
--sequencer.recover | KONA_NODE_SEQUENCER_RECOVER | false | Force recovery mode operation |
--conductor.rpc | KONA_NODE_CONDUCTOR_RPC | - | Conductor service RPC endpoint |
--conductor.rpc.timeout | KONA_NODE_CONDUCTOR_RPC_TIMEOUT | 1 | Conductor RPC timeout (seconds) |
Example Configurations
Basic Sequencer
kona-node node \
--mode=Sequencer \
--l1-eth-rpc=http://localhost:8545 \
--l1-beacon=http://localhost:5052 \
--l2-engine-rpc=http://localhost:8551 \
--chain=42161
Sequencer with Conductor
kona-node node \
--mode=Sequencer \
--conductor.rpc=http://conductor:8080 \
--conductor.rpc.timeout=5 \
--sequencer.l1-confs=6 \
--l1-eth-rpc=http://l1-node:8545 \
--l1-beacon=http://l1-beacon:5052 \
--l2-engine-rpc=http://l2-execution:8551 \
--chain=123456
Recovery Mode Sequencer
kona-node node \
--mode=Sequencer \
--sequencer.recover=true \
--sequencer.max-safe-lag=100 \
--l1-eth-rpc=http://localhost:8545 \
--l1-beacon=http://localhost:5052 \
--l2-engine-rpc=http://localhost:8551 \
--chain=42161
Key Considerations
:::tip
Sequencer Operation
- L1 Confirmations: The
--sequencer.l1-confs setting determines how many L1 blocks the sequencer waits before using an L1 block as an origin. Higher values provide more safety but increase latency.
- Recovery Mode: Use
--sequencer.recover=true when the sequencer needs to catch up after being offline.
- Conductor Integration: For multi-sequencer deployments, configure the conductor service for proper leader election.
:::
Running a sequencer in production requires careful consideration of infrastructure, monitoring, and failover procedures. Ensure proper JWT secret management and secure network configuration.
