> ## 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.

# op-conductor

> Canonical hub for op-conductor, the high-availability sequencer coordination service, covering what it does, who runs it, and where its guides, reference, releases, and source live.

op-conductor is the OP Stack's high-availability sequencer coordination
service: it manages a cluster of sequencers so that exactly one leads the
chain at a time and block production survives single-node failures.

## Overview

op-conductor runs alongside each sequencer in a multi-node cluster, typically
three nodes spread across regions or availability zones. It participates in a
Raft consensus layer to elect the leader and to store the latest unsafe
block, monitors its sequencer's health, starts or stops sequencing based on
leadership and health, and serves admin RPCs for manual recovery scenarios.
The design provides three guarantees: no unsafe reorgs, no unsafe head stall
during a network partition, and continued uptime with no more than one node
failure in a standard three-node setup. For how this works in detail, see the
[OP Conductor explainer](/chain-operators/tools/op-conductor).

It exists because a single sequencer is a single point of failure: without
coordination, a crashed sequencer halts block production, and a naive
failover can produce two active sequencers or reorg unsafe blocks.
op-conductor is not Byzantine fault tolerant: it assumes all participating
nodes are honest, performs no authentication or authorization of RPC
requests, and is intended to run inside a private network.

**Who runs it:** chain operators running a high-availability sequencer setup,
with one conductor instance per sequencer node. The rollup node's sequencer
mode integrates with it on both op-node (`--conductor.enabled` and
`--conductor.rpc`) and kona-node (`--conductor.rpc`). Chains running a single
sequencer do not need it.

## Get started

* [Setup](/chain-operators/tools/op-conductor/setup): add op-conductor to
  an existing multi-sequencer OP Stack network without downtime.

## How-tos

* [Launch a chain with fault proofs and HA sequencing](/use-cases/launch-a-chain-with-fault-proofs-and-ha-sequencing):
  the whole production path in one pass, from contract deployment through
  conductor cluster setup and failover drills.
* [Network architecture](/chain-operators/guides/management/network-architecture):
  where op-conductor sits in a production topology, including its use as a
  leader-aware RPC proxy for op-batcher.

## Configuration & flags reference

* [Configuration and RPCs](/chain-operators/tools/op-conductor/reference):
  every CLI flag and environment variable, with flag tables generated from
  the op-conductor release named on the page, plus the `conductor`
  namespace RPC methods for cluster management.

## Releases

* [op-conductor release history](/releases/op-conductor)

## Source & spec links

* [Source: `op-conductor/`](https://github.com/ethereum-optimism/optimism/tree/develop/op-conductor)
  in the Optimism monorepo; its
  [README](https://github.com/ethereum-optimism/optimism/blob/develop/op-conductor/README.md)
  documents the architecture, the conductor state transitions, and
  walkthroughs of each failure scenario.
* op-conductor is operational tooling and has no section of its own in the
  OP Stack specifications; the sequencing role it coordinates is specified
  in the
  [rollup node specification](https://specs.optimism.io/protocol/rollup-node.html).