Skip to main content
basectl conductor and basectl sequencer help you operate a high-availability (HA) sequencer cluster. A cluster runs several sequencer nodes coordinated by op-conductor, which elects one leader at a time so sequencing keeps working if a node goes down. Both commands find the cluster’s nodes from your network config when it lists them. If it doesn’t, they discover the current members starting from the --conductor-rpc URL (or discovery.bootstrap_rpc in your config).
These commands act on live sequencing infrastructure. Run status first to confirm the current state, and read the safety notes below before making changes.

​
basectl conductor

Check and control the op-conductor layer that coordinates the cluster.
  • basectl conductor status shows each node’s cluster membership, which node is the leader, whether it’s paused, sequencer health, L1/L2 heads, and peer counts.
  • basectl conductor transfer-leader [TARGET] moves leadership away from the current leader, or to a specific node when you name one.
  • basectl conductor pause <NODE> pauses op-conductor on one node.
  • basectl conductor unpause <NODE> resumes op-conductor on one node.
  • basectl conductor pause-all pauses op-conductor on every node in the cluster.
  • basectl conductor unpause-all resumes op-conductor on every node in the cluster.
Commands that change the cluster also accept:

​
Safety notes

  • pause and unpause show the exact node name and conductor URL before acting.
  • transfer-leader shows the target node (or network) before acting.
  • pause-all and unpause-all ask you to type the network name to confirm, unless you pass --yes.
  • A cluster-wide action can succeed on some nodes and fail on others. The output lists which nodes succeeded and which failed, and the command exits with an error if any node failed.

​
Examples

Conductor commands

​
basectl sequencer

Check and control sequencing on the individual nodes in a cluster.
  • basectl sequencer status [NODE] shows whether each node is sequencing, plus its health, pause state, L1/L2 heads, and peer counts. Name a node to see just that one.
  • basectl sequencer start <NODE> [UNSAFE_HEAD] starts sequencing on one node.
  • basectl sequencer stop <NODE> stops sequencing on one node.
If you run start without UNSAFE_HEAD, basectl uses the node’s current unsafe head. This matches what the node requires: the head you start from has to match the node’s current unsafe head. Commands that change sequencing also accept:

​
Safety notes

  • start shows the node name, CL RPC URL, and the unsafe head hash before acting.
  • stop shows the node name and CL RPC URL before acting.
  • After start or stop, basectl checks the node’s actual sequencing state for up to 12 seconds before reporting success, so an accepted request isn’t mistaken for the node actually reaching the desired state.

​
Examples

Sequencer commands