Running an Axelar Node: Start, Sync & Operate

Once your node is built and configured, running it is a matter of invoking axelard start. The daemon embeds CometBFT in-process and begins syncing the chain state from either genesis or a state sync snapshot. This page covers the start command and its most important flags, available monitoring endpoints, snapshot management, and the operational commands you’ll reach for most often.

Starting the Node

axelard start --home ~/.axelar

The process is long-running. It will block and emit structured log lines to stdout. Use a process manager (systemd, supervisord, or Docker) to keep it running across reboots.

Key Flags

--home

string

Path to the node home directory. Defaults to $HOME/.axelar. Override this when running multiple nodes on the same host.

--log_level

string

Log verbosity: trace, debug, info, warn, error, fatal, or panic. Defaults to info. Use debug when diagnosing issues.

--log_format

string

plain (default, human-readable) or json (structured, better for log aggregators like Loki or Datadog).

--p2p.laddr

string

P2P listen address. Default tcp://0.0.0.0:26656. Change the port if it conflicts with another service.

--rpc.laddr

string

CometBFT RPC listen address. Default tcp://127.0.0.1:26657. Bind to 127.0.0.1 unless you intentionally expose the RPC.

--minimum-gas-prices

string

Override app.toml minimum gas prices at runtime. Example: 0.007uaxl.

--pruning

string

Override the pruning strategy. One of default, nothing, everything, or custom.

--p2p.seeds

string

Comma-separated list of ID@host:port seed nodes passed directly without editing config.toml.

--halt-height

uint

Stop the node gracefully at a specific block height. Useful for planned upgrades.

--grpc.enable

bool

Enable or disable the gRPC server. Defaults to true.

Full Example

axelard start \
  --home ~/.axelar \
  --log_level info \
  --log_format json \
  --p2p.laddr tcp://0.0.0.0:26656 \
  --rpc.laddr tcp://127.0.0.1:26657 \
  --minimum-gas-prices 0.007uaxl \
  --grpc.address localhost:9090 \
  --api.enable \
  --api.address tcp://localhost:1317

Syncing Strategies

Genesis Sync
State Sync

Replay every block from block 1. This is the most trustless option but can take days for a mature chain.

# Ensure state sync is disabled in config.toml
axelard config set config statesync.enable false
axelard start --home ~/.axelar

You can monitor progress by querying the local RPC:

curl -s http://localhost:26657/status | jq '.result.sync_info'

Download a recent state snapshot from trusted peers and only replay blocks from that height forward. Much faster than genesis sync.

Find a trusted block

curl -s https://axelar-rpc.polkachu.com/block \
  | jq '.result.block.header | {height, hash: .last_block_id.hash}'

Configure state sync

axelard config set config statesync.enable true
axelard config set config statesync.rpc_servers \
  "https://axelar-rpc.polkachu.com:443,https://axelar-rpc.polkachu.com:443"
axelard config set config statesync.trust_height 12345678
axelard config set config statesync.trust_hash "ABCDEF..."

Start the node

axelard start --home ~/.axelar

Health Check

The axelard health-check command validates that the node, the broadcaster account, and the optional tofnd daemon are all reachable and functional:

axelard health-check \
  --node tcp://localhost:26657 \
  --operator-addr axelarvaloper1...

Flags

--node

string

CometBFT RPC endpoint to check. Defaults to tcp://localhost:26657.

--operator-addr

string

Validator operator address (axelarvaloper1...) to check on-chain registration.

--tofnd-host

string

Hostname of the tofnd signing daemon. Defaults to localhost.

--tofnd-port

string

Port of the tofnd daemon. Defaults to 50051.

--skip-broadcaster

bool

Skip the broadcaster account check.

--skip-tofnd

bool

Skip the tofnd connectivity check. Useful for non-validator full nodes.

Monitoring Endpoints

A running axelard node exposes several endpoints:

Endpoint	Default Address	Purpose
CometBFT RPC	`http://localhost:26657`	Block queries, transaction broadcast, node status
REST API	`http://localhost:1317`	Cosmos SDK REST / Swagger UI
gRPC	`localhost:9090`	Cosmos SDK gRPC queries and transactions
gRPC-Web	Proxied via gRPC port	Browser-compatible gRPC
Prometheus	`http://localhost:26660/metrics`	Node metrics for Grafana

Enable Prometheus

Prometheus metrics are emitted by CometBFT. Enable the endpoint in config.toml:

[instrumentation]
prometheus = true
prometheus_listen_addr = ":26660"
namespace = "cometbft"

Then restart the node and point your Prometheus scrape config at http://<node-ip>:26660/metrics.

Useful RPC Queries

# Node sync status
curl -s http://localhost:26657/status | jq '.result.sync_info'

# Latest block height
curl -s http://localhost:26657/abci_info | jq '.result.response.last_block_height'

# Node ID
curl -s http://localhost:26657/status | jq -r '.result.node_info.id'

# Connected peers
curl -s http://localhost:26657/net_info | jq '.result.n_peers'

Snapshot Management

Snapshots let you quickly restore a node without replaying the full chain history. The axelard snapshots family of commands manages the local snapshot store.

List Available Snapshots

axelard snapshots list --home ~/.axelar

Export a Snapshot

Take a snapshot of the current application state and save it to the snapshot store:

axelard snapshots export --home ~/.axelar

The node must not be running when you export a snapshot manually. For live exports, configure state-sync.snapshot-interval in app.toml and restart the node.

Load a Snapshot Archive

Load a portable .tar.gz archive into the local snapshot store:

axelard snapshots load snapshot.tar.gz --home ~/.axelar

Restore from a Snapshot

Apply a previously loaded snapshot to reset application state:

axelard snapshots restore <height> <format> --home ~/.axelar

Dump a Snapshot

Export a snapshot from the store as a portable archive:

axelard snapshots dump <height> <format> --home ~/.axelar

Operational Commands

Rolling Back One Block

If the node has diverged due to an incorrect app hash, roll back one block:

axelard rollback --home ~/.axelar

Use --hard to remove the last block entirely (not just revert state):

axelard rollback --hard --home ~/.axelar

Only use --hard if instructed by the Axelar team or if you fully understand the implications. The blocks are not removed from the chain — upon restart, transactions from the removed block will be re-executed.

Pruning Historical State

Manually prune old state without running the full node:

# Use the default strategy (keep last 362880 states)
axelard prune default --home ~/.axelar

# Keep only the most recent 100 states
axelard prune custom --pruning-keep-recent 100 --home ~/.axelar

# Archive: keep nothing (delete everything except the 2 latest)
axelard prune everything --home ~/.axelar

Resetting CometBFT State

This is a destructive operation that deletes all chain data. Use only on a node you intend to re-sync from scratch.

axelard comet unsafe-reset-all --home ~/.axelar

Running as a systemd Service

Create /etc/systemd/system/axelard.service:

[Unit]
Description=Axelar Core Node
After=network-online.target

[Service]
User=axelard
ExecStart=/usr/local/bin/axelard start --home /home/axelard/.axelar --log_format json
Restart=on-failure
RestartSec=5s
LimitNOFILE=65536

[Install]
WantedBy=multi-user.target

Enable and start:

sudo systemctl daemon-reload
sudo systemctl enable axelard
sudo systemctl start axelard
journalctl -fu axelard

Running with Docker

The official Docker image sets up the home directory at /home/axelard/.axelar and reads several environment variables:

Variable	Purpose
`HOME_DIR`	Base directory (default `/home/axelard`)
`TOFND_HOST`	Hostname of the tofnd signing daemon
`AXELARD_CHAIN_ID`	Chain ID (default `axelar-testnet-lisbon-3`)
`AXELARD_KEYRING_BACKEND`	Keyring backend type (default `file`)
`PEERS_FILE`	Path to a file containing comma-separated seed peers
`CONFIG_PATH`	Directory with custom `config.toml`, `app.toml`, and `vald.toml`
`PRESTART_SCRIPT`	Path to a shell script that runs before node startup
`NODE_MONIKER`	Node moniker

docker run -d \
  --name axelard \
  -e AXELARD_CHAIN_ID=axelar-dojo-1 \
  -e NODE_MONIKER=my-node \
  -v ~/.axelar:/home/axelard/.axelar \
  -p 26656:26656 \
  -p 26657:26657 \
  axelar/core

Get Started

Running a Node

Core Modules

Governance & Operations

Running an Axelar Node: Start, Sync & Operate

Starting the Node

Key Flags

Full Example

Syncing Strategies

Health Check

Flags

Monitoring Endpoints

Enable Prometheus

Useful RPC Queries

Snapshot Management

List Available Snapshots

Export a Snapshot

Load a Snapshot Archive

Restore from a Snapshot

Dump a Snapshot

Operational Commands

Rolling Back One Block

Pruning Historical State

Resetting CometBFT State

Running as a systemd Service

Running with Docker

Build docs developers (and LLMs) love

Get Started

Running a Node

Core Modules

Governance & Operations

Documentation Index

​Starting the Node

​Key Flags

​Full Example

​Syncing Strategies

​Health Check

​Flags

​Monitoring Endpoints

​Enable Prometheus

​Useful RPC Queries

​Snapshot Management

​List Available Snapshots

​Export a Snapshot

​Load a Snapshot Archive

​Restore from a Snapshot

​Dump a Snapshot

​Operational Commands

​Rolling Back One Block

​Pruning Historical State

​Resetting CometBFT State

​Running as a systemd Service

​Running with Docker

Build docs developers (and LLMs) love

Starting the Node

Key Flags

Full Example

Syncing Strategies

Health Check

Flags

Monitoring Endpoints

Enable Prometheus

Useful RPC Queries

Snapshot Management

List Available Snapshots

Export a Snapshot

Load a Snapshot Archive

Restore from a Snapshot

Dump a Snapshot

Operational Commands

Rolling Back One Block

Pruning Historical State

Resetting CometBFT State

Running as a systemd Service

Running with Docker