Recommended hardware: 4 CPU cores, 16 gb RAM, 50 gb disk.
Currently only Ubuntu 24.04 is supported.
Ports 8000 and 9000 are used for gossip and must be open to the public. Otherwise your IP address will be deprioritized by peers in the p2p network.
For lowest latency, run the node in Tokyo, Japan.
Save an initial set of initial peers to connect to, or use the public list:
curl https://binaries.hyperliquid.xyz/Testnet/initial_peers.json > ~/initial_peers.json
Configure chain to testnet. Mainnet will be available once testing is complete on testnet:
echo '{"chain": "Testnet"}' > ~/visor.json
For now, the non-validating node requires a configuration file. This will no longer be required in a future update when the node will verify the validator beacon chain from genesis:
curl https://binaries.hyperliquid.xyz/Testnet/non_validator_config.json > ~/non_validator_config.json
Download the visor binary, which will spawn and manage the child node process:
curl https://binaries.hyperliquid.xyz/Testnet/hl-visor > ~/hl-visor && chmod a+x ~/hl-visor
To start the visor, run ~/hl-visor. The node process will write data to ~/hl/data. The network will generate around 20 gb of logs per day, so it is also recommended to archive or delete old files.
Blocks parsed as transactions will be streamed to ~/hl/data/replica_cmds/{start_time}/{date}/{height}
State snapshots will be saved every 10000 blocks to ~/hl/data/periodic_abci_states/{date}/{height}.rmp
The state can be translated to JSON format for examination:
./hl-node translate-abci-state ~/hl/data/periodic_abci_states/{date}/{height}.rmp /tmp/out.json
Crash logs from the child process will be written to ~/hl/data/visor_child_stderr/{date}/{node_binary_index}