Skip to main content
Use systemd to run a non-validator full node in the background. This guide uses client 1.8.1-tavchain.eval1 from the earlier Linux installation, not the newer native bundle. No wallet, faucet claim, staking deposit or keyring passphrase is needed.
For a new installation, start with the installation chooser. Docker provides the current packaged background workflow on Linux x86-64. The unit and paths below belong only to this earlier native installation; do not substitute a new bundle or point them at its home without a matching service procedure.

Before you start

Complete the signature, fingerprint, archive and internal checksum checks in Run a full node. Start in the extracted tavchain-node directory after replacing its historical network files with the separately authenticated 771002 network package from that guide. You need a compatible Linux x86-64 host, systemd, sudo access and the prerequisites from that guide. The reference platform is Ubuntu 24.04 with systemd 255. This procedure creates a fresh service account and home. Do not use it to overwrite a validator, migrate an existing node or modify a shared Tavchain installation. If you have a foreground node using ports 29656/29657/29645/29646, stop it gracefully first. Leave its home intact; this new service starts its own history sync. Check those ports are free before continuing. The network is unaudited and centrally operated. Test TAV has no monetary value.

Install the binary and account

These commands stop on an existing binary path, account, state home or unit. Review any existing installation instead of overwriting it.
The service runs as tavchain-fullnode, not root. The versioned binary is root-owned; only the node home is writable by the service account. No operational validator or wallet keys are copied.

Initialize the fresh service home

Run initialization once, as the service account. Native initialization creates private P2P and unused consensus identity files. Keep them private. The account is not registered as a validator.
The configuration retains full history, connects outbound to the published peers and binds listeners to loopback. No public RPC or firewall opening is required. Network files and ports lists the exact configuration.

Create the service

Save the complete unit below. It runs the native client directly; no launcher script or upgrade manager is required.
Change --moniker my-tavchain-fullnode in your unit if you want another public node label. Keep the home, binary version and genesis pin consistent with this release.

Check status and sync

Press Ctrl+C to stop following logs; that does not stop the service. active (running) means the process started, not that it is synced. Check the node separately:
Expect native network tavchaintestnet_771002-1, advancing block height, at least one peer, catching_up: false after sync and EVM chain ID 0xbc3ba (771002). A non-validator full node normally has zero voting power. Compare a block hash at a fixed common height with the public RPC when checking agreement. Use Monitor a node for a complete fixed-height comparison and ongoing host checks. For startup failures, missing peers or restart loops, follow Troubleshoot a node.

Stop and restart

Use sudo systemctl restart tavchain-fullnode.service to perform both in sequence. Keep the same home; do not initialize again, replace genesis or delete data between starts. Stopping sends SIGTERM and waits for graceful shutdown. The unit does not force-kill a slow database shutdown. If stop remains pending, inspect the journal, process and disk before taking recovery action; do not start another process on the same home. An unexpected process failure triggers a restart after 5 seconds. Five starts within 120 seconds trigger systemd’s crash-loop protection. This controls crash-loop retries. After diagnosing and fixing the cause, use sudo systemctl reset-failed tavchain-fullnode.service, then start normally. An explicit stop does not trigger automatic restart.

Optional start after boot

To remove boot startup, use sudo systemctl disable tavchain-fullnode.service. Disabling alone does not stop an active service. After a host reboot, check service status, peers and sync before using the local RPC. Enabling a unit does not replace those readiness checks.

Operational checks

For planned maintenance, client changes and interruption recovery, follow Upgrades and recovery. Keep the same node home across ordinary restarts. Keep the node home and identity across starts. Monitor block progress, peer connectivity, free disk space and inodes. The service running successfully is only the first step; use Monitor a node to check readiness. For failed starts or repeated restarts, follow Troubleshoot a node. Do not reset the home or copy validator state as a routine repair.